Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

[P]
On the Need for "the Samba of Web Browsers"

By Caspian in Culture
Wed May 31, 2000 at 07:11:30 AM EST
Tags: Internet (all tags)
Internet

I am writing this story as a concerned Webmaster (and major Perl/PHP geek) and Linux/Unix sysadmin who believes in freedom of software choice. My topic today is something which I've been meaning to write about, but which events today have convinced me that I must speak up on immediately-- namely, the slow IE-ization of the Web, and what it means for the Unix community.

Today, I was directed to a particular Web site. Upon trying the page to which I had been pointed, I was greeted with merely a blue background. No actual content showed up in my copy of Netscape 4.73 for Linux.

Acting on a hunch, I fired up VMWare and tried the page in Internet Exploiter 5 for Windoze. My hunch was correct-- the page worked just fine there. So I fired off an e-mail to webmaster@(that domain), informing him of the URL of the page that doesn't work in Netscape.

The e-mail bounced.


We are entering a bold and rather frightening new era in the continual evolution of the Web, an era in which Netscape's long-held lead, now turned into a straggling second place, becomes ever more a thing of the past. While us geeks have turned a blind eye, Microsoft has been quietly adding "feature" after "feature" to IE, designed to break compatibility with Netscape. These are features that no other browser has-- things like support for odd graphics formats in inline graphics (for instance, .BMP), support for backslashes (!!!) instead of forward-slashes in <IMG> tags, and the like.

Since hardly anyone edits HTML by hand any more (I do, for the record, and a lot of those reading this might, but look at Joe Average Web Developer nowadays and see if he can), those browsing the Web are at the mercy of whatever garbagey HTML the latest GUI "Web editors" produce. Increasingly often, those "Web editors" are made by Microsoft-- namely, Front Page (in its various incarnations) and Word (which, while not a Web editor, can output HTML, and many people use it for this). For obvious reasons, code output by MS editors will very often render properly only in.. <GomerPyle>surprise, surprise, surprise...</GomerPyle> IE.

This, in and of itself, would not be a problem-- if only Web developers would test their pages in Netscape. But they don't. More and more often, new Web sites are being made in Front Page, Word or some other garbagey-code-spewing GUI editor. Then they are tested-- in IE for Windoze only. Then they are put online.

If you're finding yourself thinking, "Well, only the stupid teenaged "31337 d00dz" from AOL who've never run anything but Windoze do things like that", you're wrong. Just the other day, I practically had to hammer a friend of mine over the head with a clue-by-four to get him to make his site work properly with Netscape. He was using Word to output his HTML, and IE for Windoze to test the site with. When I told him that I wanted to see the page as he intended it to look, he told me that if that was what I wanted, I should download IE.

You would expect that sort of brash arrogance from a WinIdiot, but this fellow is no fool. He's a programmer, like me. In fact, he's a Unix geek. His server runs-- get this-- FreeBSD. In fact, he's a staunch FreeBSD advocate, and has been for years. Yet Windoze is making him piles of money through his job, and so he runs Windoze on the client end. And, like most people (smart or dumb, computer-literate or otherwise) who run Windoze, he neglects all those who don't do the same. When dealing with the Windoze community as an outsider, one has to remember that there are basically three sorts of Windoze users: The Masses (largely computer-illiterate AOLer types), the computer-literate folks like you and I who use Windoze because it's a good way to make piles of money, and the computer-literate folks like you and I who use Windoze merely because they are essentially forced to. The third group is dwindling rapidly. By and large, people who use Windoze are either too computer-illiterate to ever be likely to run anything else, or too busy raking in the dough to want to stop running Windoze. Either way, they don't give a flaming flip about anyone who doesn't have a Windoze box sitting on their desktop.

What we are facing here is a situation where most people run IE, and where those who do not, when faced with an increasingly large percentage of pages that require IE to function properly, are greeted with a shrug and a snide "so download IE." Likewise, those who protest that IE won't even run on their OS (e.g. GNU-Linux systems, NetBSD, OpenBSD, FreeBSD, et cetera) are likely to be greeted with yet another shrug of apathy.

Meanwhile, the pile of compatibility-clobbering "features" that solely IE offers has continues to rise. Meanwhile, Netscape is STILL a huge, lumbering beast of a process, so slow that some pages take seconds to render that render almost instantly in IE (one site I was developing had to be retooled, since even on a K6-2 500, pages were taking approximately 30 seconds to render! I managed to shave that time down to approximately 4 seconds with some simple HTML changes, but IE still beats it out, rendering both versions of the page nearly instantly), and so famously unstable that an entire small application exists whose sole purpose is to kill and restart an errant Netscape process. In fact, I'm currently typing this in my trusty copy of pico (yes, I know, I should move to vi or Emacs), since Netscape crashed-- instantly went <foom> and disappeared on me-- while entering my post directly on SlashDot's site.

The canonical response to any concerns about browsing under Unix is, of course, "So use Mozilla. A fine idea, in theory, but let's look at it in practice. Mozilla is:

  • Just as big and bloated as Netscape (remember where its codebase came from?)
  • Just as slow
  • Just as unstable
   ...And, last but DEFINITELY not least...
  • Just as incapable of handling the masses of new "IE-only" pages as Netscape is.


Another common retort is "so use Opera". Nope. I don't think so. A, Opera costs money. That's just ridiculous for a browser in this day and age. B, Opera is just as incapable of dealing with the IE-only crap as Netscape is.

What about Konqueror? Still under development, and requires KDE, and, I would imagine, just as useless at handling, say, an inlined .BMP as good ol' Netscape. What about Lynx? It's a fine browser, but even an old-schooler like me occasionally DOES like graphics.

While we've not even been paying attention, Microsoft has slowly been turning the Web from a platform-independent, browser-independent community into its own private playground, very much platform-dependent, and tied into a proprietary browser controlled wholly by none other than the Colossus of Redmond itself. Bill Gates's long-held aspiration of controlling the Web is getting closer to reality, as he is now virtually finished in conquering the critical client end-- the Web browsers that millions use to access the Web. Meanwhile, the Unix community has been lying like a slumbering beast, counting on the magic of Mozilla-- which is essentially just as bad as Netscape in all important aspects, except for the obvious fact that it is open-source/free software-- to give the Unix community the "next-generation" browser it so desperately needs.

It ain't gonna happen.

And, critically, I say "needs", since this is very much the truth. The people who run Web sites nowadays are increasingly not people like Kuro5hin readers. They are not people who run Unix. These are people born and raised on Microsoft, from DOS up through Win2K. They have no sense of the many time-honored traditions of the Net-- for instance, setting up a working "webmaster at WHATEVER dot WHATEVER" e-mail address for their sites. Even if you CAN manage to get a hold of such a person to complain that their site won't work with your copy of Netscape, you're increasingly likely to receive merely a pointer to http://www.microsoft.com/ie in return.

Netscape is crap. Mozilla is crap. Opera is crap. Konqueror is under development, and therefore unsuitable. Lynx isn't even in the same league, since it's not graphical. Meanwhile, Microsoft is getting further and further ahead in the race to add new obfuscating "features" to its HTML parser. What solution does this leave the Unix community with? Thus far, NONE.

At this point, it is highly unlikely that a stable, efficient browser could ever come out of the Mozilla project. Clearly, there is a distinct need for a new browser, coded from scratch and capable of handling all the W3C-rejecting crap that the Microsofties' HTML editors are throwing up onto the Web like so much half-digested brisket.

What we need, in short, is the Samba of Web browsers.

From the start, the Samba project has managed to keep pace with Microsoft's continual "embrace, extend, extinguish" behavior. Each time MS has one-upped the Samba team with yet another version-to-version protocol change, the venerated file- and print-sharing package has fought back valiantly, patching their code as often as necessary to ensure compatibility with even the very newest versions of Windows. Tutorials are on hand to help a Unix sysadmin set up a box with the package to communicate with any variant of Windows they could possibly encounter.

WHY IS THERE NOTHING LIKE THIS IN THE WEB BROWSING WORLD?

Like it or not, we all-- at least occasionally-- use the Web. And like it or not, Lynx alone isn't good enough for the vast majority of us. It isn't for me-- and I'm about as old-school as they come nowadays. The Unix community, in short, NEEDS to wake up and smell the coffee. If no browser can keep up with MS's continuous stream of new "features", it will soon be virtually impossible to use the Web under Unix.

And then, Microsoft's next step will be the server end...

Unix hackers of the world-- I beg you, I implore you. We NEED-- absolutely NEED-- a new browser project... one that is going to do it right. One that is going to, as Samba has, ceaselessly keep up with MS's constant compatibility-busting behavior. (I would start coding it myself, but I'm a complete neophyte when it comes to parsers and graphical coding.) We need this and we need this now-- or pretty soon, we'll all be running Windows on our desktop machines, merely because it would be impossible to use the Web-- except, perhaps, for sites like this one-- otherwise. And a (wo)man cannot live on Kuro5hin alone...

Sponsors

Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure

Login

Related Links
o Slashdot
o Freshmeat
o Kuro5hin
o PHP
o VMWare
o AOL
o pico
o vi
o Mozilla
o KDE
o Lynx
o Samba
o Also by Caspian


Display: Sort:
On the Need for "the Samba of Web Browsers" | 157 comments (157 topical, editorial, 0 hidden)
Excellent rant^H^H^H^Hwriteup, and,... (3.00 / 2) (#19)
by Qtmstr on Wed May 31, 2000 at 01:13:55 AM EST

Qtmstr voted 1 on this story.

Excellent rant^H^H^H^Hwriteup, and, for once, it's not a link to an article, but an article inof itself. I like the fact that you point out that even a *NIX geek can turn a blind eye to interoperability if he is sufficiently blinded by little green slips of paper. I do have one question, though: Why do you think it is doubtful that stable browser could come out of the Mozilla project? Do you realize that Mozilla is not even technically beta yet, and already it is better than many betas I have seen. Some people even manage to use it for everyday browsing.


Kuro5hin delenda est!

Re: Excellent rant^H^H^H^Hwriteup, and,... (3.00 / 2) (#26)
by squigly on Wed May 31, 2000 at 08:18:05 AM EST

even a *NIX geek can turn a blind eye to interoperability if he is sufficiently blinded by little green slips of paper.

I don't think all *n[ui]x geeks would though. Some might feel that if they were being paid that generously, they have an obligation to make sure the site is correct and as visible as possible. Its a professional obligation.

--
People who sig other people have nothing intelligent to say for themselves - anonimouse
[ Parent ]
*very good* even it the header's du... (1.00 / 1) (#13)
by ishbak on Wed May 31, 2000 at 01:25:09 AM EST

ishbak voted 1 on this story.

*very good* even it the header's duplicated.

I agree. One of the main reasons I... (2.50 / 2) (#15)
by Marcin on Wed May 31, 2000 at 01:25:44 AM EST

Marcin voted 1 on this story.

I agree. One of the main reasons I don't use Linux or FreeBSD as a desktop environment is because frankly there aren't any decent web browsers for it. Even not using an IE specific features IE renders faster and better looking pages than Netscape or anything else (as was mentioned).

I think a major part of the problem isn't non-standard HTML, because there isn't so much of that anymore as far as I am aware, but instead non-standard javascript between Netscape and IE. Sure the two main browsers render pages differently but people that spend ages trying to make the two look the same are missing the main point of HTML.. it's not meant to be a page layout language! (even though people will try to convince you otherwise).

Why does there need to be a new browser project? What's wrong with taking one of the existing ones (like say Konqueror) and making it the "Samba of Web Browsers"?

I look forward to more interesting discussion on this subject ;)
M.

Re: I agree. One of the main reasons I... (4.00 / 1) (#46)
by Paul Dunne on Wed May 31, 2000 at 12:28:16 PM EST

Bzzzzt! Wrong. If you don't like Linux or FreeBSD, don't use them; but don't build strawmen ot support your decision (it doesn't *need* support, it's your computer and your life). There are plenty of decent broswers for Linux, if by "browser" you define "a program to look at WWW pages using the HTTP protocol". Of course, if you define "browser" as "IE", well, since IE doesn't run on Linux, your conclusion follows -- but only then.

The real trouble with this page layout issue is, people want to get on the web and do ... page layout! They will try to force HTML to do this for them. It's HTML "fault" in a sense if it can't help them to do this. "The sabbath was made for man, not man for the sabbath".
http://dunne.home.dhs.org/
[ Parent ]

damn straight. However, Mozilla ai... (4.00 / 1) (#8)
by RJ11 on Wed May 31, 2000 at 01:30:45 AM EST

RJ11 voted 1 on this story.

damn straight. However, Mozilla aims to accomplish these things, there's just one problem: it lacks organization. That's its only problem. the problem is that new features are being added before old bugs are getting fixed, there is no project leader and everyone just adds the cool new features that he/she pleases, if they just had some organization, Mozilla could be the best browser out there.

Bugzilla (none / 0) (#132)
by JJC on Thu Jun 01, 2000 at 11:41:45 PM EST

No organisation? Check out Bugzilla. That's a very sophisticated organisation if you ask me.

[ Parent ]
This topic is about two or three ye... (3.00 / 1) (#5)
by HiRes on Wed May 31, 2000 at 01:38:22 AM EST

HiRes voted 1 on this story.

This topic is about two or three years too old.

If you're worried about IE (which is a fantastic browser, BTW) forcing more proprietary technologies down our throats, join the WaSP or write your own browser.

Me, I'll just keep using Netscape, Mozilla, Lynx, or W3M, until M$ releases an IE for *nix that is largely W3C-compliant. And when I can't view a web page, I'll complain to the webmaster and send him the WaSP link, and more importantly, tell them I won't do business with them or visit any of their sponsors.
--
wcb
wait! before you rate, read.

Okay, I was willing to live with th... (4.00 / 6) (#3)
by analog on Wed May 31, 2000 at 02:16:03 AM EST

analog voted -1 on this story.

Okay, I was willing to live with the gratuitous misspelling of Windows (playing games like that may be fun, but it kills your credibility when you actually have something to say). A lot of what you say makes a great deal of sense. However, you're displaying an outstanding level of ignorance about Mozilla, and since it turns out your whole premise depends on your Mozilla 'facts' being straight, it's gotta be -1.

Uh. you like to bash oodles of prer... (2.33 / 3) (#6)
by dieman on Wed May 31, 2000 at 02:22:36 AM EST

dieman voted -1 on this story.

Uh. you like to bash oodles of prerelease software. wow. thats a new one!
---
blah

I get the sneaking suspicion that t... (4.00 / 4) (#17)
by Prospero on Wed May 31, 2000 at 02:25:01 AM EST

Prospero voted 1 on this story.

I get the sneaking suspicion that this guy hasn't used Mozilla recently. M15 is at least as stable (if not more so) than Netscape 4.73, and are comparable in speed and size. Given that the milestone builds still include debug code, they promise to be smaller and faster. Besides, if a website starts using MS-HTML, and their site becomes unreadable, BOYCOTT THE SITE. Vote with your feet, and refuse to use their services, or buy their products. As it stands, I already boycott Java/Javascript/graphics intensive websites, simply because I can't stand flashy web design. I can't say this decision has drastically affected my browsing choices. The web is such a big place that there is no reason to accept mediocrity. If you desperately want `IE compatibility` (and I still dont see why you would, but different strokes, etc...), build it into an existing project, and have a checkbox to turn on the functionality. The mozilla codebase is a complete rewrite, and is, by all accounts, a much easier study than the old code. Don't reinvent the wheel when a perfectly good wheel already exists.
... and never, ever play leapfrog with a unicorn.

The web constantly grows, so this i... (1.00 / 1) (#22)
by OscarIommi on Wed May 31, 2000 at 02:41:25 AM EST

OscarIommi voted 1 on this story.

The web constantly grows, so this is alwys a hot issue..

Sites made in those WYSIYG editors ... (3.00 / 2) (#12)
by Commienst on Wed May 31, 2000 at 02:43:56 AM EST

Commienst voted -1 on this story.

Sites made in those WYSIYG editors are not worth really visiting in my opinion. How much time and effort does someone who is too lazy to learn the most basic or html commands anyway going to put into their site to make it worth visiting anyway?

Konqueror does not require KDE it requires the KDE shared libraries. Dont want to use KDE fine install Konqueror and the libraries and run it Konqueror by itself.

Re: Sites made in those WYSIYG editors ... (3.00 / 1) (#39)
by mattc on Wed May 31, 2000 at 10:52:51 AM EST

If you use Dreamweaver, it makes some decent looking HTML output -- it even has an option to re-format messy HTML.

Frontpage is a pile of garbage though, especially if people use it to upload to your web server.. it adds a whole bunch of directories named vnt_cnf or something similar which contain copies of all the images?!?

[ Parent ]

It would be nice and neat and fit w... (4.60 / 5) (#1)
by rusty on Wed May 31, 2000 at 03:07:56 AM EST

rusty voted 1 on this story.

It would be nice and neat and fit with many of our preferred world views nicely if this were true, but unfortunately, it isn't. The reason Netscape can't render some pages is almost always because Netscape is broken. It has bugs. Bad, evil ones that are the result of a crufty hacked up codebase, which ought to be drug out behind the barn and shot. I'm sure IE has bugs too, but they are not as bad as Netscape's, and in almost all ways, it is a far superior product. IE is more standards-compliant than Netscape. IE is more flexible, faster, and tighter than Netscape, which makes it able to easily render pages which will choke Netscape with their complexity, despite being totally valid HTML. The simple fact is, Netscape is garbage. Case in point: I actually never test K5 in IE, because I know what should work, and if it should work, it works in IE. It often doesn't work in Netscape, so I do all my testing there, pretty much. Mozilla and IE, BTW, render identically, usually.

