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

Pride and Prejudice: The Platform War Comes to the Web

By Caspian in Culture
Thu Jul 20, 2000 at 04:24:55 AM EST
Tags: Internet (all tags)

   A long time ago, when the word "Web" wasn't thought of as a synonym for "Internet", when the name " Tim Berners-Lee" was more often associated with it than the name "Bill Gates", when Netscape Navigator 1.x was the hottest program online and Internet Explorer was but a glimmer in Redmond's cyclopean eye, the fantastic new technology of the World Wide Web offered a hope for something as yet unheard of in large quantities in the computing world. That something was one of the great Holy Grails in computing, right up there with artificial intelligence, realistic "virtual reality" and the final elimination of the plague of BASIC derivatives. That mythical something, that Grail, that wonderful boon which the newly-fledged Web promised the world was what is commonly called "platform independence".

   As many of us know, "platform independence" is the concept of a program, protocol, communications medium or other computer-related technology not being tied to a particular software or hardware base-- for instance, while PCI technology is available on multiple platforms, Apple's now-defunct NuBus technology was tied to Macintosh computers; thus, NuBus was not a platform-independent standard, but PCI is. The JPG, PNG and GIF technologies used in images on the Web are good examples of platform-independent file formats-- these formats are readable by virtually any sort of computer in the world. Still another example of platform-independence is Adobe's vaunted "PDF" format, a descendent of the DTP standby "PostScript" (which also was platform-independent).

    A large part of what I do in my spare time involves running a non-profit Web (and e-mail, among other services) hosting provider called The Web Union. Aside from the lack of a profit motive, TWU is a relatively unremarkable hosting provider as far as user base goes. This means, of course, that I host a few Mac OS users, a sprinkling of Unix users, and vast legions of Windows users. Hundreds of them. Most of them use IE to get to the site, and that's alright with me. I certainly make every effort to ensure that the TWU pages, as well as my own personal site, look just as nice in IE as in Netscape (or Lynx). In fact, things were going just fine in this arena until recent events prompted me to take action.

    First, many of my users took to the habit of displaying little "best viewed with Microsoft Internet Explorer" buttons on their sites. Alright, so it's annoying, but no harm done. I made my opinion known, but otherwise did nothing-- all users certainly have the right to voice their opinions as to which browser, OS or whatnot that they prefer.

    A few months after my users started peppering their sites with pro-IE buttons, though, I began to notice a far more disturbing trend. A small but growing number of my users' sites were completely unviewable in my browser. By this, I don't mean "ugly"-- I mean broken. Some of them would display as blank pages in my browser, but would work fine when I would start up IE for Windows in an emulator (I don't actually run Windows on any of my machines, but merely use it in an emulator as needed-- which is rare for me). Others displayed (poorly), but could not be navigated. Images mysteriously overlapped text, unusual question marks appeared where quotation marks should be, some characters disappeared entirely-- but only when viewed outside of IE.

   In short, a growing number of my users' sites were unviewable to me. Not only could I not see, for purposes of personal curiousity, what sorts of uses my server (which I pay for out of my own pocket) was being put to, but I couldn't even patrol the site for violations of TWU's few and simple rules!

   For a while, I took to simply sending out e-mails to the offending users, telling them to modify their sites so that they were no longer dependent on IE (or Windows). This worked for a while, but eventually, I got fed up with having to do this. I decided to take a stand on the browser compatibility issue once and for all-- I made it a violation of TWU's policies to host a platform-specific site.

   This change in policy prompted a rather disturbing collection of e-mails from my users. The feedback I received seemed to primarily fall into three categories:
  • Users who were grateful for the free hosting I provided, and hence were asking what they could do to stay compliant-- and therefore keep their sites.
  • Linux/Unix users praising me for my stance on platform neutrality.
  • Windows users complaining about the alleged unfairness (!!!) of the new policy.
   Keep in mind that the policy was not "all of your sites must be viewable solely in non-IE browsers". It wasn't even "all of your sites must look just as nice in non-IE browsers as in IE". It was simply "all of your sites must WORK in non-IE browsers". Or, in short, I was telling my users "if you want me to pay for your hosting out of my own pocket, you have to play by my rules. One of those rules is that your site must not, within reason, exclude users-- including, but not limited to, your own sysadmin, who runs Netscape for Linux-- based on their choice of browser or OS."

   It sounds like a reasonable, level-headed policy, doesn't it? My users, apparently, didn't see it as such-- at least, those users who ran Windows. I was called everything from "silly" to "fascist"-- the latter invective flung by none other than a very old friend of mine, a long-time FreeBSD user who uses Windows on his personal workstation (and BSD solely as a server-side OS).

   Perhaps the proverbial "straw that broke the camel's back" in motivating me to write and publish this article, though, came when this particular friend of mine suggested a solution to my dilemma.

   He suggested that I change OSes.


   Take a moment to read that statement again. Let it sink in. My friend was suggesting that I install a new operating system so that I could properly use the World Wide Web. Yes, the Web, that wonderful platform-independent communications medium...which, at the rate things are going, will soon virtually (if not literally) require Internet Explorer 5 for Windows to get any meaningful work done.

   All of a sudden, things don't look so terribly platform-independent on the Web front, now, do they?

   Before I continue, perhaps you'd like to hear what Tim Berners-Lee, father of the Web, has to say about platform-specific Web design. A quote of his on the subject is available here. In a nutshell, Lee, even back in 1996 when the Web was still relatively new and platform-independent, wrote that the people who pepper their pages with "best viewed with (whatever)" buttons seems to long for "the bad old days", as he puts it. The "bad old days", that is, of platform dependence!

   Slowly, surely, without anyone noticing, the Web is bieng transformed right under the world's collective noses. It is being transformed from its initial state-- where pretty much any site would work pretty much okay in pretty much any browser, running on pretty much any OS-- to a new state, perhaps best named "Microsoft World Wide Web 1.0 for Windows". The concept that Web sites can (and should!) be made to work properly in multiple browsing environments (browsers, OSes, even font loads!) is slowly being degraded to the point where many users don't even know that it exists; when it is introduced to them, their reaction is, as often as not, apathy. What's most disturbing about the new trends in Web design is that the devolution of the Web from a platform-neutral medium to a platform-specific medium hasn't been prompted by corporate bribery, meaningful technological incentives or intrusive government regulations. It's been prompted by the collective apathy of tens of thousands of Web developers, and the collective ignorance of millions of Web users.

   I have brought up this issue (which far too many IE for Windows users-- a group which includes a somewhat disturbing 80-plus percent of all Web users by most estimates-- are completely unaware of) with many of my contemporaries, and have been snapped at virtually every time by sizable contingents of said peers. "IE is faster", they say. This is, by my experience, true. "IE is better!", they say. This is also true-- at least, from a technical standpoint. So what-- does this mean that we should require everyone on the roads to buy a Ferrari, and simply be forced to put up with the associated monetary cost? Of course not. Similarly, we should not require everyone on the Web to run Windows and IE, and therefore be forced to put up with the associated "costs": Windows's instability and, of course, being a pawn to Gates's vision of the future (a vision, not coincidentally, of a Windows machine on every desk, and Windows CE-- or "Windows Powered", as I believe it's being renamed-- in every toaster.) The plain and simple truth is that it is wrong to attempt to twist every computer user's arm until they run the browser that Microsoft wants them to run, in the OS that Microsoft wants them to run. Where I come from, that's called "bullying".