As for MS "embracing and extending", I don't have any evidence of that, anymore. They've killed Netscape, in a war of attrition in which both sides were guilty of introducing proprietary "features" into HTML, and Netscape probably more so than MS. Netscape's javascript DOM for example, is one of the most braindead things I've ever tried to code with. Maybe MS allows inline bitmaps and backslashes in image paths. I've never seen any responsible website use those things. If you see that on a site, forget about the content, because frankly, who cares what those idiots have to say. Other then that, what are the evil MS-only features?

And now a word about Mozilla. I think you unfairly dismiss the mozilla project. They do not share code with netscape-- they ditched the legacy codebase a year into the project and did exactly what you suggest-- they started from scratch. That's why it's taking so long, mostly. Secondly, the milestone releases are huge and slow because all the debugging is enabled. Production releases should be far faster and smaller. I think Mozilla will achieve what you (and all of us) desire; a real, efficient, standards-compliant browser, and on top of all that, it'll even provide a portable application base, which can be totally customized by a remote application service provider. That will simply change everything.

So, while I sympathize with your rant, I think your ire is off target. I recommend you assist with Mozilla in any way you can-- testing, debugging, anything. They need help, and people who care are the best ones to help.

____
Not the real rusty

I agree a major problem, and a head... (3.00 / 1) (#20)
by martin on Wed May 31, 2000 at 03:07:58 AM EST

martin voted 1 on this story.

I agree a major problem, and a headache to cope with the way different browsers render pages! About time the HTML 'standard's were the standards and not how a vendor interprets them. Even worse is WML support, its a mess and your WML might work on one phone/browser but another. It's almost impossible to make WML that works on all phones/browers.

Wrong, wrong, wrong, wrong and wron... (2.00 / 2) (#14)
by marm on Wed May 31, 2000 at 03:28:51 AM EST

marm voted 1 on this story.

Wrong, wrong, wrong, wrong and wrong. :-) Good troll though ;-)

Preach on, brother! I remember whe... (3.00 / 1) (#7)
by buzzbomb on Wed May 31, 2000 at 03:56:24 AM EST

buzzbomb voted 1 on this story.

Preach on, brother! I remember when I coded my first homepage and IE was OUT. It was a piece of shit and WTF is/was "J"? Now, I code for both. But Netscape is my primary. However, you do have a point. It's just too damn bad that M$ won't follow the rules and keeps adding proprietary tags. BMP in IMG tags? I can deal with that. (Any asshole who uses those huge, bloated files on a webpage should be flogged...at least it's a standard tag with a non-standard file format.) Frontpage (or whatever) can have those proprietary options and it would be fine in some cases, (company intranet, everyone running IE without a doubt) but it would be ideal if those programs came pre-configured with strict HTML 3 or 4 or whatnot. Of course, Outlook should come pre-configured not to use JS/ActiveX/fluffy shit on mail messages too. :) But fact of the matter is this: if this site you are talking about was worth a shit, the webmaster would have coded for both. Chances are that it's not a good site now and won't amount to shit in the long run.

Way too long of a rant. Could have... (2.00 / 1) (#16)
by duxup on Wed May 31, 2000 at 04:00:14 AM EST

duxup voted -1 on this story.

Way too long of a rant. Could have been done much more efficiently in one or two paragraphs. Let alone the fact that the solution amounts to: "Hey, someone else make (insert your idea) OSS project to solve (insert problem)." It's just not that simple. The problem of IE centric web development story has been much better written about and discussed before, nothing new here.

Re: Way too long of a rant. Could have... (3.00 / 3) (#35)
by bmetzler on Wed May 31, 2000 at 10:18:17 AM EST

Way too long of a rant. Could have been done much more efficiently in one or two paragraphs.

Well! I've enough people complain that the write-up was too short. But this is the first time I've ever heard someone complain that the write-up was too long. :)

-Brent
www.bmetzler.org - it's not just a personal weblog, it's so much more.
[ Parent ]
Re: Way too long of a rant. Could have... (2.50 / 2) (#48)
by Paul Dunne on Wed May 31, 2000 at 12:29:24 PM EST

Of course if he'd boiled it down to one or two paragraphs, there'd be half a dozen people complaining it was too short. Kuro5hin benefits from original content. This piece is good enough to stand as a full-length article, and kudos to Caspian for writing it.
http://dunne.home.dhs.org/
[ Parent ]
There is a perfect browser availabl... (4.00 / 1) (#18)
by feline on Wed May 31, 2000 at 05:07:59 AM EST

feline voted 1 on this story.

There is a perfect browser available, doesn't take up a lot of memory, graphical, no zombie processes, development team that keeps up with the latest html, it's called IN YOUR DREAMS.

The truth is, getting a good web browser in unix or any os for that matter, is the impossible dream, as plausible as perpetual motion.

I say, we just mishmash code from every browser we can get our hands on into one.

As far as the web-developers developing only for ie, ie is the major web browser, there's nothing anyone can do about it, it's gained too much market share for any web-developer to really give a rats ass about any of us.
------------------------------------------

'Hello sir, you don't look like someone who satisfies his wife.'

I couldn't agree more with you in g... (3.00 / 1) (#11)
by jackyb on Wed May 31, 2000 at 05:46:32 AM EST

jackyb voted 1 on this story.

I couldn't agree more with you in general. I have some specific points to add, however.

Firstly, I use Windows for choice - and I'm certainly computer literate, and I don't earn any money from it. That's because the applications I want to use are available on it; I'd love to be competent on Linux as well, but it will have to wait for my second box.

Secondly, there are a lot of things wrong with IE, and god knows that I'd like it to finish supporting W3C standards before it implements yet another buggy feature - but fundamentally it is pretty quick, pretty, and the UI is good. Netscape, on the other hand, I can't stand. (Nor Mozilla, either) It's a shame that we can't persuade Microsoft to develop a version for Linux.

Lastly, I code my HTML by hand, and even if I stick to the W3C standards, the only browser that it renders half-well in is IE. Mozilla, arguably, does a better job, but since it doesn't support Javascript, I can't do everything I want to do in it. The bad old days of coding vastly different page sources should GO! All the browser manufacturers should comply strictly with the latest W3C standards as a minimum.



Re: I couldn't agree more with you in g... (4.00 / 1) (#58)
by Anonymous Hero on Wed May 31, 2000 at 02:21:08 PM EST

Mozilla, arguably, does a better job, but since it doesn't support Javascript

Stop right there. I don't know where you got this idea. Check under (this works with M15) tasks/tools/ where you will find the Javascript console.

Mozilla supports JavaScript and the W3C DOM in a MAJOR way ... that's the main point of supporting CSS-P ....

[ Parent ]

Re: I couldn't agree more with you in g... (none / 0) (#65)
by jackyb on Wed May 31, 2000 at 03:34:39 PM EST

Well, to be fair I haven't looked at the last couple of milestone builds; I'm happy using IE for my day to day browsing. I didn't know that Javascript support was now present. But is the DOM the same as Microsoft's? (I should put that, is MS's DOM W3C compliant?)

[ Parent ]
First off, I think you've put forth... (3.00 / 1) (#2)
by driph on Wed May 31, 2000 at 05:57:00 AM EST

Driph voted 1 on this story.

First off, I think you've put forth a great idea. I would LOVE it if someone came out with a 3rd party fully compliant browser. Unfortuntely, with a couple minor exceptions, we are all forced to deal with NS and IE.

And I've gotta tell you, I LOATHE Netscape. I'm not a fan of IE-only tags, but dammit, It'd be nice if Netscape actually supported the official ones. I can't think of anything better to do on a Friday night than build workarounds because NS doesnt want to support this CSS feature or that tag.

--
Vegas isn't a liberal stronghold. It's the place where the rich and powerful gamble away their company's pension fund and strangle call girls in their hotel rooms. - Psycho Dave

So basically, you want a web browse... (2.50 / 2) (#10)
by renaud on Wed May 31, 2000 at 06:14:06 AM EST

renaud voted -1 on this story.

So basically, you want a web browser which ignores these things called 'standards' so that you can visit all those web servers designed with products like FrontPage, and see all the nifty animations ? There is no real standard for SMB (and DCE/RPC over SMB), so Samba is a Good Thing. There are standards for the web though, so torturing them will not help. You'll play the same game as Microsoft, and this is a game you can't win. So not only we would end up with a web browser which mocks all the standards, but all the new features that only run under IE would be implemented months after. So we'll always be a step behind. (oh, and IE is somehow bloated too. You don't notice that because it's loaded at startup as a part of the NT 'kernel')

Sure, it might be good for discussi... (3.00 / 3) (#9)
by soulhuntre on Wed May 31, 2000 at 06:14:18 AM EST

soulhuntre voted 1 on this story.

Sure, it might be good for discussion. However, there are a few points to consider: 1) We have the W3C. If people really care about that standard, the market will provide a browser that handles it. 2) The MS'isms DO serve a valid purpose - intranets. In a company internal net I KNOW what the browser is, and there is no reason for me to NOT use extensions specific to that browser to enhance my information presentation. There is NO REASON for MS to abandon those features simply because lazy webmasters will misuse them. 3) The Netscape problem is of it's own doing. Current, I code pages strictly to the W3C specs - and Netscape blows chunks most of the time (the older NS anyway) - that is >NOT< Microsofts fault. 4) The issue with BMP support is a red herring. If Mozilla supported say - PNG (it does I think) then will you now demand that PNG be held back because all browsers can't do it yet? So MS added support for a format common in intranets - and a compressible, lossless 24 bit format at that, this is no crime. Ken

Re: Sure, it might be good for discussi... (4.00 / 1) (#34)
by gleef on Wed May 31, 2000 at 10:16:05 AM EST

Soulhuntre wrote:

2) The MS'isms DO serve a valid purpose - intranets. In a company internal net I KNOW what the browser is, and there is no reason for me to NOT use extensions specific to that browser to enhance my information presentation. There is NO REASON for MS to abandon those features simply because lazy webmasters will misuse them.

The problem with the MS'isms isn't that IE supports them, its that Front Page and Word produce them by default. The most common tools used to develop websites produce non-standard HTML that is incompatible with anything but IE. It takes extra effort to make a compatible web site, and few people bother.

[ Parent ]

Re: Sure, it might be good for discussi... (none / 0) (#89)
by soulhuntre on Thu Jun 01, 2000 at 12:49:41 AM EST

"The problem with the MS'isms isn't that IE supports them, its that Front Page and Word produce them by default."

Actually, this is only partially true. Certainly the Word export does this, but MS provides a tool to convert that output to normal HTML.

Further, frontpage does NOT generate any IE specific code by default.



[ Parent ]
Re: Sure, it might be good for discussi... (4.00 / 1) (#37)
by YellowBook on Wed May 31, 2000 at 10:33:21 AM EST

Sure, it might be good for discussion. However, there are a few points to consider:
  • We have the W3C. If people really care about that standard, the market will provide a browser that handles it.

We do have the W3C, and the standards it puts out tend to be well thought-out. With HTML 4.0 Strict and CSS2, you can do everything that you could do with fancy table or frame driven layout in HTML 3.2 or HTML 4.0 Transitional, but have it look nice in Lynx or w3m, too. The only problem is that the browser vendors are unwilling to implement these standards. Probably more importantly, the web design software implementers aren't willing to implement these standards. What the world really needs is a Pagemaker-like web-editing package that does all its layout using CSS2 absolute-positioning and encourages the use of standard structural markup (by providing good stylesheets for headers, etc, and making it easier to define and apply styles than to insert inline styles).

  • The MS'isms DO serve a valid purpose - intranets. In a company internal net I KNOW what the browser is, and there is no reason for me to NOT use extensions specific to that browser to enhance my information presentation. There is NO REASON for MS to abandon those features simply because lazy webmasters will misuse them.

I'm not so sure about this one. In one sense it's true, but it's far too easy for practices like that to escape the intranet. Also, you have to consider what kind of software the internal net content designers are going to be using to lay out pages -- probably the same ones they'd use for external pages, or else why have an intranet at all (instead of some proprietary internal network)?

  • The Netscape problem is of it's own doing. Current, I code pages strictly to the W3C specs - and Netscape blows chunks most of the time (the older NS anyway) - that is >NOT< Microsofts fault.

I do agree with this (to some extent). Not to let Microsoft off the hook, but the Netscape 4.x series has been abysmal. I wanted to move some of our pages from a crufty, table-driven design which is rather hard to generate from CGI, to a CSS2-based layout which would also let me rearrange elements so that they are in sensible places in Lynx (basic navigation at the top, content near the top through the rest of the page, extended navigation/menus at the bottom). Looks great in Lynx, looks great in Mozilla (pre-M16 nightly builds), looks fine in IE...looks like crap in Netscape 4.x.

I do want to take issue with the article's author's description of Mozilla. Mozilla is much smaller and faster than Netscape; the rendering engine in particular is much, much faster on heavily table-driven sites like TOS. A full download of Mozilla is about half the size of a full download of Netscape Communicator. I've been using nightly builds for my daily browsing post M15, and the only places I find it lacking are authentication (basic-auth doesn't seem to work at all, and I haven't yet been able to test digest-auth), and ssl (and I don't think the basic Mozilla builds include https support at all -- you'll have to get that in Netscape 6.0, at least until September...). There are a few places where a feature is not implemented (such as the "pick app" option on a download menu). But the most important thing about Mozilla, and the reason I so much prefer it to Netscape, is CSS support. If Netscape can really push their Mozilla-based Netscape 6, they can take back market share from IE, especially with the upcoming Microsoft breakup. Widespread use of Netscape 6 would encourage wider use of CSS for more than pretty details, but make it central to site design.

IE is not without its faults. It does funny handling of backslashes that can really screw up the POSTed data that gets to CGI scripts. I have one script where a particular page won't work on IE because of this misbehavior. I've been trying for the last few minutes to find a link that documents this, but I seem to have lost it. And of course, IE dependence is a very bad thing.

  • The issue with BMP support is a red herring. If Mozilla supported say - PNG (it does I think) then will you now demand that PNG be held back because all browsers can't do it yet? So MS added support for a format common in intranets - and a compressible, lossless 24 bit format at that, this is no crime.

I don't disagree here. A modern browser ought to support image formats through some generic image-loading library like gdkpixbuf or imlib2. Using Windows BMP files on an intranet is pretty silly though --- maybe gzipped BMP files would be sane.



[ Parent ]
Interesting idea. I disagree with ... (4.50 / 2) (#21)
by Paul Crowley on Wed May 31, 2000 at 06:21:33 AM EST

Paul Crowley voted 1 on this story.

Interesting idea. I disagree with your criticisms of Mozilla though; it isn't bloated (the core is very small, all the size comes from optional pluggable extras that provide specific functionality that some people want) and by the time it's released it'll be fast and stable (why do you think they call it beta?). Nearly all of the original Netscape code has been left behind. It's possible that Mozilla's wonderful extensible architecture would make it possible to add modules that try and do what you suggest, to track these stupidities and give us web sites we can see. But no browser will cure web designer stupidity. How do you impart a clue to Web designers who ask how to make sure their images turn up even when people have turned off image loading, and achieve it (with serious intent) by using tables full of tiny coloured cells?
--
Paul Crowley aka ciphergoth. Crypto and sex politics. Diary.

Another large difference between IE... (4.00 / 1) (#4)
by Pasco on Wed May 31, 2000 at 07:11:30 AM EST

Pasco voted 1 on this story.

Another large difference between IE and Netscape is that Netscape (4.7 anyway) is a heckva lot stricter with it's HTML parsing. Surprisingly often pages that don't load properly in Netscape have actually been written by hand, but only tested in IE - and IE doesn't particularly care if you haven't closed your tables or whathaveyou. Who is in the right here? Netscape - forcing correct code, or IE - giving allowances for the odd mistake?

Re: Another large difference between IE... (4.00 / 1) (#42)
by YellowBook on Wed May 31, 2000 at 11:04:09 AM EST

Another large difference between IE and Netscape is that Netscape (4.7 anyway) is a heckva lot stricter with it's HTML parsing. Surprisingly often pages that don't load properly in Netscape have actually been written by hand, but only tested in IE - and IE doesn't particularly care if you haven't closed your tables or whathaveyou. Who is in the right here? Netscape - forcing correct code, or IE - giving allowances for the odd mistake?

I'd be much cooler with this if Netscape did a better job of rendering standard-compliant, validated HTML. I write HTML by hand in a validating SGML editor (psgml-mode for Emacs), and would like to do all my presentation by CSS. Netscape's bogus CSS implementation means I can't do that, so I have to stick to HTML 4.0 transitional with the odd Netscape "extensions" like the "topmargin" attribute of tables. I really hope Mozilla is widely adopted.



[ Parent ]
Grrr.. (4.00 / 1) (#23)
by genehack on Wed May 31, 2000 at 07:25:24 AM EST

Off Topic: Hey, Rusty, the "this story was posted while you were voting, so now your comment has vanished" thing really kinda stinks. Nothing like encouraging me to just set the pop-up and mash the button without much reflection.

On Topic: My vote was going to be +1 -- broken down into +2 for the rant and -1 for not mentioning some of the other browsers already out there that are starting from scratch, such as Gzilla/Armadillo.

Mea culpa (3.00 / 1) (#53)
by rusty on Wed May 31, 2000 at 01:18:16 PM EST

I know, it sucks. :-( I'm really sorry about that. Thank you for complaining though.

____
Not the real rusty
[ Parent ]
Re: Mea culpa (4.00 / 1) (#61)
by genehack on Wed May 31, 2000 at 02:49:22 PM EST

You're welcome. 8^)= Thanks for listening, and running K5, and all that.

Say, now that I'm not as annoyed, a potential kludge^Wsolution presents itself: I presume that some error condition triggers the "thanks for playing but you missed out" page redirect -- when that triggers, just include the text of the comment in the page. Sure, you'll lose HTML markup, but that's trivial to fix -- easier than retyping the whole comment.