   But perhaps the most oft-repeated, and in some ways most convincing, argument in favor of "everyone in the world using IE", has been that IE is more "standards compliant" than Netscape. That is, so says the pro-IE camp, Internet Explorer more closely follows the standards proposed by the World Wide Web Consortium, the standards body that has been overseeing the development of the Web's standard protocols and languages virtually from its inception. While the argument that IE is more standards-compliant is true, it also fails to recognize other important truths-- it is the truth and nothing but the truth, as a judge might say, but it is not the whole truth. This merits some explanation:

    As a reminder, Microsoft's history in the "standards" arena boils down to a series of choices between two unfavorable (well, for anyone other than Microsoft and their fans) alternatives. In some cases, Microsoft simply chooses to spurn the existing body of standards (both written and unwritten) entirely, or at least as much as they can get away with it. (Witness Windows's pitiful shred of POSIX-compliance, placed in almost as an afterthought to prevent the software titan from losing valuable government clients-- clients that demanded POSIX features.) In others, Microsoft seems, on the surface, to follow the standards...however, upon closer inspection, it's revealed that they're subtly twisting the standards in such a way that only Microsoft products work with them. This is commonly known as the "embrace, extend, extinguish" tactic, and it is what Microsoft attempted in a recent, highly publicized spat over their upcoming Kerberos implementation.

    The latter standards-compliance (or lack thereof) tactic-- "embrace, extend, extinguish"-- is what Microsoft has been using for Internet Explorer. While IE advocates frequently claim that IE is much more "standards compliant" than Netscape-- which is true-- they forget two important points. The first, and the most obvious, is that in addition to supporting the W3C's standards, IE supports all sorts of additional features which, if utilized in a page, renders them completely incapable of being viewed properly under any other browser. Two examples which come to mind are IE's support for inline .BMP files (why in the world anyone would want to use inline .BMPs is beyond me-- after all, the files are huge-- but I've seen it done) and support for backslashes (\) instead of forward slashes (/) in links, image source attributes and background attributes. None of these features are available in many, if any, non-IE browsers; given the fact that neither of them grants any actual advantage to Web developers (Why use .BMPs, which are huge, when you can use .PNGs-- which most modern browsers support, and which are much smaller, but just as non-lossy? Why use backslashes instead of forward-slashes in paths-- what's the advantage?), it is painfully clear that these "features" were included in IE merely to encourage the creation of IE-specific pages. Sadly, the tactic has worked. That is precisely what is happening today.

    Even if Microsoft decided to remove the loony new features (e.g. .BMP support, backslash-instead-of-forward-slash support, etc.)-- which they won't-- a critical point ignored by the pro-MS camp remains: Microsoft did not put those features in IE to please its users, or even simply to make IE a better product. The simple truth, as evidenced by Microsoft's tactics over the past decade, is that Microsoft doesn't give a whit about standards compliance. Just like anything else, standards compliance exists, in Microsoft's eyes, as solely another tool in the Redmond giant's box, to be pulled out whenever it can be used to gain them money or market share. The only reason Microsoft put the various standards-compliance features into IE is that Netscape hadn't put them into Communicator. Therefore, by beating Netscape to the punch in the standards compliance game, IE could be made to properly render sites which Netscape would puke on. This is what happened, and the result is that IE's market share has continued to grow and grow.

   "But wait," you might be thinking... "so those HTML features that Netscape can't handle are SUPPORTED in IE... so what? That doesn't mean anyone's going to use them! I've been using IE for Windows for ages, but I've never seen a site say that it requires it!" Of course they don't say they require IE for Windows-- the assumption nowadays is that all users are running it. Many IE for Windows users are completely unaware of what their pages look like in other browsers, other operating systems, or even earlier versions of their own browser! For the curious, here is a table showing the same sample page-- a relatively simple page containing one table, peppered with just a few of the platform-dependent features often found on IE users' sites. Some of these screenshots were contributed by volunteers, to whom I am very grateful. (By the way, to my fellow Unix users-- if you're thinking "so run Mozilla already!", look at the table-- Mozilla's results are printed there, and they're no better than the others.)

(To browse these screenshots, go to http://caspian.twu.net/ieonly/views/.)

   Back to the standards-compliance issue, though, for this is what is truly at the heart of the "IE for Windows vs. Anyone Else" debate. Don't get me wrong-- I love standards. I think that good standards help make good things better, or even possible where they would otherwise be impossible-- after all, the Internet could not exist had it not standardized on TCP/IP. However, standards aren't everything. What is more important than standards is, quite simply, pragmatism. That is-- if you are a Web developer, can your pages be viewed successfully (if not optimally) in all browsers? If you are a Web user, can you view all the pages that you try to view with your browser(s) of choice? If the answer is "no", then that is what is important-- that disruption of your work. Who cares what caused it-- the point is that Web developers should be expected to make their sites work on as many browsers as feasible (regardless of their level of standards compliance!) and Web users should be able to expect that all sites will work in their browser (regardless of its level of standards compliance!)

    It would be wonderful if all browsers suddenly complied 100% with the W3C standards, but I'm afraid that's just not going to happen any time soon. In the meantime, Microsoft is busy building ever more unusual features into IE, and Web developers worldwide are busy helping the folks in Redmond along-- by making their sites depend wholly on standards solely implemented in Microsoft's software, and often on pseudo-"standards" (like inline .BMPs) invented out of the blue by Microsoft themselves, with no other possible purpose than to make other browsers (Netscape, Mozilla, Opera, Konqueror, Lynx, etc.) less viable.

    It is petty and mean-spirited for Web developers worldwide to continue the process of rebuilding the Web as a single-platform medium. Besides that, why on Earth should developers not being given a red cent by Microsoft themselves essentially act as an unpaid standing army to help build up the ill-gotten monopoly of a predatorial megalomaniac? It makes no moral sense, and it makes no financial sense.    Some IE users have also taken great pains to point out that in the early days, when Netscape enjoyed a lead over everyone else, they too peppered the Web with proprietary tags like <CENTER> and the like. This, too, is true. I'm not claiming that Netscape (or any other company, organization or individual!) is without sin. That isn't what I am writing about. I am writing, rather, about the refusal of Web site authors to put any efforts into ensuring that their sites don't use such features. If it were Netscape on the top of the browser heap again, rather than IE, my point would stand. I would never advocate the creation of a Netscape-only Web site any more than the creation of an IE-only Web site-- and neither should you.

   The point of this essay is not to discourage the use of IE, though that would be a nice turn of events. Rather, its point is to encourage those who develop Web sites to ensure that their sites work on multiple browsers and multiple operating systems before announcing them to the world. I go to great lengths to ensure that my sites work on browsers that I would never in a trillion years use myself. Can't you, dear reader, do the same?

   Last but not least, I wish to encourage other Web hosts to do as I have, and refuse users who develop platform-dependent sites. While it may seem unusual to mention the words "moral" and "computers" in the same sentence, there are ethical and moral issues involved in computing, and I firmly believe that transforming an open, platform-independent community like the early Web into a closed, platform-dependent community dependent on a single company (particularly one that has demonstrated time and again that its sole interest is having not merely billions of dollars (this is perfectly alright), but an effective monopoly in multiple areas of the computing world (this ISN'T)) is, indeed, immoral. Don't be a part of Bill Gates's vision of the future-- support platform-independent Web design!

    In conclusion, the Web has become the victim of pride and prejudice-- non-Microsoft users being too proud to admit that there are more important things than standards compliance (for example, oh, things actually working) and Microsoft users being irrationally prejudiced against anything not written in Redmond, and all too willing to turn a cold shoulder to Netscape, Unix, Mac OS, and so on. It will take a stupendous amount of effort to reverse this trend. Want to help? Good. You can start by making some noise. Don't stand silent in the path of the MS steamroller, or you'll soon find yourself flat as a pancake.