[ Parent ]

Re: Mea culpa (4.00 / 1) (#63)
by rusty on Wed May 31, 2000 at 03:05:06 PM EST

Well, rather than kludge, I might as well just fix it. There is such a routine, and the right fix would be to have that bit also take your comment and stuff it into the normal comments table, before it tells you the story's already been submitted. I have no idea why this didn't occur to me the first time I wrote that :-/

____
Not the real rusty
[ Parent ]
Re: Mea culpa (none / 0) (#70)
by genehack on Wed May 31, 2000 at 04:15:36 PM EST

Actually, I had sort of ass-u-me-d that there was a Good Reason that it didn't happen like that.

Hey, we all have our off days, eh?

[ Parent ]

While you're at it... :) (none / 0) (#83)
by Lionfire on Wed May 31, 2000 at 09:18:20 PM EST

So long as you're fixing the comment-posting code, is there any chance of stuffing a preview in for the comments we leave while voting? It's quite needed by people who can't type properly (like me  :)

[ blog | cute ]
[ Parent ]
If they make a site that Netscape can't display . (3.00 / 1) (#24)
by Anonymous Hero on Wed May 31, 2000 at 07:26:12 AM EST

.. then they don't need my eyeballs I suppose.

Most publishers won't fool 30% or more of their audience.

Re: If they make a site that Netscape can't displ (4.00 / 1) (#25)
by squigly on Wed May 31, 2000 at 08:13:07 AM EST

A large proportion of those 30% also have access to IE. If a site is good enough, they will return with IE. All the administrators see is hitcounts, so they think thats all that matters.

--
People who sig other people have nothing intelligent to say for themselves - anonimouse
[ Parent ]
I read somewhere... (4.00 / 1) (#43)
by Anonymous Hero on Wed May 31, 2000 at 11:43:50 AM EST

That IE is getting very close to the 90% mark... and it's increasing every day. So that's only 10% now, and going down... Soon you'll get major sites saying "oh well, sod the 4% of netscape users" and making their site IE-only... Daniel

[ Parent ]
Netscape, it's crap anyway........................ (2.00 / 1) (#27)
by Anonymous Hero on Wed May 31, 2000 at 09:15:53 AM EST

It is, I work for a company that designs WWW sites/pages/etc, Netscape is the biggest problem, what displays well in IE/ Opera/ ICAB/ Voyager/ and most others, doesn't display properly in Netscape, oh well, use a better browser................................

a.thyer@ntlworld.com

Aol, er I mean, AOL, er I mean Netscape.. what Net (2.00 / 1) (#57)
by Anonymous Hero on Wed May 31, 2000 at 02:12:48 PM EST

I work for a very big company working a new web
based project. We, thank god, dismissed Netscape
support altogether and will be required IE 5.0 or
better which means I can do some simply awesome
JScript and VBScript programming.

Netscape treats HTML like some sort of paint-by-number
text that is rendered into a page, then handles the
scripting half-assed.

IE treats the page like a program. IE 5.5 is extremely
object oriented allowing you to give an id to any object,
create a new instance, refer to its parent, child, move
children. I'de love to show you what I can do, but am
not allowed.

In essence, these "features" you complain about are
really innovation. No. I'm not a MS bitch, but I think
Netscape is dead. I'll repeat that:

NETSCAPE IS FRIGGEN DEAD FOLKS!

AOL bought em, raped em, and left a few developers
to hack in Ads into an open-source project (Mozilla)
that AOL, er AOLNetscape doesnt have to pay a cent
for. Look at Netscape 6.. its Mozilla M15 with Ads.

So, as a profressional web programmer, I don't
see Netscape as a player. I look at my logs and
98% of hits are IE. Your damn right I'm going to
use IE specific features because of the awesome abilities.

Programmers of awesome sites have only 2 real options:
program in scripting languages and HTML and suffer
incompatibility or use Flash, and I'm not selling out
to plug-ins yet.

I like this Samba idea, but don't piss on MS.
They're doing their job. I'de prefer that to having
them sit on their asses waiting for NS to catch up.

-Z

[ Parent ]
FrontPage and Word assumption (4.00 / 2) (#28)
by error 404 on Wed May 31, 2000 at 09:34:52 AM EST

I assumed that the HTML generated by FrontPage and Word would favor IE. It is a very reasonable assumption.

However, that, in my experience, is not the case unless you use the extensions, which are clearly marked. I use FrontPage 98 for first drafts of templates, and the HTML it generates is crap in any browser. I've had FP generate pages that worked in Opera but not in IE. I've also used the Word export, which is unbeleivably bad. But when you are maintaining company junk for use within the company, and you have a whole lot of existing Word documents that you have to make available, well...

Under Windows, Opera is my primary test browser, because a page that works under Opera will probably work under both IE and Netscape. Except for CSS - Opera has much better CSS support then either of the others, so the fact that a CSS chunk works perfectly under Opera means only that it is compliant, not that it will work under IE or Netscape.

I guess I can't get too worked up over the specific complaints.
Supporting the backwards slashes isn't all that hard, nor does it break any standard functionality. And it makes the browser able to look at local files on an MS box. Browsers should just support it. HTML authors should use the slash that God decreed, but the browser should tolerate the infidel.
And what's wrong with a browser supporting stupid file formats? BMP is obviously a stupid format for the web (for those unfamiliar with BMP, it is a plain, uncompressed bitmap. For an example of the file sizes - my camera stores between 12 and 18 JPG shots on a floppy, or one BMP) but for an intranet or for dealing with files on your own box, BMP has uses.

Maybe FrontPage 2000 is more IE specific - I've heard that it no longer trashes good HTML just because you opened it - but I have a certain fondness for FP98. You see, it introduced me to modern open source software. I needed a FP98 module that I hadn't installed, so I re-ran the setup program, which destroyed my Win95 system. Reinstalling Win95 took so long (I spent two weeks worth of evenings and weekends fighting with it) that, out of sheer desperation to get back online, I went out and got a RedHat CD. It took me almost half an hour from poping the wrap to downloading my mail, because I was being careful and reading the manual.

I feel much better now...

..................................
Electrical banana is bound to be the very next phase
- Donovan

Re: FrontPage and Word assumption (4.00 / 1) (#79)
by David Gerard on Wed May 31, 2000 at 07:28:18 PM EST

However, that, in my experience, is not the case unless you use the extensions, which are clearly marked. I use FrontPage 98 for first drafts of templates, and the HTML it generates is crap in any browser. I've had FP generate pages that worked in Opera but not in IE. I've also used the Word export, which is unbeleivably bad. But when you are maintaining company junk for use within the company, and you have a whole lot of existing Word documents that you have to make available, well...

Microsoft products do not necessarily have anything to do with each other. I've had cause to use Frontpage Express (the freebie with IE 4) to massage the INSANELY CRAP HTML oozed by Word 97. And you know, FP Express's output was quite decent and nicely-formatted.

They could do it right ...

[ Parent ]

"They could do it right..." (none / 0) (#105)
by error 404 on Thu Jun 01, 2000 at 10:23:56 AM EST

That is the really sad part. Microsoft could, if they wanted to, create fabulous software. But would an excellent version sell any more copies than what they have now?

..................................
Electrical banana is bound to be the very next phase
- Donovan

[ Parent ]
Browsers and IE... (5.00 / 3) (#29)
by bmetzler on Wed May 31, 2000 at 09:49:57 AM EST

ARG!!

You did a good job with the length of the write-up. However, I now feel I've got to whack you over the head with the clue-by-4 a few times.

*whack* *whack* *whack*

There, I feel a lot better now!

Anyways, here's what I have to say about your mozilla comments.

    * Just as big and bloated as Netscape (remember where its codebase came from?)

Nope, I don't remember where it's code-base came from. Perhaps you can refresh my memory. FWIW, Mozilla is a component based toolkit written from scratch.

    * Just as slow

Maybe, maybe not.

    * Just as unstable

Again, hard to say. It's hard to make a conclusive judgement of something you don't have yet. Mozilla is still in development. It'll be optimized and cleaned up later, when most of the features are in. Only at that time will we know how slow, and how unstable it is. I know that the testing I've done on Windows with it against IE, it is at least as fast, and crashes much less. And you can only suppose that it'll get better.

    * Just as incapable of handling the masses of new "IE-only" pages as Netscape is.

This is probably the worst claim. Why should IE determine compatibility when we already have the w3c? Let IE play by the rules :) We *don't* need to keep trying to play catch-up to a convoluted MS standard that MS can change, and change, and change and then protect with the DCMA. No, let MS support the w3c if they want.

On my end, I just test with Mozilla. If it works there, then that's all I need to worry about.

-Brent
www.bmetzler.org - it's not just a personal weblog, it's so much more.
Re: Browsers and IE... (3.33 / 3) (#78)
by Hurst Dawg on Wed May 31, 2000 at 07:26:12 PM EST

This is probably the worst claim. Why should IE determine compatibility when we already have the w3c? Let IE play by the rules :) We *don't* need to keep trying to play catch-up to a convoluted MS standard that MS can change, and change, and change and then protect with the DCMA. No, let MS support the w3c if they want.

Theoretically thats how it should work, too bad its not how it really does work. The vast majority of people know nothing about w3c or what it does. And an even larger majority don't care if they do know or wouldn't care if they don't, they just want the programs they run to work. And if they are running windows and browsing with IE on web pages that are made _for_ IE then everything will work, and they could care less. Its when MS starts adding "features" that make their favorite sites not work that they will start to complain. This starts to form a vicious (sp?) circle though as web"masters" start coeding for the IE crowd, MS changes their HTML, they use the new non-netscape-compatible features, and the majority of internet users never notice, because they are still using IE and don't care if other people can't view the sites or whether or not MS conforms to standards

Anyway, as I notice I'm rambling, the point is that since MS has such a large presence in the internet today, many times they can just do what they want and people have to follow, and the people that don't will get left on the side of the road.
I'm not saying that we should just accept what they do, but we should raise awareness of how they mess up the standards and work on getting them to follow the w3c guidelines (however that can be done...)

-Hurstdog

[ Parent ]
Re: Browsers and IE... (3.50 / 2) (#102)
by mrbob on Thu Jun 01, 2000 at 09:43:21 AM EST

The problem is, like you said, no one cares. If they can view it fine in IE, then they have no idea. It comes down to the bottom line. If a M$ product allows a person to create a web-presence more quickly that a non-M$ solution, they're gonna go with M$ every time. You can't fault people for making that choice. M$ has made it ok for millions of people to be computer illiterate, think of all the MCSEs :-) It comes down to the bottom line, and thats where companies get hurt. If you can't buy a product you can't see (because the site is IE-ified), then that company is losing money. The best thing to do is to write the webmaster/hostmaster (whomever), like the poster tried, and tell that joe or jane that you'd have liked to browse the site and buy stuff, but you _ONLY_ use netscape and can't see their products. I'm sure the genius on the other end will figure out his company is losing money if he gets enough of these emails (considering they dont bounce). And if they do bouce, write snail mail (heaven forbid) to the addy registered with InterNIC. At the very least, they couldn't say they didn't know they were not visible to everyone.

It's alright man, I'm rambling too :-)

mrbob

[ Parent ]

Re: Browsers and IE... (3.00 / 1) (#130)
by core on Thu Jun 01, 2000 at 09:28:32 PM EST

All it takes is for a braindead author to put code on the site to look at the browser string with a CGI/servlet/whatever and print up a lovely and helpful error like: "This site requires IE blah or Netscape blah to function. Click here to download IE...." and then not allow you to do anything on the site at all unless your browser string matches. I run the junkbuster proxy and modify my browser string at will so run into this a lot. It is so frustrating. It's equally as annoying when a site will test if you have javascript enabled and will programmatically not let you continue until you enable it. Requiring cookies is another pet peeve as junkbuster prevents you from getting to a lot of sites that do not fail over to URL session IDs or something other than cookies (although sessionIDs in URLs that prevent you from passing a bookmark on a site are the other extreme of annoying). -core

[ Parent ]
write the ICANN? (4.00 / 1) (#30)
by mrbob on Wed May 31, 2000 at 09:53:38 AM EST

I thought a domain registrant was required to have a valid, functioning webmaster/hostmaster account? Couldn't you write the domain registrar and raise hell with them for not being able to get in contact with webmaster@whatever.whatever? At the very least, they could/should be able to provide more contact information for the person or corp that owns the domain. Of course this doesn't get at the problem of M$ monopolizing the web. You can't expect that M$ ever play fair; heaven forbid someone follow open specs. The only reasonable thing to do, outside of writing a new browser, is to boycott those sites who chose not "support" other OSes and Browsers; for example, stay away from CtiFinancial (from Citibank/CitiGroup) as they _CHOOSE_ not to support *nix. Bob

Re: write the ICANN? (3.00 / 1) (#32)
by Alhazred on Wed May 31, 2000 at 09:59:20 AM EST

Nope...

domains can be used for a lot of things. There need to be 3 contacts provided (they can be the same person however). The email address of that contact could be anything however. I mean how does ICANN even know your going to RUN a web server? They can't possibly demand that your domain even have a mail exchanger let alone a "webmaster@foobar.whatever"
That is not dead which may eternal lie And with strange aeons death itself may die.
[ Parent ]
Re: write the ICANN? (3.00 / 1) (#44)
by mrbob on Wed May 31, 2000 at 11:53:14 AM EST

yes but the fact that you have a domain means you have a DNS server, right? (even if you use one of those i.am go.to, or whatever services, _THEY_ have a DNS server) I think I have the wrong organization; anyway, the domain registrar (whomever it may be, ICANN , InterNIC, whatever) requires that a DNS server have a contactable person in case they find a problem with your domain, right? It is a requirement of having a domain (not that you'd even put anything on the web), right? I'm not saying that they would care, but there _IS_ a way to find someone to complain to. As an example, go to register.com and query microsoft.com. They give you heaps of info. Like I said, you may not find someone who cares, but you _CAN_ get ahold of someone somehow. blah blah blah, I don't care. Let M$ assimilate the world. No one gives a sh*t anyway. We Geeks are the minority; big business will always win because they have more money.

[ Parent ]
Re: write the ICANN? (4.00 / 1) (#52)
by LetterJ on Wed May 31, 2000 at 01:18:14 PM EST

You don't have to specify a DNS server when you register. Obviously for the site to work, you do, but not just to register the name. For example, I recently registered phpgeek.com and haven't decided where to host the thing and so it has no DNS servers yet.
"If you can't explain it to an 8 year old, you don't understand it." - Albert Einstein
[ Parent ]
Re: write the ICANN? (none / 0) (#116)
by Alhazred on Thu Jun 01, 2000 at 01:19:53 PM EST

Ahhhh, grasshoppa, there is much you do not know...

Actually ALL domains have to be directed to a DNS server somewhere. In the old days that meant you had to have one of your own, but the new registration services will "park" your domain for you. If you actually use whois and look up your domain it WILL list a name server, and a zone transfer from that server with nslookup will demonstrate that it exists, though it may be an "empty" domain.

The root name servers would be swamped due to "lame delegation" problems otherwise.
That is not dead which may eternal lie And with strange aeons death itself may die.
[ Parent ]
Re: write the ICANN? (none / 0) (#114)
by Alhazred on Thu Jun 01, 2000 at 01:16:00 PM EST

Yeah, sure. All I meant was webmaster@whatever is not guaranteed to exist, nor hostmaster, postmaster, root, etc. though most or all of those are good things to have, along with spam_abuse and a few other standard ones.

Perhaps someday there will be a "network accreditation to connect to the net" that teaches net admins about needing all this stuff... hehe.
That is not dead which may eternal lie And with strange aeons death itself may die.
[ Parent ]
Citibank (3.00 / 1) (#40)
by mattc on Wed May 31, 2000 at 10:58:43 AM EST

Another reason to avoid Citibank is that they like to spam people. Somehow they got my email address and hit me with a couple of spams!! Usually you expect spammers to be some fly-by-night operation.. I was suprised to see a real company spamming.

[ Parent ]
Lets be realistic here (4.66 / 6) (#31)
by Alhazred on Wed May 31, 2000 at 09:53:58 AM EST

I'll boil your argument down first:

1. MSIE is a non-standard browser
2. It has most of the browser market share
3. MS Web development tools essentially build "IE Only" pages
4. All other browsers suck
5. Once M$ dominates the client side completely they are free to lock everyone else out of the server and tool markets.
6. We thus need a totally new OSS browser project.


I reject your conclusion catagorically.

I agree with you on points 1,2,3 and 5, but your conclusion is I believe unwarranted.

First of all lets look at market share. This is a highly ephemeral thing. We are on the brink of a revolution in client side computing at least as significant as the advent of the PC. I refer to the proliferation of hand held devices and appliances. In 3 years Microsoft's dominance of the PC desktop browser market will be virtually meaningless, and certainly not significant enough to give them the ability to make the web proprietary. Their most significant advantage in fact will probably be the fact that PC's will still be used as authoring stations and I don't see them loosing the desktop anytime soon. Still, the entire technology is about to go into massive flux once again, and market share now is not going to necessarily mean diddly squat. Let MS waste their resources on that, good for them.

Second, professional web designers, the ones that build the big sites, are never going to drop compatibility with all browsers, and they are insisting on and slowly getting total compliance with W3C standards. This will accelerate as the market diversifies on the client end, and as application service provision comes to dominate the economics of the market. Things will HAVE to run everywhere. If there is a fully W3C compliant browser, it WILL WIN. Thus MS will be forced into compliance. Their dreams to the contrary are simply short sited stupidity. Remember, MS was caught by surprise by the advent of the net. People attribute too much smarts to Mr G. He knows how to smash competition, but he is not a visionary.