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


Related Links
o ZDNet
o a synonym for "Internet"
o Tim Berners-Lee
o The Web Union
o personal site
o FreeBSD
o World Wide Web Consortium
o W3C
o http://cas pian.twu.net/ieonly/views/
o Mozilla
o Also by Caspian

Display: Sort:
Pride and Prejudice: The Platform War Comes to the Web | 69 comments (62 topical, 7 editorial, 0 hidden)
50 million people doing a dumb thing is still a du (5.00 / 1) (#3)
by ramses0 on Thu Jul 20, 2000 at 12:50:20 AM EST

(this is really a topical reply to joshv's comment)

I'd vote +2 on this if I could, and I'll definitely send links to my friends. Editorially, the only two problems were the length, and the gratuitious "no-table" thing.

As for your claims that IE is the best browser- I agree. !!!*BUT*!!! as the author says, IE supports things that are just plain *WRONG* in order to protect it's dumbass userbase from tripping over HTML.

If 80% of the web uses IE/Windows, and 10% of those people use http:\\microsoft.com\support instead of http://microsoft.com/support, then 8% of the web is *BROKEN* for 20% of the population. And don't tell -me- that I should change my software in order to bend over for Microsoft, because then I get to call you the fascist. ;^)=


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

Nice write up, but... (4.00 / 4) (#4)
by Dacta on Thu Jul 20, 2000 at 12:59:40 AM EST

Didn't you post a rant about this a month or so ago?

Yes, here we are. I might have voted for the article anyway (despite the fact that I disagree with your premise), but you ended up being pretty rude in some of your comments.

Even Rusty had to say:

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

Normally, I wouldn't care, but your attitude really made a bad impression on me.

Re: Nice write up, but... (3.00 / 2) (#6)
by Anonymous Hero on Thu Jul 20, 2000 at 01:42:46 AM EST

Very good comment, but ... um, shouldn't this be an editorial comment?

Lately, it seems like people have been applying the editorial/topical comment tags at random! C'mon folks, those tags are there for a reason, and it ain't for turning your comment pretty colors!

[ Parent ]

Authoring tool? (3.00 / 1) (#9)
by ppanon on Thu Jul 20, 2000 at 01:59:41 AM EST

What's the authoring tool being used to generate most of these broken pages? Are these people maybe upset because they can't actually fix the HTML because they "composed" it with the latest version of MS FrontPage? Hey, Microsoft practically gives away Frontpage - I would have expected to find it bundled with Office by now if it wasn't for the fact that they prefer to rev it with IE to take advantage of all those IE proprietary extensions.

If they are using FrontPage for web authoring, you might want to explain to them that their page doesn't just render poorly on IE browsers, but that FrontPage has historically been notorious for generating poor HTML. Any browser not using the standard US/North American character set mapping is likely to see garbage on many characters, including quotes. That means that their site probably looks unprofessional to many people outside of North America - where Internet growth is currently highest (in percentage terms) as the North American market approaches saturation.


Re: Authoring tool? (1.00 / 1) (#13)
by bobsquatch on Thu Jul 20, 2000 at 03:06:07 AM EST

I'm generally in favor of any mean-spirited, obnoxious, holier-than-thou, clueless sanctions against FrontPage users.

After all, FrontPage generally creates mean-spirited, obnoxious, holier-than-thou, clueless HTML. Fair is fair.

[ Parent ]

Re: General Cluelessness [rant] (4.30 / 3) (#35)
by shadowspar on Thu Jul 20, 2000 at 10:19:57 AM EST

I recently had a scrap on the other site with someone about a lot of the same topics mentioned here. (He made some decent points about standards, but his continual slagging of the lynx developers really pissed me off. [1])

There are two groups who I generally find have the worst webpages on the net[2]:

  1. HTML-Clueless Newbies, and
  2. Seemingly-neurotic web developers.

The pages of the first group tend to be the ones with the neon-green text on orange backgrounds, with everything in 24-point letters and blinking. Their crumby webpages are due to a lack of design skills and a general lack of cluelessness about the Internet in general. (I use lynx and w3m, one of their chief advantages being that they filter out such mind-numbing design choices.) Such pages improve once the said people gain clue [3].)

The members of the second group are generally the ones employed by less-than-net-savvy companies who want them to make whizz-bang web pages to be viewed by those in category 1. Their pages tend to look great in (say) MSIE 5.01/Win32 and Netscape 4.71/Win32, but are crap in IE/NS 3/4 or on non-Windoze systems, and heaven forfend if you're one of the heathens that uses Mosaic or Lynx. They seem to be twisted to make them as incompatible as possible with all browser versions except the current majors - one swears that their HTML design is motivated by malice.

I don't actually think that these people screw me around on purpose (as much as I sometimes feel that way[4].) Being a webmaster entails a lot of stress, with $PHB forever nagging you to wrest the impossible from the current leading excuse-for-a-browser. But bloody hell, wasn't the point of HTML, the web, and the internet in general, to make as much information as possible available to as many people as possible?

As much as I realize that the future of the net (supposedly) grows by more and more non-geekfolk getting online, there are still days when I be thrilled to see them all kicked off - first all the clueless and unrepentant companies[5], and then all the clueless and unrepentant users[6].

[1] Called them lazy because Lynx doesn't support JavaScript / ECMAscript yet.
[2] Yes, this is an overbroad sweeping generalization.
[3] Having bobbed[?] on a university helpdesk, this is not generally my experience, but it happens on occasion.
[4] The most frustrating thing, though, is writing to a company about their b0rken web pages, because you pretty much always find out that the really don't give a rat's ass about you.
[5] Think of the drop in spam volume.
[6] Think in the drop of bobs' and sysadmins' stress levels.
-- Drink Canada Dry! You might not succeed, but you'll have fun trying.
[ Parent ]

Backslashes and .BMP files (2.00 / 5) (#10)
by PrettyBoyTim on Thu Jul 20, 2000 at 02:32:22 AM EST

There are very good reasons for the backslashes and .bmp files.

In case you're not aware of it, Microsoft operating systems use backslashes as serpeators instead of forward slashes. Therefore, if you're trying to access a file on the local machine, it makes a lot of sense to allow the user to type in a path in the local format.

I know that Opera (for windows) allows forward slashes as well, and I wouldn't be suprised if Netscape does in it's windows version.

As for .BMP files - The browser is no longer used entirely for browsing content from the web - it's used for browsing content on the local machine as well. Microsoft is pushing the browser to be the centrepiece of their operating systems - they already use it for browsing local files, and they allow thumbnails to be viewed of local files, and that's why they've got the .bmp support.

And so no... I don't think they've put it in there to just piss off those who don't use IE...

Re: Backslashes and .BMP files (none / 0) (#26)
by WattsMartin on Thu Jul 20, 2000 at 08:56:26 AM EST

The problem with this logic is that the browser can obviously tell if what's being viewed is local or non-local. Using backslashes instead of forward slashes in a local filepath should be perfectly acceptable (just like using colons to delineate the path should be on a MacOS browser). Using backslashes instead of forward slashes in a URI violates the standard and the browser shouldn't support it.

BMPs are a fuzzier issue; as far as I know, using GIF, PNG and JPEG images online is merely a matter of convention. Suppose you're browsing on a system like BeOS, where the image handling might be done by the operating system--you just pass the image binary data to the translator and have it return a bitmap to you. Do you add a check to the web browser to deliberately break all but the most common image types?

[ Parent ]
Re: Backslashes and .BMP files (none / 0) (#30)
by Caspian on Thu Jul 20, 2000 at 09:33:07 AM EST

The problem is that IE for Windows will happily accept a remote Web page-- not a local file, but A PAGE OUT THERE ON THE WEB SOMEWHERE-- with inline .BMPs and backslash-filled img srcs. This encourages Web developers to create sites that only work in IE, which they are doing in disturbingly increasing numbers.
Hacking code for the people!
[ Parent ]
Re: Backslashes and .BMP files (none / 0) (#34)
by squigly on Thu Jul 20, 2000 at 09:59:15 AM EST

I think the backslashes thing is more down to a desire to handle broken html. I've never been able to decide whether I think this is a good idea or not. On one hand, if a program can easily guess what is meant when presented with broken data, I generally approve of it fixing it. On the other hand, if a piece of software is used to test data, it shouldn't try to fix things. Web browsers are used for both.

BMP's don't worry me. If people start using them, then I can probably get a plugin or ignore the images as long as they put an alt tag in. I'd rather allow some stupid images than be locked into jpeg, GIF and PNG if something better comes along.

People who sig other people have nothing intelligent to say for themselves - anonimouse
[ Parent ]
Re: Backslashes and .BMP files (none / 0) (#41)
by kkeller on Thu Jul 20, 2000 at 12:09:02 PM EST

There are very good reasons for the backslashes and .bmp files.

I, for one, am still waiting for these reasons to be listed. You cited a reason that the browser might understand backslash in a file: URL, but *not* in an http: URL. Again, the browser can support BMP locally, but people *writing* HTML should not *expect* BMP support from all browsers.

Checking out the renderings at http://caspian.twu.net/ieonly/views/, I can see the browsers that support these features.

--Backslash in path: IE 5/Windows (*NOT* on MacOS), Mozilla M14/Windows (*not* any other OS)

--Inline BMP: IE 5/Windows, IE *3*/MacOS (*not* IE 5 on MacOS!), Opera b7/Linux (*not* Opera/Windows!)

I think this is clear evidence that HTML authors who use these ''features'' are either clueless authors who don't know or don't care that non-IE 5/Windows users can't view their pages correctly, or are spiteful bastards who want to force *you*, the internet browser, to switch browsers and possibly even platforms, to view the page as the author intends it to be seen.

I think there is an important distinction here: Caspian is complaining about HTML authors who use these awful ''extensions'' to HTML. Just because MS provides them doesn't mean that all authors should use them.

(Re: the backslash deal: MacOS uses : as a directory delimiter (when it uses one at all). Should the colon also be valid in HTML? What if I write an OS to use a plus sign as my directory delimiter? Now what should authors do?)

[ Parent ]

"It sounds like a reasonable, level-headed po (3.50 / 6) (#11)
by PrettyBoyTim on Thu Jul 20, 2000 at 02:54:19 AM EST

Looking at the Rules of TWU, it looks like you've actually made use of your service platform-dependent.

it is PROHIBITED to host browser-specific and/or OS-specific sites on TWU. Sites which look "ugly" as viewed by non-Windows/non-Mac OS/non-IE users are alright; however, sites that do not function properly (e.g. links break, pages render as blank screens, etc.) on the sysadmin's browser (Netscape Communicator 4.x for Linux) and, upon testing, function properly under IE for Windows will be nuked for being OS/browser-specific.

Sites which look "ugly" outside of Windows/IE are acceptable. Sites which are unusable in any way (e.g. a link will not work, a page consistently crashes the browser, a page displays as nothing but the background) outside of Windows/IE are not. If you are using TWU, you have a responsibility to make your site accessible to users of all OSes and all major browsers-- not merely Microsoft products.

By insisting that all pages work on your browser, you are effectively foring your users to use Linux to be able to test them - hardly platform-independence, is it?

The other aspect of it is this: People wish to have control of their own sites. All of your other rules are to very reasonable, and have to do with bandwidth usage, not spamming etc... You have suddenly forced users to bow to your browsing habits. The fact is, if you needed to look at any of the sites you host (because of legal action or something), then you could.

Now, I appreciate that you are funding this out of your own pocket (it's a very noble project), but the people who joined your service would have been under the impressions that they had 'editorial freedom' over their sites. By forcing their hand, you have taken that away from them.

Re: "It sounds like a reasonable, level-heade (4.00 / 1) (#14)
by Caspian on Thu Jul 20, 2000 at 03:21:54 AM EST

Poppycock. I don't run Windows OR IE for my own use, but I DO test all of the sites I create in IE for Windows. How, you ask? Simple. I run Windows in an emulator.

I'm not demanding that ANYONE change OSes. I AM, however, demanding that if my users want me to subsidize their sites out of my own pocket-money, that that same money NOT go towards helping Gates along the road to a MS-only Web. I consider this quite fair.
Hacking code for the people!
[ Parent ]
Re: "It sounds like a reasonable, level-heade (3.50 / 2) (#15)
by PrettyBoyTim on Thu Jul 20, 2000 at 03:35:41 AM EST

I would say that setting up Linux on an emulator is considerably harder than writing HTML, and may be out of reach for many users. Your stated aim is to allow students to have their own webspace, and yet students may well not have their own computers to run Linux on. If they're using university computers, or computers in an internet cafe, they may have no choice but to use Windows.

As for your second point, fair enough. But you shouldn't be suprised when those who already had accounts on your system before you made the change to the user agreement are annoyed to find out that their webpages now have to conform to your political views.

[ Parent ]
Re: "It sounds like a reasonable, level-heade (3.00 / 1) (#16)
by Caspian on Thu Jul 20, 2000 at 03:47:28 AM EST

I will reply by saying two things:

1: If it works in Netscape for Windows or Mac OS, chances are 99.9% that it will work in Netscape for other platforms as well.

2: I see no reason why anyone-- be it a private individual such as myself or a publically held company-- should act as MS's lackeys unless they'd paid to do so. (I wouldn't do so even if I WAS paid, but that's me.) As for my "political views", I beg your pardon if I dislike the idea of the entire Web eventually becoming completely dependent upon a single company-- a possibility which is rapidly becoming an inevitability.

Hacking code for the people!
[ Parent ]
Re: "It sounds like a reasonable, level-heade (none / 0) (#45)
by PrettyBoyTim on Thu Jul 20, 2000 at 01:08:24 PM EST

1) Um okay. fair enough

2) Creating a webpage that happens only to work on IE is hardly 'acting as a Microsoft's lackey'...

I suppose the thing that strikes me as odd about it is it seems similar to putting up a condition that says "All pages must be written in English", or something like that...

[ Parent ]
Re: "It sounds like a reasonable, level-heade (none / 0) (#53)
by Anonymous Hero on Thu Jul 20, 2000 at 04:49:23 PM EST

Perhaps you'd prefer a simpler test.
There is one. Does it validate?

While validation won't absolutely guarantee that everything will be right, it's a big step in that directiom.

I applaud the action taken, to eliminate browser/OS-specific pages. I use Windows. I even use IE (though not by choice). And guess what? I still don't like seeing IE or NS only type pages. The greatest line I've seen was...

This page optimized for [ANY DAMN BROWSER].

[ Parent ]
Re: "It sounds like a reasonable, level-heade (none / 0) (#56)
by PrettyBoyTim on Thu Jul 20, 2000 at 05:57:33 PM EST

This page optimized for [ANY DAMN BROWSER].

Amen to that!

[ Parent ]

Re: "It sounds like a reasonable, level-heade (4.00 / 2) (#20)
by squigly on Thu Jul 20, 2000 at 07:02:14 AM EST

Couldn't this be made a little less aggressive. Something like
"For compatibility, Users should only use forward slashes in urls.
Use of non-standard image types is discouraged for reasons of bandwidth and comaptibility. Recommended file types are PNG, JPG and GIF.
All users, particularly Javascript users are strongly encouraged to test their websites with at least one other browser."

An outright ban on BMP's for inline images is not a good idea. It sets a bad precedent (And I'm not sure whether the html specification says anything about valid types)

The policy in itself is not too bad, but far too forceful. It might be helpful to offer assistance in making these sites compatible.