Now, point 4, all other browsers suck. This is just patently not true. Your beating your breast. I use Netscape exclusively, and I do a LOT of stuff online, heck I LIVE on the net practically and I don't need IE and I don't miss not having it. I grant you that IE5.x is in many ways a superior browser. Fine. Netscape is more than just adequate (though on Linux its way crashy as you noted). As for Opera, there is no reason to suppose it can't support all the features that MS invents, but I think they are smart enough to realize that it is a waste of time to do so. Mozilla cannot fairly be compared to anything as it is a work in progress. Netscape 6 is actually quite impressive as a technology demonstration, though you are correct when you say it is not yet a useable browser. We might lament the long development time for this new offering from Netscape/Mozilla, but thats life, and the thing is HIGHLY portable, which IE is not, and incorporates technologies which have merit on their own, and are already being adopted into other projects. In much the same way Konqueror is a work in progress and shows great promise. In fact it may actually be the best of all, though its tied to KDE. On that subject Konqueror should be able to handle ANY media type that KDE in general supports via Kparts technology, so I wouldn't worry about .bmp files or whatever...

Thus we come to your conclusion that we need another OSS browser project. Why? Browsers are extremely complex software. Mozilla is being developed by a very talented group and there is a damn good reason its taken over 2 years to make an almost-beta quality product, its FREAKIN HARD! Konqueror has an advantage in that KDE itself provides them with a lot of the infrastructure to build on, but even so they aren't any further along than Mozilla.

A new OSS browser will just mean starting over and spending ANOTHER TWO YEARS building a browser. Look at the dismal track record of word processors and spread sheets. I know I'm going to get flamed for this, but all the OSS ones SUCK ROCKS IN HELL. They aren't even within an order of magnitude of the functionality of MS Office products. I don't have any reason to believe an OSS browser project will fare any better. By all means go out and prove us wrong, I'd love to see it, but I won't hold my breathe! Instead I'd urge people to support Mozilla and get it out there in all its glory. It does NOT suck!
That is not dead which may eternal lie And with strange aeons death itself may die.
Re: Lets be realistic here (4.00 / 2) (#51)
by Anonymous Hero on Wed May 31, 2000 at 01:04:42 PM EST

Browsers are extremely complex software. Mozilla is being developed by a very talented group and there is a damn good reason its taken over 2 years to make an almost-beta quality product, its FREAKIN HARD! Konqueror has an advantage in that KDE itself provides them with a lot of the infrastructure to build on, but even so they aren't any further along than Mozilla.

I don't think browsers are particularly complex ('course, I've never tried writing one, so I may be sticking my foot in my mouth for the whole world to see here!) A browser in its simplest form is nothing but a glorified file viewer, after all.

Trouble is, MS and NS have made life hard for themselves and for anyone coming after by writing their browsers so that they take any old line noise and start parsing it like it was well-formed HTML. That wasn't clever. They should have put the message out: you will conform to a DTD or we will find where you live, burn your house to the ground, and sell your loved ones for dog meat. At least, that's what I feel like doing every time I see an HTML page generated by some kewl (cough) product that litters its output with <FONT FACE> and other bollocks. Then again, I guess the Redmond and Mountain View boys were too busy implementing such must-have features as <MARQUEE> and <BLINK> to do the job properly.

I guess the only real solution was the one outlined in a previous post: a Junkbuster-style proxy that sits between your browser and the net and tries to massage the inrush of net.vomit into something approximating well-formed HTML. Maybe XHTML will have some effect when it gets going, as it specifically forbids some of the crimes page authors have been allowed to get away with for years.



[ Parent ]
Re: Lets be realistic here (3.50 / 2) (#60)
by matthead on Wed May 31, 2000 at 02:39:25 PM EST

I don't think browsers are particularly complex ('course, I've never tried writing one, so I may be sticking my foot in my mouth for the whole world to see here!) A browser in its simplest form is nothing but a glorified file viewer, after all.

Um, not quite. A browser, first of all, has to parse HTML and "render" the page... file viewers do not. Second, if [this browser] is to be any more functional than, say, lynx, it also has to parse and execute Javascript, recurse links (for retrieving images and stylesheets and suck and so forth). Then there's plugins to consider...

I agree with a lot of what's been said here: WYSIWYG editors, for the most part, do not function as advertised, and even when they do, almost never in more than one browser. Competing "standards" are hell for those who try to keep up compatibility, but another way to look at it is: hey, at least it makes more jobs for web developers.

Someday I'll try to write a web browser. Not, you should understand, in order to gain market share or whatever. More as a learning experience, to figure out for myself just how difficult it is to parse and render HTML.


--
- Matt
I'm at (0.3, -2.5). Where are you?
[ Parent ]
Re: Lets be realistic here (none / 0) (#98)
by Anonymous Hero on Thu Jun 01, 2000 at 05:23:54 AM EST

Um, not quite. A browser, first of all, has to parse HTML and "render" the page... file viewers do not. Second, if [this browser] is to be any more functional than, say, lynx, it also has to parse and execute Javascript, recurse links (for retrieving images and stylesheets and suck and so forth). Then there's plugins to consider...

Sorry, I should've made it clear that when I mentioned file viewers, I meant things like GhostScript and xdvi as well as the common-or-garden bitmap viewers. HTML isn't unique in needing to be rendered.



[ Parent ]
Re: Lets be realistic here (4.50 / 2) (#71)
by rusty on Wed May 31, 2000 at 04:17:24 PM EST

And a weblog is just a glorified database frontend. :-) It's the "glorified" bit that's really hard. I expect that writing a browser is several orders of magnitude harder than you might expect.

Secondly, about <FONT FACE> tags-- have you looked at the HTML here lately? Currently, the simple choice in web design is, it can be ugly, or it can be a kludge. <FONT FACE> and such tags are a kludge, but they pass the Pragmatism Test: they work, on all known browsers. CSS? No way. So you can have the "gray page, default text" look, or you can kludge around all the brokenness in the current crop of browsers. I don't fault anyone for either choice, but it does irritate me when people claim that there's some magic third option. There isn't.

I do agree with you that most WYSIWYG generators output total garbage. As someone else pointed out, writing html is more an art than a science, and the correct choice of tags is usually very context-dependent. I still wonder why people don't bother to learn the basics before they turn the crank on some frontpage-style HTML sausage factory. I mean, cripes, five tags and you have 80% of the functionality of HTML mastered. It can't be any harder than learning what all of the buttons on a WYSIWYG generator do!

Sigh. Four years now, and the battle rages on :-/

____
Not the real rusty
[ Parent ]

Re: Lets be realistic here (4.50 / 2) (#95)
by Anonymous Hero on Thu Jun 01, 2000 at 05:14:26 AM EST

And a weblog is just a glorified database frontend. :-) It's the "glorified" bit that's really hard. I expect that writing a browser is several orders of magnitude harder than you might expect.

Eek! Touched a nerve there, I think. rusty, no criticism, implied or otherwise, of kuro5hin was meant.

I'm not really denying the difficulty of writing a browser-- even thinking about how you'd handle table layout alone makes me queasy)-- I'm just saying that it's probably not as hard as people think. MS and NS employ lots of PR droids to convince the stoog^H^H^H^H^Husers how studly their developers are. Add to that the general aura of techno-mystique that surrounds the Web in the public mind and we're talking some serious FUD generation. Web browsers are the pinnacle of technology!!! You cannot write one!!! Your pitiful talents will crumble before the intractable problems faced by our programmers every day!!! Don't try this at home, kids!!! Give me a break.

Secondly, about <FONT FACE> tags-- have you looked at the HTML here lately?

No; you are using the "nested tables to make coloured table borders" trick, therefore your source is unclean! Get thee behind me, HTMLSatan. (-: But, seriously, I bet you made sure that your generated HTML was properly nested before rolling it out, rather than expecting the browser to Do What I Mean (tm)?

Currently, the simple choice in web design is, it can be ugly, or it can be a kludge. <FONT FACE> and such tags are a kludge, but they pass the Pragmatism Test: they work, on all known browsers.

Check out this page to see the true evil of <FONT FACE>. I've stopped bothering to set fonts on pages. It is ignored by the 99.9% of users who never bother changing the default browser font to something other than Times New Roman and pisses off the 0.1% who care about these things and set the font to something nice.

We could go on, but this is the sort of discussion that flamewars are made of and I don't think anonymously flaming a site's creator on his own site is going to win me any netiquette points any time soon. (-:



[ Parent ]
Re: Lets be realistic here (none / 0) (#101)
by Anonymous Hero on Thu Jun 01, 2000 at 09:39:19 AM EST

I'm not really denying the difficulty of writing a browser-- even thinking about how you'd handle table layout alone makes me queasy)-- I'm just saying that it's probably not as hard as people think

Table layout is not that difficult if you think hard at it (I did). I'd beleive the most difficult part of a browser is the overall architecture and the general uncertainty of the environment. I mean you are not just displaying HTML, you want to do it fast, you want to make the layout incremental, you want to support half-broken HTML, you don't want to hang if an image stop in the middle, you need multi-threaded stuff (or a rather complex state machine), you need to deal with forms (make your own widgets ? use the ones of the OS ?), you need to be standard compliant (ie: CSS or DOM stuff is not really light), you need javscript, etc, etc ad nauseaum.

The single idea of anyone rewriting this from scatch again make me smile. Mozilla is a fantastic achievement, and working on stabilizing it is much more worthwile than starting any other project. Sure mozilla won't be ready before mid-2001, but there is no other choice.

And mozilla is way superior to IE5. Mozilla is a foundation. For instance, XUL is something very promising (not for the chrome stuff, but in association to RDF. It opens nice possibilities for thin-client/fat-server javaless applications)

Stay tuned to discover people using mozilla to do something else than a browser. (ie: specialized apps that are modification on the browser concept)

My real gripe against mozilla is that it is ugly beyond belief. It have a consistant ugly look on every platform, and chrome don't change anything to that (aphrodite is just ugly in a different way). IE5/Mac is real nice, but it won't take that long to wrap gecko in platform-specific shells, to make mozilla shine.

And about this samba-like project: why dont you patch gecko to behave more like IE5 if a preference is set for a given URL ? This won't take 2 years, and won't be wasted time...

Cheers,

--fred

[ Parent ]

Re: Lets be realistic here (none / 0) (#136)
by rusty on Fri Jun 02, 2000 at 12:06:24 AM EST

I don't feel flamed at all-- my example of the weblog was just the handy one that I have personally experienced. I.e. I went into it thinking "it won't be all that hard!" and it wasn't, for a while, but then there's a point at which it suddenly gets much harder. I suspect a browser would be the same way, only worse. :-)

As for the FONT FACE stuff, ok, that page explains one good reason not to use it. Luckily, I'm not using it to try to display a language with complex characters. Just good ol' ASCII text. Secondly, the font stuff is added on display, not before. All the actual data here is stored totally plaintext (except for any markup you add to comments or stories, but that's heavily restricted). Basically, I use fonts because I know that the majority of people don't change their font from times new roman, and I like how verdana looks a lot better.

Why not use CSS? It screws things up. Most current browsers have broken CSS implementations, and while you might be able to pull off 80% of what you'd want to do, that just doesn't cut it for me. Believe me, I'll be thrilled when I can convert the whole HTML mess in Scoop into CSS, but that day isn't here yet.

And also, one final note-- users with accounts can choose their own font face and size, in their preferences here. So if you think my choice is too small, set your prefs to "3". Wanna see it in Caslon OpenFace (which you happen to have installed)? Go right ahead. So there isn't the issue of "I don't have that font", which I also try to avoid by having multiple choices, from most to least specific. At worst, everyone will have a "sans-serif" font, which is the last option.

So, yes, it's a kludge, and I do know it. I think the crucial difference is that there are some of us (like me) who push HTML beyond what it ought to be doing in order to acheive certain usability effects, in full knowledge that there ought to be a better way, and we'll change when one becomes available, and then there are people who just have no clue, and don't care. Ask any regular lynx user how much I care about the real-world usability here. :-)

____
Not the real rusty
[ Parent ]

What about this (3.00 / 1) (#33)
by Emacs on Wed May 31, 2000 at 09:59:43 AM EST

Is there an engine out there that can take MS html as input and stream html that netscape understands as output ? I'm thinking of something along the lines of a junkbuster type proxy, except it looks for some of the nasty stuff that is known to blow up netscape and turns it into harmless html.

Yeah... sounds like a kludge, I know, but it's an idea.

Check this Out (5.00 / 1) (#62)
by matthead on Wed May 31, 2000 at 03:00:07 PM EST

Check out John Walker's site, Fourmilab. He has a utility ("Demoronizer," listed under UNIX utilities) that works more like unix2dos than a filter, but I'm sure you could kludge it. He has some good miscellaneous anti-MS reading, too, if you like that sort of thing.


--
- Matt
I'm at (0.3, -2.5). Where are you?
[ Parent ]
Re: Check this Out (2.00 / 1) (#143)
by Anonymous Hero on Fri Jun 02, 2000 at 12:04:46 PM EST

That's the first (and IMHO probably the quickest) solution that I thought of when I read the original post. A filtering layer between the browser and the network connection could intercept the HTML code, strip and/or replace all noncompliant / dangerous code, and let the resulting code pass through to the browser. I'm already using a similar solution on my Win and Linux boxen to filter banner ads and dangerous Javascript code. (Proxomitron, Junkbuster, among others are available solutions for this purpose.) The sophistication of the filtering engine of Proxomitron, Junkbuster etc... probably wouldn't allow for the sort of filtering that I'm proposing here, since entire sections of HTML might have to be changed, or code changed depending on context, etc. However, a similar local proxy solution seems to be the most promising solution right now. This local proxy could even have a plugin architecture, allowing different filter "files" to be loaded depending on which browser you're using. It could have a Netscape filter, a Mozilla filter, an Opera filter, hell, a Lynx filter, that would filter the incoming code to be most compatible with the browser of your choice, and only then let it through. Anyone like the idea, mail me at wiggles@888.nu for further discussion.

[ Parent ]
Perspective from a professional... (3.80 / 8) (#36)
by octos on Wed May 31, 2000 at 10:30:44 AM EST

I've been building web sites professionally for about 3 years now. I started learning HTML back when Netscape3 was the browser of choice. I will agree that IE is not the most standards compliant browser, but it is moreso than NS4. If you want to talk about the HTML extensions game, we can talk about Netscape. They started it very early on. In many ways, it was a good thing because some of these things eventually became part of the spec. With IE on the scene, netscape really lost it with NS4. Have you heard of the LAYER tag? That is an example of empbrace and extend by NS. At the same time, IE implemented CSS fairly close to the working draft that the W3C was working on. I'm not a huge Microsoft fan, but they currently make the best browser. The nice thing about HTML is that you don't have to use the proprietary extensions and if you do, anoter browser is supposed to ignore the tags it doesn't understand. Netscape will sometimes try to render non-supported tags or crash trying. I don't want to solely beat up on Netscape, but they aren't much better than MS in the area of supporting the HTML spec. The reason a lot of pages don't render in NS is that it doesn't toloerate sloppy code while IE loves it and asks for more. This reflects coding standards more than the browser. All the people I've worked with code by hand. It's the only way. The reason WYSIWYG editors spit out crap is because they are trying to reproduce all the workarounds for browser incompatibility that a pro web builder knows. It's more art than science. As for the alternatives, Opera is a wonderful browser that supports the spec to the letter, has a footprint less than 5 meg and renders pages FAST. I don't mind paying for quality. There was a time when you paid for browsers and your own thinking has been tainted by IE since you think a browser should be free. (I'm for open source too. The story author talked about using VMware-- a commercial product, so I don't believe he is against purchasing software) Mozilla trashed the old Netscape codebase and pretty much started from scratch. I'm writing this post using the M15 build. It is stable and renders as fast as IE and aims to be 100% standards compliant. The footprint is 10 meg-- Not too bad for a beta. MS isn't stupid. I believe that they understand that the next browser war will be waged over standards compliance. Take a look at the reviews of IE5 for the MAC. It is very compliant. The heart of the problem is standards. People that code pages don't follow the spec. A small part of that is workarounds for bad implementations by the browser makers. Some of it is people using Frontpage and Word, but a lot of those are personal pages by people that don't know better. That doesn't make it right, but as the very angry developer community pushes for more standards compliance, the rest will catch up. The internet has been in a low period, but it is getting better.

Pear o' graphs (none / 0) (#66)
by kmself on Wed May 31, 2000 at 03:41:29 PM EST

Paragraph

breaks

considered

useful.

--
Karsten M. Self
SCO -- backgrounder on Caldera/SCO vs IBM
Support the EFF!!
There is no K5 cabal.
[ Parent ]

The Microsoftization of the Web (4.00 / 1) (#38)
by mattc on Wed May 31, 2000 at 10:39:45 AM EST

One area I've personally found annoying are car manufacturers' web site. Last year when I was researching a new car, I was unable to load almost ALL of the car manufacturers' sites. They all wanted "Internet Explorer with blahblah-windows-only plug-in".. or else they were nearly impossible to navigate. Since the only web sites I could get working were Toyota and Ford, I ended up buying a car from Ford.

I don't think I'm the only one with this problem either. There are a lot of Windows users who simply refuse to install the latest gee-whiz plugin (especially after all the virus problems), not to mention Mac users and Linux users who are dependant on the easily-crashed netscape browser. These businesses need to wake up and realize that if no one can see their web site then they aren't going to be selling anything.

On another topic, I think the author is overly harsh on Opera and Konquerer. As for Opera, I'd be happy to pay $35 for a web browser (assuming it works). I spend so much time on the web, it would be worth paying $60-100 if it meant I could have a decent web browser. Konquerer is another good browser. While I haven't used it for a while (I've been messing around with Gnome too much), I think it has a lot of potential, and IMO could unseat Mozilla as the leading Linux web browser.

I agree with the author that the AOL web browser.. er.. Mozilla.. is probably a lost cause. It is just as bloated as Communicator.. Have you tried running Mozilla on a 486?.. or even a Pentium?? It's slow as molasses.