People who sig other people have nothing intelligent to say for themselves - anonimouse
[ Parent ]
Why your accusation is false (5.00 / 3) (#23)
by Anonymous Hero on Thu Jul 20, 2000 at 07:28:03 AM EST

Umm no. It's quite possible to write good standard HTML that every browser on the planet will be able to handle without testing your page in every browser. You may need access to multiple OSs and browsers to make sure your page looks GOOD in all of them, but you definately don't need that to make sure your page isn't broken. All this requires is a very basic understanding of HTML. HTML is NOT a layout language! If you must have total control of how your page renders, use PDF not HTML. These problems (assuming you aren't using BMPs and forward slashes) almost always originate from attempts to misuse HTML as if it were PDF!

See The Campaign for a Non-Browser Specific WWW for all the information that you need to use HTML properly.

This sort of crap doesn't only bother Linux users - I dual boot and spend at least half my time in windows, largely because of the lack of any usable graphical browser in Linux (NutScrape is an abomination, and Mozilla doubly so) and yet I run into this all the time myself, using Opera (and sometimes Lynx) on Windows! In most cases I take a moment to view the page source when I hit a page that is broken, and in every single case I have seen so far the problem has been immediately apparent, and involved someone that was misusing HTML, that either didn't understand what sort of language HTML is or didn't care. Hell, before I deleted IE I used to use it to load these sites, and half the time they were still if not totally broken at least for all practical purposes unusable in IE because I wasn't running the same resolution as the idiot that made the page was! HTML is NOT a layout language! It is a logical language that requires you to format your document LOGICALLY so that the browser on the viewers system can then make the layout decisions! Once you understand this simple fact, you can make proper Web pages on any platform you desire, and be confident they will be viewable by all.

The sad truth is that the people that make these travesties and place them on the web are often just well meaning but ignorant. Placing or linking to some good information on how to use HTML properly might help those. Another sad point of course is that very often these people are so hopeless they are using a so-called WYSIWIG HTML editor (which is a contradiction in terms) and never even look at the garbage they are posting. Again, education is the only realistic solution.

On this subject, Wired is now totally broken (it used to just be ugly) in Opera. Wired, of all sites, should be expected to do better.

[ Parent ]
such coincidence (3.00 / 1) (#18)
by marks on Thu Jul 20, 2000 at 05:01:31 AM EST

This article is very well-timed. A couple of days ago I mailed the 'Stichting Trudo' (they rent houses in Eindhoven, The Netherlands), because I'd very much like to view their page, which shows the houses that are for rent this week, but I'm only able to do so on someone else's computer with window$ and IE. I mailed both to the company's email address and to the creator of their pages, but as of yet, I haven't had a reply. I'm starting to think they'll just ignore the mail.

As can be seen in the html, their pages are created using Adobe GoLive. I've never used this program, but I always thought Adobe made some nice programs (I certainly enjoyed photoshop before my Linux days), but it seems that GoLive produces IE-only pages. This is indeed a Bad Thing (tm), but I don't know what to do about it (besides sending another email, quoting this article).

Re: such coincidence (none / 0) (#22)
by xypher on Thu Jul 20, 2000 at 07:23:50 AM EST

Actually GoLive itself doesn't produce IE only pages, but it does have the potential to do so, I've produced quite a few cross-browser pages with GoLive myself. It would be nice if some HTML editors had browser check plugins available however, so they could warn you what did and didn't work in other browsers/platforms.

[ Parent ]
Earlier versions of browsers (5.00 / 1) (#19)
by Cat Mara on Thu Jul 20, 2000 at 06:47:16 AM EST

Many IE for Windows users are completely unaware of what their pages look like in other browsers, other operating systems, or even earlier versions of their own browser!

This strikes a chord with me. As the sole user who still has a copy of IE4 on my floor, I'm continually bothered by the Web heads to act as a guinea pig for their latest and greatest FrontPage-generated atrocity. More often than not, their masterpiece collapses in a slew of JavaScript errors and badly-positioned content.

Forget the users of other OSes for minute, how can Microsoft do this to their own customers? This reminds me of the cynical tricks they pull with the Office file formats every two years or so to force users to upgrade. Clearly they have no qualms about making their users' surfing experiences miserable ones, all in the name of "innovation".

Cat Mara
Love me, I'm a liberal!

i use IE 4 at work (none / 0) (#21)
by Anonymous Hero on Thu Jul 20, 2000 at 07:05:43 AM EST

Should the ad in the upper corner be blank or is kuro5hin just "broken" OS(browser) specific page.

Not that i mind the ads missing ;)

nRe: i use IE 4 at work (none / 0) (#31)
by Inoshiro on Thu Jul 20, 2000 at 09:36:34 AM EST

K5 currently hasn't had its Scoop engine checked over for HTML correctness by anyone. Rusty is hardly the master of pedantic standards following like I am, so I may take a stab at it sometime so that you, too, can enjoy the full bredth of the ads ;)

Of course, you could just use Junkbuster and avoid ads everywhere :)

[ イノシロ ]
[ Parent ]
Netscape sucks. Get over it. (1.00 / 1) (#24)
by AndrewH on Thu Jul 20, 2000 at 07:32:02 AM EST

Netscape’s shoddy excuse for a web browser scarcely represents a legitimate standard to base web pages on, unless you really beleive that Marc Andreesen deserves to set Web standards. Personally, although i regard open standards as a matter of honour when working on my corporate network, I will not put in Netscape workrounds ther, and I put as few as possible on the open Web.

All too often, “cross browser compatibility” means imposing a Netscape Tax on the majority whose browsers conform far better to W3C standards, particularly on users of minority browsers that the authors haven’t tested the page on.

Go get yourself a decent browser. If you use Linux, have a look at Konqueror. Galeon may show some promise as well. If you don’t mind paying, there’s always Opera.

John Wilkes Booth, Lee Harvey Oswald, John Hinckley Jr — where are you now that we need you?
Re: Netscape sucks. Get over it. (none / 0) (#25)
by Caspian on Thu Jul 20, 2000 at 08:37:51 AM EST

That was really a nice touch. You know I run Netscape for Linux-- and you seem to have included some of the very same broken code in your response that my browser can't view. Check out this screenshot.

Anyhow, isn't the Web all about NOT HAVING TO USE ANY ONE BROWSER? Why should I have to change browsers just to use it? Any modern browser should be consciously and conscientiously supported by any decent Web site designer.

And by the way, Opera is NOT yet available for Linux. It's in a "technology preview"-- a highly unstable beta. When I tried to view that test page that I made for the article in it, IT HUNG.
Hacking code for the people!
[ Parent ]
Re: Netscape sucks. Get over it. (2.67 / 3) (#32)
by Inoshiro on Thu Jul 20, 2000 at 09:43:01 AM EST

I agree that being browser independant is good, but I wouldn't force your users into not having webspace because it crashed Netscape v4. Netscape v4 is a classic example of software which can't handle broken input in any sane way. BUT it also can't handle standards compliant pages in a sane way, either. The Day the Browser Died by Jeffery Zeldman illustrates quite nicely how Netscape v4 is a fundamentally broken browser.

No, I'm not an IE booster. I use Netscape v4 myself most often (as well as Opera and Lynx -- I refuse to let IE code reside on any of my computers) . I just happen to know its limitations better than you seem to.

On a side note, those are valid HTML v4 entities he has. It is only because Rusty has not thought to include an HTML v4 DTD at the top of all of the pages that Netscape does not render them. Please consult w3c standards on such matters before assuming that the other person has done something bad. You might reconsider making it a rule that all users' pages must have a DTD, and must verify reasonably well with the validators available on the web.

[ イノシロ ]
[ Parent ]
Inoshiro's side note is correct (none / 0) (#66)
by mattm on Sun Jul 23, 2000 at 10:40:15 AM EST

The curly quotes and such which AndrewH used are completely legitimate HTML 4.0; I hope it isn't too much trouble to modify scoop to generate correct DTD's. Would it be preferable if AndrewH had used MS-Windows-specific non-standard numerical ASCII codes &145;like these&146;, which on other systems generally appear either as question marks or as nothing at all?

And incidentally, lynx renders the “curly quotes” just fine. :)

[ Parent ]
Opera & Linux (none / 0) (#52)
by PrettyBoyTim on Thu Jul 20, 2000 at 04:36:33 PM EST

Unfortunately, Your experience with Opera is not unique to Linux 'technology previews' I've had quite a few crashes with v4 of Opera on Windows... *sigh* - it's a damn shame - I really like Opera as a browser, but it does hang *far* too often...


[ Parent ]
NuBus? PCI? (1.00 / 1) (#27)
by Anonymous Hero on Thu Jul 20, 2000 at 09:17:46 AM EST

Jackass. NuBus was developed by MIT. Apple adopted it in the spirit of open architecture. PCI was developed by Intel and marketed by Intel in the hopes of making them lots of money, which it has. You're as bad as those idiots at Slashdot.

Re: NuBus? PCI? (none / 0) (#29)
by Caspian on Thu Jul 20, 2000 at 09:30:21 AM EST

So what? The same PCI standards are used in Power Macs, x86 boxen, Alphas and, I would imagine, other architectures as well. Now find me an architecture other than 68K-based Macintoshes that uses NuBus. I didn't say that NuBus was "proprietary", you git. I said that it is not "platform-independent"-- for it isn't. NuBus is available on only ONE platform.
Hacking code for the people!
[ Parent ]
Re: NuBus? PCI? (none / 0) (#58)
by Anonymous Hero on Fri Jul 21, 2000 at 12:06:36 AM EST

NUBUS was available on other platforms, but the only one I can think of right now is the Merlin III.

[ Parent ]
It's tough (3.80 / 5) (#33)
by Inoshiro on Thu Jul 20, 2000 at 09:47:15 AM EST

Being a standards compliant person on a web gone mad. My homepage, for example. I recently redid it (I dreamed up a rather nice layout somewhat like fluffy grue's homepage), and was pleased. I'd verified it with the HTML 4 and CSS 1 parsers on validator.w3.org, and went to show it to some people.

The problem was that a suprisingly large number of browsers can't even render tables correctly (as I did then). Netscape 4 did it, Opera v3 padded the table, Opera v4 did it (but it didn't render the title="" attribute of the HTML 4 anchor tag), IE and Mozilla overflowed their tables. By overflow, I mean they somehow computed the width as a negative value, leading to an overflow such that it shows a page 65,536 pixels (aprox) wide.

I managed to get the attention of one of the fellows who contributes to Mozilla's development (outside of Netscape), and he submitted a patch to fix it (and a patch to fix Mozilla not handling the title attribute). So right now, the only browser which isn't standards compliant enough to render my page is IE (and things like Net+ which don't support CSS).

With the current semi-horrid implementation of documented standards, it's no wonder people resort to kludges to make things look right. That fellow's page you try to render has NO DTD.  Which is horrible, because Netscape assumes HTML 2. Leaving out a specifier like that leaves room to misinterpretation. Assuming things will be one way always is so wrong. Many programs written for Windows (even MS ones) improperly assume which colours are being used (hardcoded values). Something similar happens on the web. Set your browser background colour to a non-default (I like the gunmetal gray Netscape used until v4 of their browser as a default myself). Then browse around. All the pages by FrontPage hacks and another mental rejects tend to have images which white all around the edges, and look really tacky.

I've mentioned before that Dr. Jakob Nielsen did some research into browser usage patterns that could present a way to avoid the problems of incompatible HTML. All we have to do is have a standards compliant 6th generation browser for the 1st of January, 2001. We have a headstart on MS because Gecko is here today with the source open to people who can help fix it and get it out the door. We can beat the marketters at their own game of embrace, extend, extinguish.

All we have to do is have the Gecko engine working, and provide a nice wrapper to it (such as the Galeon Gnome wrapper for Gecko). The only catch is that you also need to have a Windows version, or you can bank on MS being able to force people into using IE 6.0.

[ イノシロ ]
Big deal (3.00 / 1) (#36)
by Jon Peterson on Thu Jul 20, 2000 at 10:36:00 AM EST

In 1993 I was trying to access the Internet, and do you know, hardly any of the services available worked on my computer (A Mac classic). The guys at the University IT helpcenter

suggested that I change OSes.


Shocking. It was like I had to use an RS6000 station (poor me :-) to use the Internet.

Sure, its sad but not that sad.

Shorter essay on the same topic (none / 0) (#37)
by Caspian on Thu Jul 20, 2000 at 10:58:31 AM EST

If you'll go here, you can read a shorter counterpart to this essay. Short, sweet, to the point and written entirely from scratch.
Hacking code for the people!
Facism? (2.50 / 2) (#38)
by Anonymous Hero on Thu Jul 20, 2000 at 11:00:34 AM EST

Who's the facist here? Have a look at these two hypothetical conversations:

Person 1: You're pages must work with all browsers and not just IE.

Person 2: I refuse to support non IE browsers.

next conversation -

Person 1: This job you're advertising must be available to blacks, asians, hispanics etc, and not just whites.

Person 2: I refuse to hire non-white employees.

Two very similar conversations addressing different issues, but which person is the facist?

One world, one people, one program. - Bill Gates
Ein Reich, ein volk, ein fuhrer. - Adolf Hitler

Yikes (4.00 / 3) (#39)
by ryry on Thu Jul 20, 2000 at 11:40:35 AM EST

First off, I'm going to say that you're writing about a very wonderful subject - something I think everyone should know about. But you sully your article with its verbosity (like I'm one to complain, but that's another story :-). Also, it's fine if it was meant as an editorial but looking at the issue objectively is an important thing too, and your anti-MS feelings come out strongly and make that impossible.

And taking a look at what's posted of your user policy, I believe that if you are going to force users to not use IE-only tags, or whatever, you should give them some alternatives or some information, maybe point them to some tips on equalizing a web page among browsers. Just some tips, take them as you want :-)

Otherwise, though, a more compact and to-the-point article would be awesome.

--too lazy for a .sig--
Possible Solutions (4.00 / 1) (#40)
by baka_boy on Thu Jul 20, 2000 at 11:48:54 AM EST

I know that this thread is directed more towards discussing the fairness/approriateness of the policy restriction IE-only sites, but I just wanted to weigh in with one comment about the technology this all hinges on.

Yes, it is true that IE supports a number of tags that other browsers don't. However, software has one wonderful property that seems to be overlooked all too often: what it taketh away, it giveth. We have one advantage against MS here that we don't in most areas: since HTML source is always visible, and people have to be able to write pages for IE, the 'proprietary' tags that they choose to use can be identified and neutralized.

So the Redmond boyas have slipped inline BMP images and backwards slashes? Why, someone might just have to write a lil' Perl/Python/Tcl script that my favorite lightweight local proxy (Junkbuster, etc.) could run over every incoming HTML page, checking it for such nasties and lifting them before my browser chokes. Take the offensive, all you tech-savvies, and stop feeling victimized.

Openness is nice... (none / 0) (#70)
by Anonymous Hero on Sun Jul 23, 2000 at 08:17:49 PM EST

And HTML has to be, right?

So what about Java apps?

It's made learning Java a real bear, I can't look at functional designs and lift/modify to my heart's content.