Re: The Microsoftization of the Web (4.00 / 1) (#55)
by Anonymous Hero on Wed May 31, 2000 at 01:50:27 PM EST

And did you remember to tell all the other car company's why you decided to buy Ford? The fact is, a lot of people who make IE-only web pages just don't know any better. We need to *nicely* let them know that there are people interested in their product who are not using IE and/or not using Windows. Otherwise they'll never even know they're doing anything wrong.

For example, the other day I received an e-mail with some Word documents. I politely sent back a reply that I cannot read Word documents and could they please send the attachements as Text, HTML, XML, PostScript or PDF (yes, I have AbiWord and StarOffice... but I object to getting Word documents as attachements from an ISP). Admittedly, I haven't heard back from them, but the fact is they know why they lost my business. Now, one person doesn't matter, but if everyone who doesn't use Windows also refused to read Word documents, eventually they will change. But only if you let them know.

[ Parent ]
Re: The Microsoftization of the Web (2.00 / 1) (#56)
by Anonymous Hero on Wed May 31, 2000 at 01:51:16 PM EST

Ick. Company's. Sorry. Should be companies (obviously).

[ Parent ]
Re: The Microsoftization of the Web (2.00 / 2) (#87)
by sar on Thu Jun 01, 2000 at 12:02:26 AM EST

I don't think it's true that they don't know anybody. Most web sites for traditional companies get designed by comitee where marketing wants the the latest gee-whiz feature thrown in, and they dictate this to the poor slob who got stuck implementing it all. Or they are sold wholesale to the company by an outside firm who throws in full page animates gifs and wierd popup navigation system to wow the customer. The truth is that in some companies IT isn't even in charge of the web site, pr and marketing are.

I am not a web designer, but I have complained several times about the fact that our site fails miserably in netscape (it works with mozilla) but i have been told that we are ignoring what they call "the idiots still using netscape". The odd thing is that the web site is a part of our "revenue model" where we " leverage synergies" and "integrate ERP with legacy systems" or whatever buzz word marketing happens to hear that week.

There is even a press release from our marketing dept where we refer to our site as a "porthole for the industry"

[ Parent ]
Mozilla on low-end hardware (3.00 / 1) (#69)
by Demona on Wed May 31, 2000 at 03:51:55 PM EST

I have successfully run Mozilla (nightly build as of sometime last week) on a Cyrix 586/100 with 48Mb RAM. It took a goodly long time to start up, but its performance once loaded was what I would call acceptable, if not exactly fast. This in itself is an amazing improvement since as recent as one month ago. If it continues to improve at this rate, I have little worry that the finished Mozilla will run well on 486-class hardware.

[ Parent ]
I just don't know what to say. (4.33 / 3) (#41)
by eann on Wed May 31, 2000 at 10:59:11 AM EST

I have a rush of thoughts about this, and none of them really strong enough to surge to the top. It all kinda smears together.

I used to be very strongly anti-IE. But they have improved it a lot since I first ran afoul of it, and I have to acknowledge, somewhat grudgingly, that it really has surpassed Netscape in quality even if it does propagate "tag soup". I even use it occasionally (more often than Netscape, anyway). I am concerned, as others have pointed out, that you're short-changing both Mozilla and all the other browsers that exist.

There are, in general, two kinds of web browsing: looking for information and having fun. Sites that are providing information, if they're run by anyone halfway respectable (that is, worth getting information from), will make that information accessible to everyone. These are the kinds of people that participate in the Web Standards Project, and that use Opera and iCab and IE5 (for Mac), and eagerly await Mozilla. They use editors like HomeSite and BBEdit, often with a site design/management tool like Dreamweaver (which, for the most part, doesn't spit out crap for code).

Sites that exist simply for fun (even if, for instance, it's someone's personal sandbox with "information" about them) aren't held to that restriction. There's no reason they should be. It's their sandbox, they can make the rules.

Amazingly, there's even something coded into the HTML standard that allows for this dichotomy. It's called DOCTYPE (see the excellent article at O'Reilly Network). How will we get people to use it? Dunno. But your proposed Samba of Browsers (SoB?) would use it to determine which way to render the code. Much like IE, if it got the "strict" DOCTYPE, it would stick with strict interpretation, otherwise, it'll slow down into sloppy mode.

The wonderful thing about an open source project is that there's usually someone who can take a meaningful or interesting suggestion and turn it into reality. The first step towards the SoB would be to have a list of the things that IE does that aren't in the HTML 4.01 spec. Even without much knowledge of coding, if you were interested in contributing this idea to a project, you could work on that part. Some things, like translating backslashes to "real" slashes in path names before asking the server for the file, would be fairly easy to implement under the "sloppy" DOCTYPE (of course, that particular example would also be pretty easy to implement on the server side). Support for inline BMPs (or whatever other media format) should be handled with plug-ins. Without actually looking, it'd surprise me if there's not a plug-in for Netscape that does such things already.

In short, the only reason no one has done this is because no one has made a well-researched, well-reasoned request. You're most of the way there, and the issue is obviously important to you. Collect your thoughts and start something. If you don't, who will?

Opera's worth the money, BTW. I'm a fan of Open Source and/or Free Software, but I also understand that others aren't, and they have a right to whatever business model they choose so long as it doesn't unfairly/illegally impinge on competition in the Free market. And, if I like their product, for whatever reason, I may have to pay for it. It's worth US$35 (US$18 academic) for me to have a small, fast browser that I can use consistently 8 to 10 hours a day, and only crashes once every few months (usually because of someone else's sucky Javascript code).

And always write code that conforms to the HTML 4.01 spec. If you don't, you have no moral high ground to stand on.

Our scientific power has outrun our spiritual power. We have guided missiles and misguided men. —MLK

$email =~ s/0/o/; # The K5 cabal is out to get you.


The catch is... (4.00 / 2) (#50)
by Anonymous Hero on Wed May 31, 2000 at 12:48:27 PM EST

And always write code that conforms to the HTML 4.01 spec. If you don't, you have no moral high ground to stand on.

The catch there is, that given that Netscape hasn't fully kept up with HTML 4.01 and IE doesn't particularly try, it's not uncommon to have a situation where if you want a page to display correctly under IE and Netscape, you HAVE to write non-conforming code. I can think of several examples - in table formatting, for instance - where IE doesn't follow the spec, and if you want your pages to display for IE users, well, you're writing non-standard code. Style sheets would make some of this go away, if only Netscape would support them properly.

I'm writing this on my HPUX workstation at work with Netscape, so I'm terribly familiar with non-Windows, non-IE headaches. Many of the fonts that have started being specified aren't available to me on this box, so I'm getting all too used to seeing random ?'s show up in place of hyphens, quotes, letters, etc.....


Alarmo, having login problems....


[ Parent ]
Re: The catch is... (3.50 / 2) (#80)
by Anonymous Hero on Wed May 31, 2000 at 08:09:21 PM EST

Many of the fonts that have started being specified aren't available to me on this box, so I'm getting all too used to seeing random ?'s show up in place of hyphens, quotes, letters, etc.....

Quotes, hyphens, etc. should exist in every font. The ?'s don't have anything to do with the font, I'm using Windows TrueType fonts on Linux and still see them. I think it has something to do with the character sets, or possibly the "smart quotes" (or "smart hyphens", i.e. combining "--" into a bigger dash) Word likes to insert. I've heard that even IE on Windows sometimes displays ?'s, although I don't remember ever seeing them.

[ Parent ]

Re: The catch is... (4.50 / 2) (#88)
by magney on Thu Jun 01, 2000 at 12:31:51 AM EST

You are correct, sir. There are several characters in the high ASCII range that are, according to any non-Microsoft character-set standard, non-coding. Microsoft uses these for forward- and back-quotes, as well as a few other glyphs. This allows Windows to have those characters in a document without going to Unicode (which has &lquot; and &rquot; defined, but they're beyond 255).

The consequence of this, though, is that the characters don't work in any non-Windows OS.

Do I look like I speak for my employer?
[ Parent ]

MS is not alone. (3.00 / 2) (#104)
by eann on Thu Jun 01, 2000 at 10:11:12 AM EST

Oh, it's not just a Microsoft thing. Apple does it too. That's why stories from That Guy on That Other Site often have strange characters in them, too.

Our scientific power has outrun our spiritual power. We have guided missiles and misguided men. —MLK

$email =~ s/0/o/; # The K5 cabal is out to get you.


[ Parent ]
doctype-sensitivity (4.00 / 1) (#131)
by JJC on Thu Jun 01, 2000 at 11:15:47 PM EST

Amazingly enough I read on the Web Standards Project (this page in particular) that Mac IE5 already does doctype-sensitive rendering which is definately the way to go if you ask me. The problem is which standard should you go for when using the "sloppy" renderer. Should Mozilla have an IE5-emulation mode which it defaults to when not given any doctype? Or, how about even going down to being generator META tag sensitive, ie render Word and Frontpage stuff as IE5 would but do everything else properly. Now that both rendering engines are embeddable objects (in Windows) I reckon it would be reasonably simple to hack together a front end which does some doctype and/or generator checking and then chooses the "best" renderer. That doesn't help us GNU/Linux, *BSD, whatever users but could be useful, or at least interesting as an experiment. I suppose you could do a similar thing with Mozilla, which would start up VMWare to render with IE5 (I don't suppose you can run IE with WINE, I haven't checked) but that's inpractical at best.

[ Parent ]
But what sites? (4.00 / 1) (#45)
by Paul Dunne on Wed May 31, 2000 at 12:26:25 PM EST

Can anyone think of a worthwhile site that can only be displayed in IE? For me, if a site doesn't work properly in lynx, that's a point against it, and I probably won't bother with it further, beyond perhaps a "sarcarstic" letter to the webmaster (which'll maybe bounce). If I want badly enough to use the site (case in point: Amazon Associates reports, which need https, which lynx doesn't support without a patch I'm too lazy to apply), then I'll fire up Netscape 3.01 (one of the more stable Linux versions). If, that doesn't work with it, then too bad, the site will have to do without me. But this has never happened to me. Where are all these IE-only sites?

On the Mozilla question, OK, I give in, finally I am going to download that grotesquely-huge tarball and try it out.

Oh, a minor point: don't point at vim when you mean vi. This is like thinking Red Hat is Linux.
http://dunne.home.dhs.org/

Re: But what sites? (3.00 / 1) (#54)
by analog on Wed May 31, 2000 at 01:44:48 PM EST

Can anyone think of a worthwhile site that can only be displayed in IE?

I think the word worthwhile is the key here. IMNSHO, if a site is browser specific it isn't worthwhile by definition. I've seen a few sites that probably don't render as intended in Netscape under Linux (they seem to be fine in the Windows version), but I've only been to one site that flat wouldn't render in Netscape; it was a Disney site that my son wanted to see, and as an error page loaded that stated "This site requires Internet Explorer and Windows 95", I doubt the problem was with Netscape. Screw 'em; my son didn't see their site, and he now has a much lower opinion of Disney. Their loss.

I don't like dealing with some of Netscape's problems, but I don't seem to have the troubles other people do creating cross browser code. It can be a pain sometimes, but I have more trouble dealing with the limitations of HTML and the web in general than I do with any specific browser.

My big gripe right now is the number of sites I see that are completely broken if Javascript is turned off; I've come across two or three in the last week that just display a blank page, with no indication of what's wrong. What's especially irritating is that frequently I can (and have) duplicated said sites with static html. What the hell they're using scripting for is beyond me. I actually sent a 'fixed' site to the webmaster of one; I got an email back saying "thanks for the input, we'll try to be better in the future". Yay.

[ Parent ]

e-commerce (none / 0) (#64)
by kmself on Wed May 31, 2000 at 03:24:20 PM EST

A number of sites I hit, Charles Schwab comes to mind, are distressingly browser/feature specific. Not restricted to IE, but definitely not intended for broad compliance.

--
Karsten M. Self
SCO -- backgrounder on Caldera/SCO vs IBM
Support the EFF!!
There is no K5 cabal.
[ Parent ]

Re: e-commerce (3.00 / 1) (#68)
by Demona on Wed May 31, 2000 at 03:48:38 PM EST

Heh. Ever since I compiled Lynx with SSL support, I find there are very few websites that I can't do "e-commerce" with.

[ Parent ]
Javascript... (none / 0) (#94)
by kmself on Thu Jun 01, 2000 at 05:13:40 AM EST

...is the specific issue I had in mind. My browser supports it, but I hate it. I'm working on a rant which may end up here at some point...

--
Karsten M. Self
SCO -- backgrounder on Caldera/SCO vs IBM
Support the EFF!!
There is no K5 cabal.
[ Parent ]

Re: Javascript... (4.00 / 1) (#107)
by analog on Thu Jun 01, 2000 at 11:02:53 AM EST

I'm working on a rant which may end up here at some point...

Okay, just my li'l nudge; DO IT!!!

I used to think Javascript was a pretty cool idea (ah, years ago, when the web was young...). Nowadays I tend to think it's evil incarnate. Be nice to have something I can wave under people's noses; "see, I'm not the only one who hates it..."

[ Parent ]

Re: Javascript... (3.00 / 1) (#110)
by rusty on Thu Jun 01, 2000 at 11:30:49 AM EST

Javascript is still a pretty neat idea, that has suffered from an unrecoverable number of bad implementations. Ever done any serious work with netscape's DOM? It's insanely stupid. And javascript, as a language, just sucks.

When will a browser let me run client-side perl (in a safety sandbox, of course!)? :-)

____
Not the real rusty
[ Parent ]

Re: Javascript... (4.00 / 1) (#112)
by analog on Thu Jun 01, 2000 at 12:43:20 PM EST

When will a browser let me run client-side perl (in a safety sandbox, of course!)? :-)

I used to have a plugin installed in Netscape that allowed this (can't remember if it was sandboxed, though; don't think so. I remember being nervous about it). I don't even remember why I got it. It would allow for some really neat possibilities, though, no doubt about that (might even get me to learn Perl ;).

I'm not sure it matters though. I have a feeling that a major shift is coming, and the supremacy of the standalone, monolithic browser may be ending. Look at some of the stuff the Mozilla project is doing as far as making it a platform, and componentizing everything. A lot of people think it's all a waste of time, and will amount to nothing. While I can't deny that possibility, I'm really beginning to think the reason that they look so clueless to some people is that they're so far out in front that noone can get a handle on it.

You may not get Perl on the client side, but I'd be willing to bet you can generate whatever you do get with it. ;)

[ Parent ]

Re: Javascript... (none / 0) (#145)
by rusty on Fri Jun 02, 2000 at 02:08:25 PM EST

Yeah, I think Mozilla id going to really change things. All the XUL interface stuff is just way too freaking cool. I can't wait until I can write my stuff as a mozilla-based network app instead of a web page... HTML as an application toolkit very much resembles what JWZ once said about Xlib:

"Trying to make an application out of this is like trying to build a bookshelf out of mashed potatoes."

I eagerly await the day I will no longer be constrained to the "mashed potatoes" of HTML as an application interface!

____
Not the real rusty
[ Parent ]

Server side v. client side (ob: PHP plug) (none / 0) (#118)
by kmself on Thu Jun 01, 2000 at 02:45:12 PM EST

The problem with any client-side processing of web pages is that you're dependent on the client side to do your processing for you, correctly. Show me any two versions of the same client, let alone competing products, that perform to spec.

The cool thing about server-side tools, including CGI, Perl, even (ugh) ASP, and particularly PHP, is that you (the webmaster) control the interpretation. Regardless of presentation, you can at least guarantee your program will run properly (or know when it doesn't).

--
Karsten M. Self
SCO -- backgrounder on Caldera/SCO vs IBM
Support the EFF!!
There is no K5 cabal.
[ Parent ]

Re: Server side v. client side (ob: PHP plug) (3.00 / 1) (#120)
by rusty on Thu Jun 01, 2000 at 03:20:30 PM EST

I don't really want client-side processing per se, I want tools that will let me make the interface more reactive on the client side. I.e. I want to build *applications*, not web pages, but which are run off a server. Perl would be kick-ass because there's no implementation issues. It runs on every known platform (pretty much), and if a plugin implemented a sandbox, you could pare it down to a reasonable subset of cross-platform functionality.

____
Not the real rusty
[ Parent ]
Perl palm port? (none / 0) (#134)
by kmself on Thu Jun 01, 2000 at 11:51:57 PM EST

Point being -- all applications? Sufficiently broad coverage, maybe. Universal -- you'll have to do the fiddling at the client side.

--
Karsten M. Self
SCO -- backgrounder on Caldera/SCO vs IBM
Support the EFF!!
There is no K5 cabal.
[ Parent ]

Re: Perl palm port? (none / 0) (#144)
by rusty on Fri Jun 02, 2000 at 02:04:57 PM EST

What do you mean, "all applications"? I don't think all applications should be run off a network, no, but I do think many would benefit from being ASP-run instead of local. I also think that if it were easier to write these kinds of things, we'd discover whole new categories of app that we never thought of before they became possible. :-)

____
Not the real rusty
[ Parent ]
Clientside Perl (none / 0) (#154)
by Dacta on Sat Jun 03, 2000 at 09:58:23 AM EST

You can run it now (on IE4+). It's called PerlScript, and it uses MS's scripting host in Internet Exploror. This is the same technology that lets you use Perl to do Active Server Page development.

You can also use Python, and (I believe) REXX.

I've heard some informed rumors that Visual Basic 7 (the next version) is going to be based on an expanded concept of this, which will allow you to code in the VB IDE in any language (which has Active Host support), and will still allow full compliation.

I guess similar capabilities will be supported in all languages (VB7 is supposed tio be fully OO), so I suspect you'll be able to do inheritance & interfaces in all languages that support them, but you won't be able to do multiple inheritance, for instance. (OTOH, the VB compiler is the same backend as the Visual C/C++ compiler, which does allow MI. Hmmmm.)