-- Ender, Duke_of_URL

[ Parent ]
Not new, but very very true (none / 0) (#42)
by Anonymous Hero on Thu Jul 20, 2000 at 12:26:51 PM EST

Have to say I agree with almost everything here (although for some reason even I haven't worked out, I'm still pretty fond of Netscape - must be a nostalgia thing). Just remember that most people who produce horrible HTML do so out of ignorance rather than malice - even though I now try to write clean HTML, my site probably still has some yucky bits from the days when I thought Netscape Composer was a cool way to write for the web or even - gasp - used MS Word's "Save as HTML" function. Let's face it, most web-page authors these days don't want to have to learn HTML, so perhaps the only alternatives are to encourage them to use something like Amaya, or if they simply have to use Front Page, persuade them to buy the full version, which allows version control (I am told).

Re: Not new, but very very true (none / 0) (#48)
by Anonymous Hero on Thu Jul 20, 2000 at 02:31:13 PM EST

It's laziness and ignorance too though, you can't blame it all on brain dead composer programs. No matter what program you are using (I use Netscape Composer which is as bad as anything you'll find for spitting out crappy html) you really must suck the output into notepad/emacs/whatever and do a little quality control before you post it to the web. It's that simple. The shame is on the people that don't do this out of arrogance instead of ignorance.

[ Parent ]
Netscape platform independent? (4.00 / 3) (#43)
by Anonymous Hero on Thu Jul 20, 2000 at 12:28:36 PM EST

Netscape's browser has a history of breaking standards that is as extensive as Microsoft's.

IE is better engineered, faster, has a better user interface, and is progressing at a much faster rate that Netscape's slow, clunky, bug-addled browser, despite its flaws. Furthermore, Microsoft has led the implementation of client side XML.

Netscape has also produced a range of mediocre products and technologies including: server side javascript, netscape enterprise server, and their ldap server.

I don't see any argument against Microsoft's actions here. Indeed, many of the original posters' non-displaying pages can probably be traced back to IE's -superior- handling of brain damaged html, which Navigator just gives up and dies on, with a blank page.

Re: Netscape platform independent? (3.50 / 2) (#47)
by KindBud on Thu Jul 20, 2000 at 02:22:40 PM EST

I use IE on the PC. Navigator on the Sun. I have used IE for Solaris, and it renders windows-specific pages very nicely, and fast - faster than a contemporary Netscape equivalent (4.7x is much better now, however). Oh, and it renders ordinary pages fast, too. :)

Unfortunately, IE4 for Solaris did not support Java at all, and IE5 for Solaris supported Java, but with the Sun jdk, so applets could not appear inside the window, but always ran external to the browser. It was also rather eerie to see an application lib directory littered with strangely familiar as well as alien names: libwsock32.so, libriched32.so, libole32.so, libctl3d.so ...

One thing that really bugs me, however, are sites that require Javascript to be functional. It really irks me when I see
(or however you do it) instead of a simple href tag. A lot of pages simply do not render at all if you block javascript.

just roll a fatty

[ Parent ]

Javascript only pages, like Sony? (none / 0) (#69)
by Anonymous Hero on Sun Jul 23, 2000 at 08:14:54 PM EST

You mean like Sony.com?

They're on my blacklist for that. They don't even have something telling you that you need javascript, and webmaster@sony.com dies. And I'm not spending the time to tell them they're fucked in the head.

I usually just view the code, then plug in the page title.

-- Ender, Duke_of_URL

[ Parent ]
Brain Damaged HTML... (5.00 / 1) (#50)
by MrEd on Thu Jul 20, 2000 at 03:02:57 PM EST

... should be given up on, and have a blank page returned. That way the brain-damaged web designer would know it was broken and fix it.

Watch out for the k5 superiority complex!

[ Parent ]
Lynx support? (none / 0) (#44)
by aphrael on Thu Jul 20, 2000 at 01:05:46 PM EST

I stopped being livid about this a few years ago, but it was infuriating for a while, and now it's funny: how hard is it to put in ALT tags behind images? Or provide text-only versions of sites that people who are still using 9600 connections can use without downloading the binaries? Even those sites which are fully functional under lynx (say, for example, the New York Times, or /.) are annoying to use with it.

Windows POSIX (5.00 / 1) (#46)
by KindBud on Thu Jul 20, 2000 at 02:07:16 PM EST

(Witness Windows's pitiful shred of POSIX-compliance, placed in almost as an afterthought to prevent the software titan from losing valuable government clients-- clients that demanded POSIX features.)

Ironically, the agencies that lobbied for POSIX support did so merely to get GAO approval for using Windows NT. Once the contract was done and the product shipped, the POSIX DLL was ignored, and applications were written directly to Win32, not POSIX. It was an end-run around government standards for government agencies, by government agencies. They don't even follow their own rules (but you knew that :).

just roll a fatty

Neat trick.... (none / 0) (#49)
by KindBud on Thu Jul 20, 2000 at 02:33:50 PM EST

For a Windows-only, Microsoft-enhanced version of this post, click here.

OK, so who is still on-line after that? :)

just roll a fatty

I've seen those on /., etc... (none / 0) (#68)
by Anonymous Hero on Sun Jul 23, 2000 at 08:11:11 PM EST

But I haven't been on a windows machine to check it until now.

A nice fatal exception error, but I was still online (modem working). Relaunching Net$crape afterward was taking too long, so I tried to kill the process... But things were acting up, so I had to reboot.

So yes, windows machines can still be online after that :)

-- Ender, Duke_of_URL

[ Parent ]
OT: NUBUS (4.00 / 1) (#51)
by Anonymous Hero on Thu Jul 20, 2000 at 03:54:12 PM EST

Apple's now-defunct NuBus technology was tied to Macintosh computers; thus, NuBus was not a platform-independent standard, but PCI is.

I have a few points* about that:

Apple didn't create NUBUS; MIT did.

The NUBUS BUS was actually an open'ish design.

NUBUS and PCI were not direct compititors (nubus is around a decade older than PCI.)

PCI Wasn't that useful until the v2.0 standard was formalised in April 1993 (less than a year after v1.0).

Apple announced it would support PCI almost as soon as v2.0 was announced (the reason the first Power Macs were NUBUS based is because they were in development prior to the V2.0 standard being formalised and the fact that customers actually like having a little backwards compatability.)

NUBUS didn't die because it was proprietary and PCI wasn't; it died because it was old.

* Please note I do actually prefer open technology standards. The reason I'm posted this is because I just happen to have a little knowledge of this area.

[This rant brought to you by POPSI: the taste of the GNU generation]

The sad part of all this (3.00 / 2) (#54)
by Anonymous Hero on Thu Jul 20, 2000 at 04:58:38 PM EST

...is that, true as Caspians points are, he is a total hypocrite about it. Check his earlier article, read the post where someone pointed out the problems with his page, and his vitriolic, and hypocritical first and second replies. Caspian doesn't write his page for any browser, he writes it for 3 browsers (IE, NS, and Lynx) and positively went balistic when a poster questioned this practice.

Caspian, You need to realise that your practice of writing for 3 browsers is different only by a matter of degree from the practice of many who write for 1 browser. Forget about the browsers - write proper HTML (which means not just formal compliance but also following good coding practices by following the spirit of the standard - that your pages should be designed to degrade gracefully, and that HTML is a logical language, not a layout language, in particular) and let the browser do its job, which is to make the layout and rendering decisions.

Your hypocritical stance doesn't render your article untrue, but it does render you a very poor person to be advocating platform-independence. Please, pull the splinter out of your eye, and THEN have another go at the logs the IE only crowd are suffering from.