Coming back on topic... here's the FAQ on PerlScript from ActiveState. You can use it in IIS 3+, Peer Web Server (That's that crappy personal webserver thing MS has), IE4+ and Windows Scripting Host (WSH).

WSH for those of you who aren't Winidiots (as Caspian so polietly describes windows users) is a cool but very scary thing that that lets us write interperted scripts to automate COM/ActiveX components on your (and other) systems. It's kind of like COM based shell scripting.

Gee.... I love Linux, but some of this Windows stuff is pretty cool, too. If you use Linux exclusivly, give it a try. Some of the ideas are not too bad.

[ Parent ]

Two possible outcomes of the MSization of the web. (3.30 / 3) (#49)
by Greener on Wed May 31, 2000 at 12:45:02 PM EST

Here are two possible outcomes driven by one of the largest groups of users of the web, AOL newbies.

These are assuming AOL 6 will use Mozilla or at least the Gecko rendering engine. AOL 5 is based on IE 5

  1. AOL 6 is based on Mozilla. Millions of AOL users complain to web masters that they can't view IE specific pages. Web masters begin coding to the HTML spec. (good)
  2. AOL 6 is based on Mozilla. Millions of AOL users complain to AOL that they can't view IE specific web pages. AOL reverts to IE based browser. AOL cancels Mozilla project (they're not making any money on it anyway). WWW becomes MSWWW. (bad)

These are two outcomes I can see. I'm sure there are many others.

My Ideal Browser(tm) would be a Gecko engine with forward and back keyboard shortcuts, a bookmark list and a location bar. No mail, no chat, no extra features, no bloat. Small, lean and fast.

Re: Two possible outcomes of the MSization of the (3.00 / 1) (#59)
by Anonymous Hero on Wed May 31, 2000 at 02:35:29 PM EST

My Ideal Browser(tm) would be a Gecko engine with forward and back keyboard shortcuts, a bookmark list and a location bar. No mail, no chat, no extra features, no bloat. Small, lean and fast.
Is there anything at all like this under development anywhere? I've been constantly amazed by my complete inability to find it, if there is...the Gecko engine just has too much potential here for it to go ignored.

[ Parent ]
SIMPLEBROWSER: Yes, it exists! (5.00 / 2) (#67)
by Demona on Wed May 31, 2000 at 03:47:20 PM EST

I'd been hearing people talk about this 'simplebrowser' executable that was as stripped-down a Gecko-based browser as an end user could obtain, but for the life of me I just couldn't track down a copy. Finally thanks to someone on IRC, we have the knowledge. So here are step by step instructions on how to get it running under Linux:
  1. Download the latest Mozilla nightly build.
  2. Unpack it.
  3. Put the Mozilla directory in your LD_LIBRARY_PATH environment variable.
  4. Go into the Mozilla directory and enter the command

    ./TestGtkEmbed

  5. Prepare to be amazed.
I could really use some help figuring out how much of Mozilla's distribution can be deleted and still have this thing function. 'ldd' tells me which libs it needs, but when I try just keeping the executable and those libs it gives me a nasty GTK crash. And 'strace' is a bit of a bitch to use for a non-coder like me.

With enough attention and effort focused on it, TestGtkEmbed (which I still think of as SimpleBrowser) can truly be a graphical Lynx.

[ Parent ]

Re: SIMPLEBROWSER: Yes, it exists! (4.00 / 1) (#73)
by rusty on Wed May 31, 2000 at 04:26:42 PM EST

Hey wow! That *is* cool. :-)

____
Not the real rusty
[ Parent ]
Re: SIMPLEBROWSER: For Windows (3.70 / 3) (#77)
by Greener on Wed May 31, 2000 at 07:14:56 PM EST

It seems this exists in windows builds as well.
$path_to\mozilla\bin\viewer.exe.

[ Parent ]
Re: SIMPLEBROWSER: Yes, it exists! (4.00 / 1) (#85)
by ramses0 on Wed May 31, 2000 at 10:37:46 PM EST

I was wondering where simplebrowser went for linux ... I'll have to run that when I get back to my computer. In the meantime, it's actually really neat to play with mozilla chrome, "Minimalist" at the bottom of the next link is closest to the older "EvilBrowser" chrome. EvilBrowser rocked, it didn't even have forward and backward buttons visible, but it had keyboard shortcuts, and you could right-click and bring up the "back/foward/last-five sites" menu.

http://www.mozillazine.org/chrome/gallery.html

or maybe...http://www.mozillazine.org/chromezone/, but it looks like some chrome is in the process of breaking while they get their new build stuff together.

--Robert
[ rate all comments , for great justice | sell.com ]
[ Parent ]

Re: Two possible outcomes of the MSization of the (3.50 / 2) (#75)
by Greener on Wed May 31, 2000 at 05:22:00 PM EST

Through the magic of VB and OLE I made my own lite browser in about 20 minutes. It hooks into an IE 5 DLL to do all the parsing and rendering in an object frame and a little bit of VB can add a working location bar. I haven't had the time or inclination to add forward and back capabilities or bookmarks but it's a nice little browser to use when I have to.

Wait... I just found how to do the same with the npmozctl.dll here. Instructions seem to be for Windows only in VB, Delphi and C++ but if I can get it going tonight I'll sure be happy. Let me know if you want a copy of the .exe if don't have VB.

[ Parent ]

Re: AOL/Mozilla (none / 0) (#146)
by Anonymous Hero on Fri Jun 02, 2000 at 04:08:22 PM EST

Two issues with scenario number two. First, AOL!=Mozilla. Yes, the lion's share of Mozilla development is done by Netscape employees, and AOL could tell the lot of them to stop wasting their time (or company time, anyway) on that fool open-source project. However, AOL can't kill the project outright, since it's <i>not theirs</i>.

Secondly, AOL has a vested interest in Mozilla's success. Without Mozilla, AOL will actually have to actively develop Netscape 6 if they want it.

[ Parent ]
Re: AOL/Mozilla (none / 0) (#156)
by Greener on Sat Jun 03, 2000 at 10:27:23 PM EST

I realise that Mozilla does not belong to AOL but if netscape were to end their involvement in the project it would eally set the project back. Possibly up to a year even at this late stage in the path to version 1.0. This delay would most likely give Mozilla a lot of bad press as far as John Q Public is concerned and allow IE to gain even more market share.

[ Parent ]
I'm a little confused. (1.00 / 1) (#72)
by narrowhouse on Wed May 31, 2000 at 04:20:24 PM EST

First I want to say that I understand the thrust of your article I just disagree on several keys points. 1) I was under the impression that Mozilla was mostly new code and not based on the earlier versions of Netscape. 2) I think that the "not windows" communitiy (Linux, BSD, Mac, BeOS, *ix, etc.) needs to be careful about playing catch up with Microsoft. Letting them set the standard and then following is a bad precedent, we need to lead with standards. Mozilla is more standards compliant in many ways than IE, many things that show correctly on it are broken in IE, that is a problem with IE not Mozilla. SAMBA is a special case its purpose is to work with Windows it HAS to follow Microsoft's lead. 3) The fact that some people will write broken pages because thay only know/use Microsoft products is not a new thing. It is up to the designer to check other browsers, not up to the browsers to include every trick and break in every major browser. Netscape has 38% of the market last time I checked, if you write a page that 38% of the browsers out there can't parse you should rethink your strategy or resign yourself to angering a large portion of your potential viewers.

Re: I'm a little confused. (none / 0) (#86)
by Anonymous Hero on Wed May 31, 2000 at 11:42:21 PM EST

A few quick comments...
  1. Mozilla isn't just mostly new code, it's 100% new code.
  2. I agree that letting Microsoft set standards is a very bad thing(tm); and Mozilla's goal is to be 100% standards compliant (that is, the standards set by the W3).
  3. Sadly, I think Netscape's market share is down to 25%. It's still ridiculous to imagine a business turning away one customer out of every four, but I thought I'd make the clarification.


[ Parent ]
*Sigh* (4.00 / 2) (#74)
by DemiGodez on Wed May 31, 2000 at 04:51:43 PM EST

I feel sorry for Caspian. This is a really nice essay on a serious problem with the web. And a lot of you seem to feel that if a site is IE only then it isn't *worth* viewing. Or that the percentage of sites that you can only view in IE is so small that it doesn't matter.

Personally, I use Windows and I like IE. But if I did run Linux or Unix and use Netscape (or another browser) I would be concerned. Why you ask?

The other day, I installed NT 4.0 from a CD. It is an old CD and the only browser it installed by default was IE 3.0. I thought to myself, well I should be able to use IE 3.0 to download Netscape or a new version of IE. NOPE. Not a single page on microsoft.com was accessible using IE3. Some netscape.com pages were, but none that actually let me download a browser. I ended up using ftp.

My point: Web designers used to make sure sites would work with IE 3. But now a days so few people use IE 3 (after all IE 5 and Netscape are free) that the cost of the extra work doesn't justify the minute increase in accessibility.

If IE continues to take over the internet like it has to date, before too long web designers will do the same cross/benefit analysis and decide it is just not worth it to support Netscape. (Obviously not *all* sites, but I'm speaking in generalities).

I agree with the essay that this is a legitimate concern. I teach classes in HTML and JavaScript at a local community college and although I stress the importance of testing in both browsers (and more if possible) almost all my students (who are mostly taking classes on behalf of their companies) are being told by their companies that they only need to support IE. This is where things are going.

Re: *Sigh* (none / 0) (#109)
by rusty on Thu Jun 01, 2000 at 11:26:37 AM EST

One of the big reasons no one designs for old IE's is that it's *hard* not to upgrade IE when MS wants you to, if you run windows regularly. They bundle the upgrades with practically every piece of software they release, and it's all so linked in there (err, "integrated") that the number of people using old versions of IE is tiny.

And the second reason is that, as you point out, they tend to change a lot between releases, to the point where their own homepage won't display on an old IE. I design for old versions of netscape, and the current IE (as well as lynx and mozilla), but I don't know anyone who bothers with old IE. It's not worth the time, and also, if you can get stuff working on older netscape, then older IE will probably work too.

Oh yeah, the other problem is that it's pretty much impossible to install multiple versions of IE on one machine. So how many boxen are we going to devote to testing with IE? When I realized that MS wouldn't let me set up a test machine with all relevant versions of IE on it, I pretty much gave up testing for older versions.

____
Not the real rusty
[ Parent ]

Not everybody is an code-cracker. (3.00 / 1) (#76)
by noek on Wed May 31, 2000 at 06:52:53 PM EST

I am writing this as an author of a site and as an editor for The Open Directory Project.

I can understand many people in the OSS community will dislike MS and IE. It is true IE finishes the page for the reader so the reader will not see any errors. I would like to see that more browsers supported this feature. Why?

1) Because most people who are out there and who have something to say aren't code-crackers. They simply want to share their knowledge. They don't want to work through the W3C specification of CSS, they want to tell the world about their collection of antique shares or killerbees.

2) When I use NS I find I can't read pages or pages get killed by a simple javascript. I'll admit MS makes us used to unlogic things this way, but it is fairly idiot proof. For my volunteer work i'll have to use it.

Everyone, meet Caspian! (2.50 / 2) (#84)
by mattm on Wed May 31, 2000 at 10:09:57 PM EST

Caspian is the sysadmin of twu.net, if anyone's wondering. But that's beside the point. Let me try to respond to this long, long rant as succinctly as possible:

We NEED-- absolutely NEED-- a new browser project... one that is going to do it right. One that is going to, as Samba has, ceaselessly keep up with MS's constant compatibility-busting behavior.

Go right ahead, then.

(I would start coding it myself, but I'm a complete neophyte when it comes to parsers and graphical coding.)

Then go forth and learn.



Re: Everyone, meet Caspian! (2.00 / 1) (#92)
by Caspian on Thu Jun 01, 2000 at 02:04:40 AM EST

Oh, so if I can't do it myself, I shouldn't even bother asking for it? What the bloody hell sort of an attitude is that? I suppose that this means that starving people shouldn't ask for food-- instead, they should grow their own... Or, perhaps, that people in cities with lousy airports shouldn't ask the city to fix the airports up-- they should build their own? Don't you see how arrogant and unproductive your point here is?
Hacking code for the people!
[ Parent ]
Re: Everyone, meet Caspian! (2.00 / 1) (#108)
by rusty on Thu Jun 01, 2000 at 11:21:00 AM EST

Hey Caspian-- thanks for the article... but no thanks for your comments so far. Dude, you seriously need to dial down the flame-o-meter here. His point was merely that learning to code will be a productive activity for you. I'm always glad when people point out problems, because there are always developers in need of a project. But how much more good you could do if you could help hack on such a project yourself? I don't think the suggestion was nearly as offensive as you took it to be.

And for your metaphors: Yes, starving people should learn to grow or find food. That seems reasonable. No, I don't think that people should build their own airports, but that case is hardly applicable here. Point being, some things you can only complain about, and you should. Some things you can complain about *and* help fix, and if you care, you should do both of those as well. I'd recommend though that you get involved with Mozilla. Do standards-compliance testing, do documentation. Do anything! And complaining counts as the first step in doing something. :-)

____
Not the real rusty
[ Parent ]

Re: Everyone, meet Caspian! (none / 0) (#113)
by Anonymous Hero on Thu Jun 01, 2000 at 12:53:35 PM EST

Oh, so if I can't do it myself, I shouldn't even bother asking for it? What the bloody hell sort of an attitude is that?

Well, not completely unreasonable. You shouldn't ask for it if you're not willing to do SOMETHING to help it get done. The first step you've already taken -pointing out a need- there are many places you can continue to help, and based on your grasp of English syntax I'd say you're capable of learning to write real code (HTML isn't real code - OK that was flamebait). But even if you can't learn coding for some reason you can still help by volunteering to beta-test and/or write documentation. Good tech writers are very scarce.

I suppose that this means that starving people shouldn't ask for food-- instead, they should grow their own... Or, perhaps, that people in cities with lousy airports shouldn't ask the city to fix the airports up-- they should build their own? Don't you see how arrogant and unproductive your point here is?

So are your examples. Although starving people will ask for food (unless they are from a culture that rejects charity - I have relatives who'd starve first) that does not absolve them from their responsibility to try to feed themselves and their families. Do you know the kid's fable about the little red hen? As for airports, I hate them. Here in the US, they are horrible sumps of pollution as well as a means that the federal goverment can regulate and restrict free trade and the free movement of a supposedly free people. My tax dollars pay for the airports - your rhetorical city does NOT generate income outside of taxation, and I believe the federal government funds the majority of airport projects - and I DO NOT WANT public airports - I barely acknowledge the right of freely associated individuals to build private airstrips, since they degrade the quality of life of their immediate neighbors. Your comment implies the right of wealthy non-productive drones (the people who care whether the airport is "lousy" or not - businesses just want to get where they need to go fast, so bullet trains are fine for them) to spend MY MONEY on things I DON'T WANT. That may be legit in a socialist or democratic republic (such as the USA) but the world of free software development is at best a meritocracy or anarchy.

Rant mode off. If you don't want to learn to code (I still think you'd be good at it) set up a web site for your proposed "samba of browsers". Or join the Mozilla group, write some doc for them, and then you'll be able to steer them towards a more samba-like philosophy since you will be a valued contributor towards their project.

--Charlie

[ Parent ]
Did I really read all this? (none / 0) (#129)
by noek on Thu Jun 01, 2000 at 09:22:00 PM EST

Did I really read all this? (Just for all those who also read it.)

[ Parent ]
Some statistics on Browser marketshare (3.00 / 1) (#90)
by Anonymous Hero on Thu Jun 01, 2000 at 12:50:29 AM EST

Firstly, I'm a bit of an IE fan. I always used to use Netscape - all through the IE4 series, and then IE5 came out. If you use Windows and haven't tried it - give it a go. It is SO fast, and I can open 7 windows of www.yahoo.com on it before it reaches the same memory usage as Netscape 4.6 with one. As a programmer, I can't argue with that.

It also is a lot closer to proper standards compliance than Netscape, if you bothered to check.

Some people here have been claiming Netscape had a 30%+ market share. I heard that claimed back in early 1999, so I was pretty sure that it had dropped by now. Here's my research (It's pretty hard to find good numbers on this, BTW):

From Positive Support Reviews for May 2000, IE3, 4 and 5 have a combined market share of nearly 78%. (Over 55% of the total market uses IE5).

A quick search using google found a few public web server logs. For example Here is the log for the UK Training Standards Council for July 1999. It shows IE usage at about 80%.

Another log, December 2000: IE share looks to be over 80%.

Here's one from the acofi.edu (whatever that is). It shows IE usage to be around 60% for May 2000, but counts AOL usage at another 10% or so.

There are countless numbers of these logs around (see for your self: Search for browser market share summary express analysis on Google)

They all show IE share increasing, but more importantly they show Netscape's share decreasing.

I guess this has wandered off topic a bit, but I think that the idea that we need a "Samba of Browsers" is a bit wrong. Web Authors are well aware of the need for their pages to be visible in all browsers, but more and more of them don't see Netscape as a market they need to chase. As soon as another brwoser comes along that gains some market share (and I do beleive that is still possible), they will start checking it in that, too.

Re: Some statistics on Browser marketshare (1.50 / 2) (#91)
by Caspian on Thu Jun 01, 2000 at 02:01:01 AM EST

What in the HELL is your point? That we should all switch to IE? Did you even READ my article? I RUN DEBIAN GNU/LINUX. IE IS NOT EVEN AVAILABLE FOR MY OS. More importantly, I would NOT run it if it WAS available! I do not appreciate the tactics that Microsoft is using in pushing IE, and I do not appreciate what it wishes to do with IE-- namely, through capturing the client end, eventually capture the server end as well! Sorry, but I'm NOT going to fold!
Hacking code for the people!
[ Parent ]
His point is very clear (4.00 / 1) (#93)
by PrettyBoyTim on Thu Jun 01, 2000 at 03:42:43 AM EST

His point is very clear. He's simply pointing out that IE has a vast proportion of the market share, and has gone to the trouble of digging up a few statistics to back his point.

And Caspian, get over yourself. If you post something to k5, don't get all pissy with anyone who doesn't take the discussion in a direction you don't like.

And now for my gripe on the article... ;)
Although Caspian brings up some very good points about the non-IE web discrimination that goes on out there, I think the article would be much better without the constant attacks on those who don't use the same OS as he does. To complain about predujice on one hand and then come up with comments like "You would expect that sort of brash arrogance from a WinIdiot" and "By and large, people who use Windoze are either too computer-illiterate to ever be likely to run anything else, or too busy raking in the dough to want to stop running Windoze" seems rather doublefaced.

[ Parent ]
Thankyou (none / 0) (#96)
by Anonymous Hero on Thu Jun 01, 2000 at 05:15:35 AM EST

I did the post your talking about.

I thought the same about the article. I must admit, mu post was a bit pointless, though.

The points I were trying to make were:

  • I use IE on windows because it is better than anything else for Windows or Linux (and yes, I use Mozilla M15 on both Windows and Linux)
  • IE has such a huge market share that it is reasonable for webmaster to make sure their pages work best on IE. No, I don't like this.
  • One point I didn't make were that most of the enhancements you were talking about (thing like allowing slashes that go the wrong way) are good, because they allow display of pages that are incorrect. This is a good thing, because it allows the user-agent to compensate for mistakes the designer shoudln't make, but also for things like incomplete downloads. Ever tried to fix a table-based page that half downloaded from a dial-up server in bulgaria so it would render in Netscape and you could see what the point of the page is? Fixing problems like that is a good, intellegent thing to do.

    [ Parent ]
  • Re: Thankyou (2.50 / 2) (#100)
    by Caspian on Thu Jun 01, 2000 at 09:37:06 AM EST

    Stupid, stupid, stupid. How the f*** are these "features" to benefit me if I CANNOT EVEN RUN IE?

    Again I ask-- is your suggestion that I START RUNNING WINDOZE MERELY SO I CAN VIEW CERTAIN WEB PAGES? Do you realize how ABSURD that suggestion is?
    Hacking code for the people!
    [ Parent ]
    Re: Thankyou (3.00 / 2) (#103)
    by Anonymous 242 on Thu Jun 01, 2000 at 10:06:02 AM EST

    Caspian stated: Stupid, stupid, stupid. How the f*** are these "features" to benefit me if I CANNOT EVEN RUN IE?

    The point of the previous posters was that some of these 'proprietary' enhancements are good enhancements. Given a broken web page, it is good and reasonable for a browser to attempt to render it readably. As the previous poster pointed out, there are reasons for being able to do this other than author error. Corrupted packets, broken connections in the middle of page and more can make unreadable pages.

    Caspian also stated: Again I ask-- is your suggestion that I START RUNNING WINDOZE MERELY SO I CAN VIEW CERTAIN WEB PAGES? Do you realize how ABSURD that suggestion is?

    The suggestion was not that you (or I or anyone) start using Windows. The suggest was merely that it is not unreasonable for a web designer to test only one platform. I would rather everyone test on multiple platoforms, but I'd be willing to wager that the cost of testing and redesign to support all the platfroms out there is not worth the time or money for most (not all) designers. If one chooses to use a non-mainstream platform, one must be willing to deal with the consequences.

    BTW, anyone know if the Mememonic browser project is still around? A few years ago they grabbed a tremendous amount of publicity through some well placed press releases and interviews. Then Netscape went Open Source and it seemed like Mememonic just sort of withered.

    Just my two cents.....

    [ Parent ]

    Re: Thankyou (3.00 / 1) (#117)
    by PrettyBoyTim on Thu Jun 01, 2000 at 02:09:23 PM EST

    The previous poster was not suggesting that you would gain benefit from features in a program that you cannot run. What he was saying is that many of those features are *useful*. Simply because you cannot use them does not prevent them from being useful to others.

    Nobody is trying to suggest that you start running an OS which you obviously hate - none of these posts have suggested that at all. There's really no need to go around calling people stupid simply because they have a differing point of view.

    [ Parent ]
    No (5.00 / 1) (#138)
    by Anonymous Hero on Fri Jun 02, 2000 at 05:39:05 AM EST

    I'm suggesting you get over yourself, and write/get a browser that does the same stuff on Linux.

    Mozilla does some of this stuff already. It renders incomplete/incorrect tables, and it deals with slashes going the wrong way.

    I'm pretty sure that most versions of Netscape do the slash thing, too.

    Someone (kmself?) suggested that if the web page is incorrect, an error message should be displayed, and perhaps this is true. I still stand by my claim that to attempt to render incorrect HTML into something readable and at least visible is the correct thing to do.

    I don't care what browser does it, but just because IE seems to do it better than any browser (apart from Mozilla, maybe - but M15 still has other problems) YOU SHOULDN'T BLAME MS. (Also, maybe you shoudl try Mozilla before you rant like this?)

    MS Word producing bad HTML is another story, though - that is pretty bad, and there is no excuse for it.

    [ Parent ]

    Ignoring bad HTML is good? (Re: Thankyou) (none / 0) (#125)
    by Anonymous Hero on Thu Jun 01, 2000 at 07:02:57 PM EST


    Is "fixing errors" by accepting bad HTML really such a good thing?

    Think in terms of good programming practices. Should my functions be encrusted with extra code so that your improper calls to them still return the results you expect?

    Or should you just learn to call them according to their definitions?

    People should not come to depend on *bugs* as *features*.


    What good is a "standard" if people don't actually feel themselves held to it?



    This road best driven with Mazda(tm) vehicles. Please go get a Mazda and come back.



    [ Parent ]

    Dealing with bad HTML (none / 0) (#133)
    by kmself on Thu Jun 01, 2000 at 11:50:30 PM EST

    IMO, bad HTML should result in:

    • An error message on the page rendered, with a brief description of the problem. Not an error log, just a message, though making this configurable would be cool.
    • The rest of the page content, rendered minimally, but usably.

    With most programs, you want an error to appear as an error. With failure mode though is worth considering -- do you fall over hard, or try to handle the remaining portions of the problem? How would you want a data validation routine to handle unanticipated input? How would you want an air traffic control system to handle an out-of-resources conflict?

    There isn't a single answer. For HTML, given you probably can't meaningfully change remote data inputs, gracefully handling boundary conditions is a Good Thing ™, though letting you know they're there (so you can rant to the webmaster) also helps.

    --
    Karsten M. Self
    SCO -- backgrounder on Caldera/SCO vs IBM
    Support the EFF!!
    There is no K5 cabal.
    [ Parent ]

    developer arrogance (4.00 / 2) (#97)
    by hiip on Thu Jun 01, 2000 at 05:23:03 AM EST

    Excellent article. Many web developers have completely unbearable attitude towards less mainstream browsers, or any brosers they don't happen to be using themselves. I think it's strictly developer's responsibility to see to that the page renders correctly on all browsers and platforms they can get their hands on. I never get used to seeing these "optimized for ie4, 16 bit colors and resolution 800x600". It's just ridiculous. Face it, who would change resolution just so that someones crappy homepage would look better. It's just a lousy excuse for having an ugly page. If you aren't prepared to test on all browsers and platforms, write standard html. That has always worked for me.

    I was recently thinking about applying a job at a local web developing firm so I surfed to their website to look if there were any job opportunities. However, I just got an annoying page telling me I needed to download and install Flash 4 plugin. No link to some non flash page, nothing. Now, that would be bad enough if I didn't have flash plugin, but I do, and it works fine. I don't know what sort of browser/plugin detection system they have but it obviously didn't work for my Netscape 4.72/Linux. I never saw the page and I never applied for the job. Well, it's their loss. Weird enough, I've never had any problem with the websites that contain actual information. I guess they don't have to resort to fancy tricks to get visitors to their site, so they keep it simple.

    I disagree with the conclusion, though. It's just plain unfair that the open source community should try to keep up with Microsofts unreasonable extensions (backslash in IMG tags?! is it really true? And anyone who allows clueless users to put BMP images on the web should die a horrible death. They're HUGE.)

    I also try to be more optimistic when it comes to Mozilla, I think it renders pretty fast already, though it takes forever to start up. I presume (hope) it'll get faster before beta. And from all I've heard it should conform to standards pretty well, far better than Netscape 4. If all browsers would be strict with HTML and agree on Javascript there wouldn't be any problems as the problem is sloppy coding. standard html looks pretty much the same on all browsers.

    Lotta




    how about a proxy filter? (3.50 / 2) (#99)
    by semis on Thu Jun 01, 2000 at 06:54:34 AM EST

    What if someone wrote a nifty little proxy, that netscape'ified HTML in realtime? Could this be our answer?

    Don't be ashamed of using pico! (3.00 / 1) (#106)
    by zsazsa on Thu Jun 01, 2000 at 10:31:13 AM EST

    Instead, make the most of it. I use nano, which is a pico clone that also has some much-needed functionality (such as search and replace, ability to seek to an arbitrary line number, etc).
    http://www.asty.org/nano/

    zsazsa

    working at the wrong end (4.00 / 1) (#111)
    by Anonymous Hero on Thu Jun 01, 2000 at 12:32:46 PM EST

    Some good points, but I can't help thinking we're going at this backwards. As quite a few replies have pointed out, the responsibility lies with webmasters to write compatible code, not with browsers to understand every weird extension under the sun. Again, it's been pointed out here that these days many webmasters are unwilling to make the effort to write clean code or judt don't know HTML (I probably still have some messy pages on my site from the early days when I thought Netscape Composer was really cool - Wince!). What _would_ be useful would be a simple utilty for webmasters that strips incompatible code and replaces it with something sensible (an expanded version of the "Demoronizer" Perl script which gets rid of MS dumbquotes, for example). This would be easier and more constructive than trying to write a whole new browser. Robin

    Re: working at the wrong end (5.00 / 1) (#115)
    by analog on Thu Jun 01, 2000 at 01:16:07 PM EST

    What _would_ be useful would be a simple utilty for webmasters that strips incompatible code and replaces it with something sensible

    You rang?

    [ Parent ]

    Re: working at the wrong end (none / 0) (#127)
    by Caspian on Thu Jun 01, 2000 at 08:16:18 PM EST

    You're missing something. WOULD THE WINIDIOTS USE IT? I'd argue "no".
    Hacking code for the people!
    [ Parent ]
    Re: working at the wrong end (5.00 / 1) (#137)
    by Anonymous Hero on Fri Jun 02, 2000 at 12:43:30 AM EST

    Stop calling them idiots. You arent exactly arguing from a position of competence yourself. The issue you have raised is completely contrived. It is not an issue for ANY webmaster worth the title.

    (1) As a user, you can use whatever browser you wish.
    (2) As a webmaster, you can code for whatever browser you wish.
    (3) Where the previous two dont cross, its money (fame, glory, agenda, whatever) out of your pocket and into that of your competitor.

    Where's the fscking problem? Correct, there isnt one. I have yet to visit an even semi-authoritative site on any subject that had a problem rendering on any browser including recent versions of the 3 major text browsers out there.

    This is silly. Your whole arguement boils down to "There are idiot webmasters out there and we need to write a browser to accomodate them."

    Nonsense.



    [ Parent ]
    Draft Javascript Rant (3.00 / 1) (#119)
    by kmself on Thu Jun 01, 2000 at 03:00:13 PM EST

    As it's been requested, the following is a draft of the rant I routinely send to webmaster@fsckingbrokenjavascriptusingsite.com. It's a draft, which means I'm still thinking of ways to make it hurt more without killing the victim (some forms of suffering are good), and maybe actually getting them to change their evil ways.

    It doesn't need to be longer (it probably wants to be shorter), it does need to be better. Suggestions welcomed.

    Here goes.

    Draft Javascript Rant

    Welcome. You are the lucky winner of my Standard Java/Javascript Rant.

    Your website uses one or more of the Java and/or Javascript, including for such basic functionality as hyperlink anchors. This is poor design and renders the site useless under commonly used browsers and configurations. http://www.stanford.edu/~dbrumley/Me/javascript.htm

    Javascript is a known source of a number of security holes and attacks on web browsers:

    http://www.cert.org/vul_notes/VN-98.06.ms_jscript.html http://www.cert.org/advisories/CA-97.20.javascript.html http://www.uncle.com/emailgrab.html http://www.upenn.edu/computing/security-privacy/java/java_security.html http://www.crackmonkey.org/fanmail.html http://www2.merton.ox.ac.uk/~security/bugtraq-199904/0074.html http://netsecurity.about.com/compute/netsecurity/msub35.htm?once=true&

    You might also try searching "javascript exploits" and "javascript bugs" in a major search engine sometime:

    http://www.google.com/search?q=javascript+exploits http://www.google.com/search?q=javascript+bugs http://www.google.com/search?q=javascript+security

    ...or go through the CERT and BugTraq archives.

    http://search.cert.org/ http://www.securityportal.com/list-archive/bugtraq/

    Javascript is a nonstandard extension, not supported by all browsers, and in particular not supported by text-based browsers, handheld and PDA devices. It may raise ADA (American Disability Act) compliance issues by being incompatible with common text-to-speach software.

    http://www.libertyresources.org/news/news_22.html

    Even where it is not a particular risk, it greatly reduces stability and usability of a number of browsers. Though not a direct risk, it can contribute to problems when the browser fails during a transaction or other process.

    Hyperlinks are an html feature already more than adequately supported via HTML anchor (<a href="URL">) tags.

    Consequently I will not browse arbitrary sites with Javascript enabled.

    Arguments I've heard for including Javascript are many, including:

    • It reduces server load on our site with over 1 million impressions/month

      If you're serving a million sessions a month, that's about 1/30th the load of servers I'm familiar with, which certainly don't rely on javascript. You're getting maybe 23 sessions a minute average, figure peaking max 100 times that (I'm being generous), it doesn't wash. Sorry.

      Your site uses shtml, asp, cgi, foo? You're already doing plenty of server-side processing

      ...and besides, you're using javascript for anchor tags>? Um, just how much load does clicking a link put on your server?

    • Javascript provides an easy way of providing flexible, configurable, interactive content? Try PHP on for size. Plus, you're guaranteed it will work since *you* are doing the processing, in a known environment.

    If you insist on adding chrome to your site, do yourself the favor of not disabling core functionality in the process.

    --
    Karsten M. Self
    SCO -- backgrounder on Caldera/SCO vs IBM
    Support the EFF!!
    There is no K5 cabal.

    Re: Draft Javascript Rant (none / 0) (#122)
    by analog on Thu Jun 01, 2000 at 05:07:36 PM EST

    Hmmm. That's a bit polite to be considered a rant, I think. It almost sounds like you're doing them a favor (which, considering the number of people who are probably using javascript in ignorance, I guess you are).

    Right off hand, I'm not sure how to really improve it (vs. just changing it, if you know what I mean). You might want to put how many hits the google searches return in parentheses next to the links; you'd think it would have to get their attention that "javascript exploits" returns over 2000 pages.

    In the area where you mention handhelds and PDA's, maybe mention that by some estimates up to 75% of all web accesses will be done through these types of devices by 2003 (can't remember where I saw that; I'll try to scare up a link).

    Maybe make the line about you not browsing with javascript enabled a little more generic. Something along the lines of s/I/knowledgable individuals. They may not care about losing your business, but if they think savvy customers may see the site and think they don't know something they should, they might be more willing to rethink things.

    Actually, you've kind of given me an idea. How about using this as a basis for a rant-o-matic? You (I, someone) could set up a web form that allowed for choosing which specific things a site had done, what consequences it had/could have, and given a webmaster's email address it would generate an email with appropriate comments and references and mail it off. While I have sent webmasters rants of my own (not as extensive and well supported as yours), I don't do it every time I come across one of these sites. If it were a little easier to put one together that was specific to the case at hand I might send them out more often. Whatcha think?

    [ Parent ]

    Re: Draft Javascript Rant (4.00 / 1) (#126)
    by Caspian on Thu Jun 01, 2000 at 08:15:18 PM EST

    Already planning this. I've registered WinHTML.com for combating clueless "requires this, that, and the other" (usually, one of "this, that and the other" is "Micro$oft Windoze") sites. A "rant-o-matic" will be part of this. It will bother the person about the site until they fix it, and talk to me to get the bothering turned off.
    Hacking code for the people!
    [ Parent ]
    Re: Draft Javascript Rant (none / 0) (#135)
    by Anonymous Hero on Fri Jun 02, 2000 at 12:02:17 AM EST

    Ummm, not quite what I had in mind. I was thinking of a convenient way to point out specific problems with a site to its webmaster. What you're contemplating sounds a lot like harassment (which, if you do it right, could easily carry legal consequences; fwiw).

    Try thinking in terms of constructive action...

    [ Parent ]

    Re: Draft Javascript Rant (none / 0) (#147)
    by KindBud on Sat Jun 03, 2000 at 01:34:05 AM EST

    Having just gotten my web-enabled PCS phone, I can tell you no ordinary HTML web site is visible on them, because they use a new format called WAP. You can only visit sites that emit WAP markup with your web enabled phone.

    But that's a fact that supports the main argument of this thread. HTML has been so corrupted by the browser wars - Netscape and Microsoft are both culpable, and Netscape started it with <center> and <blink> - that the cell phone people had to dream up a new markup that was lightweight enough to be practical for small devices.

    XML is what the commercial web wanted HTML to be, but it wasn't, and the attempts to make it so have ruined it for the purpose it was originally intended. And the W3C went along with it.

    It's a lot like the situation with domain names that another thread is talking about.

    --
    just roll a fatty

    [ Parent ]

    Re: Draft Javascript Rant (none / 0) (#150)
    by rusty on Sat Jun 03, 2000 at 02:19:28 AM EST

    Psst-- we're working on WAP-enabled Scoop. You could read K5 on the toilet, and isn't that really the dream of modern technology? :-)

    ____
    Not the real rusty
    [ Parent ]
    Slight Problem (none / 0) (#121)
    by Anonymous Hero on Thu Jun 01, 2000 at 03:59:13 PM EST

    You are obviously very unhappy about the fact that many sites are optimised for MS Internet Explorer and many of your points are valid. There are a few problems with your arguement. You assert that b/c your operating system doesn't have an option to use Internet Explorer no one should design Internet Explorer only sites. While i hope that most people would take the time to design their pages to work in Netscape also the fact remains that it is only 1/3 of the market and thus many people don't have the time or the care to tweak their page. YOU! have made a decision to use a non-mainstream desktop/workstation OS so unfortuantly YOU have to deal with the an effect of that choice. Also, I take great offense to the assertion that all Microsoft products are ran by "WinIdiots" or people who are too busy making money on Windows. I for one am not computer illiterate and have plenty of experiance with Unix variants but choose to run Win2K on my workstations. While i disagree with their marketting stradegys they have made a damn fine OS that rarely crashes, opens all web pages, runs games, runs Office 2k, runs photoshop and supports strange oddball hardware. Win98 can't stand up to a comparision to a highly stable server platform like Linux and the BSD's, it wasn't designed too so it isn't fair to compare the two. Win2k really does work for many power-users and you shouldn't be so quick to declare "WinIdiots" it makes you sound petty, shallow, and childish. Your ONLY real point seems to be that the *nix's browser's suck so their should be a new one, well don't complain, it was your choice, when you go agianst a majority (no matter how much better and enlightened you percieve yourself to be) you are going to have to deal with some annoyances. DEAL WITH IT, not everybody needs to cater to you. ShaunBaker@houston.rr.com BTW: a browser SHOULD deal with the \ and should finish/fix pages written poorly or corrupted durring transfer. Maybe us "WinIdiots" have the better product and use it

    Re: Slight Problem (3.00 / 1) (#124)
    by error 404 on Thu Jun 01, 2000 at 06:17:29 PM EST

    If a webmaster makes the concious decision that IE optimization is worth blowing off some percentage of the potential viewership, that's fine. A valid decision under some circumstances.

    But a webmaster should be aware of the choice.

    And where possible, the site should degrade with grace - that is, when it is viewed with a browser that isn't up to snuff, something decent (although not as cool as it could be) should appear. Common courtesy, which in commercial environments translates into cold, hard cash.

    Degrading with style also tends to make the site much more friendly to users with special challenges like vision or bandwidth problems. I, for example, often browse with graphics off because I have no patience. When I find what I want, I turn graphics back on. And if your site doesn't provide me the clues I need to realize I've found something good, well, you lose my eyeballs. That has nothing to do with my selecting a weird OS, (in fact, I tend to browse that way more under NT than under Linux, because turning graphics on and off in Netscape bites) there are lots of people like me, and we spend money.


    ..................................
    Electrical banana is bound to be the very next phase
    - Donovan

    [ Parent ]
    Re: Slight Problem (none / 0) (#140)
    by FoodMike on Fri Jun 02, 2000 at 09:30:35 AM EST

    Argh....
    I really think that people who hold opinions similiar to yours are the reason why the technology market is so screwed up. "Computer literate" people are the ones who should be advocating choice. Urging competition, not saying, "DEAL WITH IT." The market for computers and software is just beginning to take shape. Why in the heck are some "educated" people in the field supporting a movement away from industry standards and toward proprietary, patented, and secret protocols and formats. That is what you are doing. You may not think so, but you are. When everything starts to sucks really bad, we can thank everyone who thinks like you.

    Thanks.

    [ Parent ]
    Konq & bmp (4.00 / 1) (#123)
    by Anonymous Hero on Thu Jun 01, 2000 at 05:35:17 PM EST

    I have not looked (I can't find such a thing as a bmp on a page), but: If konqueror doesn't support BMPs, it's trivial to add it, as long as there is any unix codebase that supports reading them. As a side note: Konqueror "requires KDE" in the same way Mozilla requires gtk. It requires that some libraries be installed. Those libraries use less than 15MB of disk space. That's not even 10 cents of disk space. Get real, that requirement is nothing.

    Code for HTML 2.0 (3.00 / 1) (#128)
    by TomG on Thu Jun 01, 2000 at 08:41:20 PM EST

    Write web pages in HTML 2.0. End the insane browser wars.

    And if you want to have a little fun, put a \ in your url, IE will clobber it into a /, producing a 404.

    Heh, you think I'm joking.



    Okay.. if HTML is so easy.... (none / 0) (#139)
    by Dacta on Fri Jun 02, 2000 at 06:05:13 AM EST

    Sorry to have to do it, but when you write an article like that, you have to expect some bastard to do this.

    Caspian's home page is at http://caspian.twu.net/. So lets validate it......

    Okay, some errors, most of which are pretty minor (Error: value of attribute "VALIGN" cannot be "CENTER"; must be one of "TOP", "MIDDLE", "BOTTOM", "BASELINE", stuff like that)

    But then we look at his table code... Basically the W3 validator is saying it is stuffed. Looking at his code, it looks correct to me, but the W3 validator doesn't lie, so your vim written HTML is really just as bad as the MS Word crap you hate so much, but it happens to render correctly in Netscape.

    Here's the output (just cut & paste from teh source - sorrry about the <span> tags):

    <h1>HTML Validation Service Results</h1> <h2>Document Checked</h2>

    Below are the results of attempting to parse this document with an SGML parser.

    • Line 1, column 1:
        <code class=input><html>
         <span class=markup>^</span>

      <span class=error>Error: Missing DOCTYPE declaration at start of document (explanation...)</span>

    • Line 140, column 4:
        <code class=input></td></tr></table>
            <span class=markup>^</span>

      <span class=error>Error: end tag for "DIV" omitted, but its declaration does not permit this</span>

    • Line 16, column 0:
        <code class=input><div align="left">
        <span class=markup>^</span>

      <span class=error>Error: start tag was here</span>

    • Line 160, column 22:
        <code class=input><tr><td valign="center" align="left">
                              <span class=markup>^</span>

      <span class=error>Error: value of attribute "VALIGN" cannot be "CENTER"; must be one of "TOP", "MIDDLE", "BOTTOM", "BASELINE"</span>

    • Line 162, column 96:
        <code class=input>... /images/hebword1.jpg" width=135 height=40 border=0></a>
                                                              <span class=markup>^</span>

      <span class=error>Error: required attribute "ALT" not specified</span>

    • Line 164, column 95:
        <code class=input>... ./images/chnword2.jpg" width=44 height=40 border=0></a>
                                                              <span class=markup>^</span>

      <span class=error>Error: required attribute "ALT" not specified</span>

    • Line 167, column 18:
        <code class=input><td valign="center" align="center">
                          <span class=markup>^</span>

      <span class=error>Error: value of attribute "VALIGN" cannot be "CENTER"; must be one of "TOP", "MIDDLE", "BOTTOM", "BASELINE"</span>

    • Line 175, column 18:
        <code class=input></center> </center></div> </td>
                          <span class=markup>^</span>

      <span class=error>Error: end tag for "DIV" omitted, but its declaration does not permit this</span>

    • Line 168, column 0:
        <code class=input><div align="center"><center>
        <span class=markup>^</span>

      <span class=error>Error: start tag was here</span>

    • Line 175, column 18:
        <code class=input></center> </center></div> </td>
                          <span class=markup>^</span>

      <span class=error>Error: end tag for "TABLE" omitted, but its declaration does not permit this</span>

    • Line 159, column 0:
        <code class=input><table width="100%" border=0 cellspacing=0 cellpadding=0>
        <span class=markup>^</span>

      <span class=error>Error: start tag was here</span>

    • Line 175, column 30:
        <code class=input></center> </center></div> </td>
                                      <span class=markup>^</span>

      <span class=error>Error: end tag for element "TD" which is not open</span>

    • Line 176, column 18:
        <code class=input><td valign="center" align="right">
                          <span class=markup>^</span>

      <span class=error>Error: value of attribute "VALIGN" cannot be "CENTER"; must be one of "TOP", "MIDDLE", "BOTTOM", "BASELINE"</span>

    • Line 176, column 33:
        <code class=input><td valign="center" align="right">
                                         <span class=markup>^</span>

      <span class=error>Error: document type does not allow element "TD" here</span>

    • Line 179, column 9:
        <code class=input></td></tr></table>
                 <span class=markup>^</span>

      <span class=error>Error: end tag for element "TR" which is not open</span>

    • Line 179, column 17:
        <code class=input></td></tr></table>
                         <span class=markup>^</span>

      <span class=error>Error: end tag for element "TABLE" which is not open</span>

    <hr>

    Sorry, this document does not validate as HTML 4.0 Transitional.



    Re: Okay.. if HTML is so easy.... (none / 0) (#149)
    by Caspian on Sat Jun 03, 2000 at 01:58:51 AM EST

    What the hell is your point? I don't write to please the HTML validator. I write to please Netscape *AND* IE *AND* Lynx. My page works in all three. So what the hell is the problem?
    Hacking code for the people!
    [ Parent ]
    The point (none / 0) (#151)
    by Dacta on Sat Jun 03, 2000 at 03:41:24 AM EST

    The point is that there is probably a browser out there that it won't work for. You might not use that browser, but you've been making the point that browsers that correct/compensate for webmaster errors are in error, so that means that you are no better than those people who rely on IE correcting errors in their HTML.

    Okay, so you check it in Netscape & Lynx as well, but what about Netscape 1 or something? Don't you dare claim "But no one uses Netscape 1" - what if they were on an old 386 with Windows 3.1? Nescape 1 was my browser of choice when I had one of those, and I guess there are still some people (esp in countries less well off) that use it. (Any you can't run Linux/X in less than 5M of memory - I've tried. It was in about 1994, but I don't think it's got any smaller since then!)

    You've been claiming that market share should be ignored, so ignore it, and code for all browsers.

    That's my point.. and a hell of a point it is! *S*.

    If you disagree with it, then you must agree that browsers that compensate for mistakes are a good thing, and IE does that better than anything else.

    [ Parent ]

    Re: The point (none / 0) (#155)
    by Caspian on Sat Jun 03, 2000 at 08:32:39 PM EST

    Idiot, idiot, idiot. First of all, I DO try to make my pages work properly in Netscape 1.x. Secondly, who CARES how good IE is? I STILL CAN'T RUN IT.
    Hacking code for the people!
    [ Parent ]
    Re: The point (none / 0) (#157)
    by Dacta on Sun Jun 04, 2000 at 03:28:03 AM EST

    Idiot, idiot, idiot?

    Hmmm. Are your trying to start a flamewar?

    Okay. Now we come to the point of your arguement:

    Secondly, who CARES how good IE is? I STILL CAN'T RUN IT.

    I care how good it is. I appreciate quality software, and I think IE is a pretty good example of that.

    As for not being able to run it - in your post you pointed out quite well how you could fire up VMWare and run it from there.

    Okay, it's not on Linux etc etc and we can go on all day like this.

    Your arguement is Full of holes.

  • Internet Exploror is the most standard compliant browser in exsitance, except for Mozilla which you admit you haven't tried. (It's only a 6.5M download, BTW)
  • IE exploror has some features that allow rendering of incorrect HTML. All browsers have these features, but IE does it best.
  • Netscape is a piece of crap so far as HTML compliance is concerned, and it has a quickly dropping marketshare, which means that many webmasters will increasingly stop supporting it.
  • Finally, Mozilla is what we need, not a "SAMBA of webbrowsers". Mozilla is standards compliant, and will render stuff correctly when Netscape won't.

    There's my points. Back to that idiot thing - I think you are an unintelligent person who has show themself to have a mind closed to new ideas. You have been proved to be wrong in many of your points and have resorted to name calling and meaningless personal insults instead of trying to learn. I find behaviour like that beneath contempt, and I hope one day you learn that your IRC-cool bluster isn't going to get you very far.

    [ Parent ]

  • HTML can be adjusted - M$ protocols can not (none / 0) (#141)
    by pacc on Fri Jun 02, 2000 at 11:20:04 AM EST

    A page not working on Netscape is just plain negligance from the webmasters side.

    Another deal altogether is corporate intranets. Websites on NT servers usually autheticate users with NTLM, only existing in Windows API's and not used by Netscape. Thus many many intranets is switched to IE-only overnight when the internal webserver is and the support just tell you to use IE.

    Maybe the success of this made Microsoft to try the Kerberos hack? I don't judge any browser over another, but once behind I belive netscape will have a hard time now that thousands of users have to convince their admins to get another security system for their intranet webpages.



    frontpage has support for netscrape to you know.. (none / 0) (#142)
    by MrSparkle on Fri Jun 02, 2000 at 11:38:18 AM EST

    Frontpage, or at least frongpage 2000 has an option to make it output code that is compatable with netscape or ie, or both. Frontpage still sends out crap html code but at least it i a step in the right direction.
    -----------------------------------------------
    I'm the guy that has a problem with everything.
    Vote with your visits (none / 0) (#148)
    by skim123 on Sat Jun 03, 2000 at 01:47:17 AM EST

    Rather than complaining that a Web site doesn't work with your browser, show your support for those that do work with your browser by visiting those sites. If a site doesn't work with Netscape, inform the Web master, letting him/her know that you would visit regularly, but the site cannot be rendered with your browser. If it is any decent webmaster, they will work on it to make their visitors happy and continually visiting.

    What you shouldn't do is get all pissy and demand that folks conform to your OS/browser choices. If you cannot view it in your browser on your OS, use a different browser/OS. If you don't want to do that, use lynx or telnet to port 80 on the Web server and download the textual HTML and pick through it yourself. If you don't want to do that, don't visit the site. Plain and simple.

    Agreed, browsers for UNIX suck. I am a big IE fan and would love to see a great IE for Linux. Perhaps this will come about if MS is split up...

    Money is in some respects like fire; it is a very excellent servant but a terrible master.
    PT Barnum


    WINE and browser plugins (none / 0) (#152)
    by spiv on Sat Jun 03, 2000 at 06:51:54 AM EST

    Here's a small thought...

    This is kinda tangential to the issue, but part of the problem is the lack of support for browser plugins which are Windows only.

    How hard is it to get WINE to make Netscape plugins for Windows work with Netscape for Linux? If I recall correctly, the plugin API for Netscape is essentially the same between platforms, so all that's needed is for WINE to intercept the Window's specific stuff...

    I'm tempted to think it must be possible, because I recall seeing on freshmeat some months ago a VQF module for XMMS - it didn't actually play VQFs itself, but rather used WINE to access the .dll of the WinAMP plugin!

    The open-source nature of Mozilla should make this sort of hackery (relatively) straightforward to accomodate...

    -Spiv

    Vote with the visits (none / 0) (#153)
    by WWWWolf on Sat Jun 03, 2000 at 09:21:55 AM EST

    "The customer is always right."

    If I can't see some company's web site, I ask them to fix the site.

    If they refuse, I tell them "You failed to serve me, so maybe I'll buy my stuff from some other place."

    ...and I snail them a letter, telling their customer service stinks.

    Web site usability is integral part of customer service. If the customer doesn't see the page, then they'll go away. Thus, if the web designers have the slightest bit of stuff sparking between their ears, they will sooner or later realize that if the site isn't visible to some users, those users will not buy their stuff and that's a Bad Thing!

    Netscape, even in Windows world, still has a big market share, so ignoring them will result in decreased sales...

    W3C is there to set the standards. Not Microsoft. Or, actually, as far as I know Microsoft does participate in W3C's work, but then again, so does Netscape.

    PS. Have you ever seen how much space BMPs take compared to GIFs/PNGs/JPEGs? They take a LOT of space. That will conviniently scare the modem users away, no matter what OS. =)

    PPS. "They all suck. Except for MSIE, which sucks raw eggs through a very thin straw."

    -- Weyfour WWWWolf, a lupine technomancer from the cold north...


    Not so easy (none / 0) (#158)
    by Anonymous Hero on Wed Jun 07, 2000 at 01:33:08 PM EST

    Samba team does great, but they are resolving much easy problem - Microsoft cannot change their proprietary network protocols just for kill Samba. Eventually they do, but with old 95 and even 3.x users in mind. If that people once figure out, that new Windoze WILL NOT work with their stuff they may stop buy anything from M$. Internet browsing is not a market anymore. Microsoft completely unstoppable from change their proprietary formats every day and even every hour. More and more windoze boxes now automatically download any updates from Internet - people even may have no idea that their browser changing everyday. Building browser from scratch to handle all W3C + all M$ is utopia. But if you would succeed in it and it become a treat to M$, they simply start to change their formats and protocols in so HIGH rate, that even supe-hacker team will give up. And some little Mozilla advocacy. M15 milestone is definitely tirn point, because this browser as reliable as Netscape, but faster and capable to render richer amount of standards. It is not true that Mozilla origins from Netscape. It based on Gecko rendering engine, rewritten from scratch. Almost all code now new, completely rewritten. Yes, it still slow, but reason is support for a lot of standards, not bad code. IE will become slow too. Only Mozilla weakness is cross-platformness - Windoze people dont need it, so why to waste time? On other hand good-coded cross-platform app never can be as faster as good-coded single platform app, because cross-platformness implies additional abstraction layer over different OS/GUI environment calls.

    On the Need for "the Samba of Web Browsers" | 157 comments (157 topical, 0 editorial, 0 hidden)
    Display: Sort:

    kuro5hin.org

    [XML]
    All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
    See our legalese page for copyright policies. Please also read our Privacy Policy.
    Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
    Need some help? Email help@kuro5hin.org.
    My heart's the long stairs.

    Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!