Re: The sad part of all this (4.00 / 1) (#59)
by Caspian on Fri Jul 21, 2000 at 12:39:25 AM EST

You apparently didn't read my points very closely. What's important isn't standards. For all of IE's "standards compliance", what it has become is merely a way of designing sites that only work on Microsoft's software. WHAT IS IMPORTANT IS "DOES THIS WORK IN MY BROWSER?" I design for browsers for all platforms and of all types (text or graphical).I don't give a flying flip what the W3C's validator has to say about my site, AND I WON'T UNTIL ALL MAJOR BROWSERS ADHERE TO THE W3C GUIDELINES 100%.
Hacking code for the people!
[ Parent ]
Re: The sad part of all this (5.00 / 2) (#60)
by Anonymous Hero on Fri Jul 21, 2000 at 04:02:00 AM EST

Caspian, I want to first thank you for your reply (although I find it completely unacceptable, the fact that we can discuss this still leaves hope and is a good thing) and to remind you that we are on the same side. I was defending you a little earlier in a thread a little lower down the page. (I guess I really should create an account - soon.) But your reasoning is in some areas completely wrong headed and yes hypocritical, and that does make you a very poor advocate for platform independence until you get straightened out, and for that reason I feel compelled to do my best to do some straightening. Because Platform Independence is a very Good Thing (TM) that I have been fighting and educating in favour of for years now, and as much of a force for good as you could be in it, you can also be just as much a force for bad by making your case in such a flawed way. Which I guess is my long winded way of saying "I hope you understand this is nothing personal." And excuse me if I rant. But this is critically important stuff.

You apparently didn't read my points very closely.What's important isn't standards.

On this you are 100% completely wrong. Standards are the only viable route to platform independence. Think about this a moment. Platform Independence doesn't just mean that something will work on Windows, Linux, and Mac. It doesn't just mean that you can use Netscape, Lynx, or IE. Those three OSs and 3 browsers may make up the bulk of what's out there at the moment, but "supporting the majority of systems" is NOT the same thing as "supporting all systems." If you can use Platform Independence to mean working with the three browsers and three operating systems with the most users, why can the IE only folks not use it to mean supporting IE on all versions of Windows? It's only a matter of degree, and the fact is that IE on Windows DOES support the majority of users. I bring this up because this is the argument you can count on hearing from IE only supporters, and you are opening yourself up to it by accepting a mild variation on it yourself. If supporting most users is good enough, then there is no need to support Mac, Linux, Netscape, or Lynx - IE on Windows IS the majority of the market currently.

Platform Independence isn't just supporting most users, but ALL users. It isn't just supporting 1, or 3 Operating Systems, it isn't just supporting 1, or 3, Browsers, it's supporting any browser. When you yourself are happy with supporting 3 browsers and 3 OSs and with them perhaps 90% of users, you yourself are constructing the slippery slope that IE only proponents can and will happily slide down to get to their 1 OS 1 browser 80% of users position.

But how on earth could anyone possibly support every OS and every browser out there? How many people have the resources to test their page, not just on IE, Netscape, and Lynx, but also on Arachne and Opera and Emacs-W3 and kfm and a dozen other browsers neither you nor I have probably ever heard of? Not just on Windows, Mac, and Linux, but on Solaris and Amiga and Vax and OS/2 and Tru64 and BeOS and NeXT and all the BSDs and FreeDOS and DR-DOS and even yes MS-DOS and freaking CP/M for that matter? Sounds like an impossible task, even for a rich man with tons of time and money, let alone the average joe, right?

If this is an impossible task, then you better buy yourself a Windows Box with IE right now and quit your whining.

But, the fact is, this is not an impossible task. Things like PPP and TCP/IP and ASCII text do exist, and they are truly Platform Independent. Every OS I mentioned, and more you or I have never heard of, can use PPP, can use TCP/IP, can use ASCII. How is such an "impossible" feat accomplished? One word. Standards.

You apparently didn't read my points very closely.What's important isn't standards.

No, I read your points very closely, I just happen to disagree with them. Standards are precisely what is important, if Platform Independence is the goal, standards are the most important thing there is.


So if you used IE on Windows, you wouldn't care about Platform Independence anymore and would be perfectly happy with an IE only web, right? After all, what's important is "DOES THIS WORK IN MY BROWSER?" - so the only reason you can give why an IE only web is a bad idea is because you, personally, do not use IE. Well, guess what? All those users writing shoddy pages that you can't get to render, they ARE using IE, and what's important to them is "DOES THIS WORK IN MY BROWSER?" By your own criteria, they are right to build their pages without a care in the world what the resulting site looks like in Nutscrape for Linux, because they aren't using it! Most people use IE, and if the important thing is "DOES THIS WORK IN MY BROWSER?" then for most people an all IE web is the way to go.

I design for browsers for all platforms and of all types (text or graphical).

But the fact is you don't. You've confirmed this in earlier posts. You design for the platforms and browsers that you consider important - Windows, Mac, Linux, IE, Netscape, and Lynx. I'm sure you are sincere in thinking that this is the same thing, but think about that! It's not true! Those platforms account for the majority of users, certainly but Windows/IE alone accounts for a strong majority of users WITHOUT the others! This is why I say your position differs only in degree from the IE only designers.

Writing for one, or for three, platforms can be done with the experimental method, and both you and the IE only writers rely on that method. To judge a pages worth, the IE only crowd just pulls it up in IE. You, being a more sophisticated person, pull it up in Netscape, and in Lynx, and in IE. But the procedure, the values, the assumptions involved are essentially the same.

Writing for ALL platforms, however, simply cannot be done the same way. NO ONE can pull up their page in every browser on every OS that is in use, and that will be in use in the future, to test a page. It's impossible.

But it's not impossible to write for ALL those platforms and browsers. Standards are what make this possible, and that's why standards are so critically important. Write a web page in standard HTML, observing the spirit as well as the letter of the standard, and you are writing for all platforms. A page written that way is accessible to any platform (within the limits of hardware and software available, it is true, but the fact that a text mode console doesn't support graphics does not make it incapable of supporting the web, as you surely know since you support Lynx, it's even possible to support HTML on hardware with no display at all, think about browsers for the blind - and the standard is precisely the mechanism that makes producing the browser software on the less popular platforms possible - it is far easier to write a browser based on the HTML 2.0 spec than by trying to reverse engineer and reproduce every bug in Netscape or every non-standard element IE supports.) It's accessible to the blind. It's accessible to the guy in Slovakia with DOS and Arachne. It's accessible to the blind guy with a voice browser. It's accessible to systems you have never seen or heard of. It's accessible to systems that haven't been invented yet. That's Platform Independence.

I don't give a flying flip what the W3C's validator has to say about my site, AND I WON'T UNTIL ALL MAJOR BROWSERS ADHERE TO THE W3C GUIDELINES 100%.

And that is, as I hope you might be beginning to realise, a fatal mistake. As long as you hold that attitude, you are an enemy of Platform Independence in just the same way that every Windows/IE only Helsinki Syndrome afflicted luser who is offended by your "fascist" policy against IE only sites is. They just happen to outnumber you.

What exactly is it that you think is wrong with being standards compliant? Please answer that.

Apologies for being so long winded. But this issue is so important, economy seems very unimportant in comparison.



[ Parent ]
Re: The sad part of all this (none / 0) (#63)
by Caspian on Fri Jul 21, 2000 at 10:13:15 AM EST

Hey-- I'd love to discuss the issues you brought up in this very long and well-thought-out post-- mostly since you did neglect to mention the only reason why I HAVEN'T coded to the HTML standard, and I'd love to hear what you have to say to that. Could you e-mail me at caspian@twu.net?
Hacking code for the people!
[ Parent ]
Re: The sad part of all this (none / 0) (#64)
by Arker on Fri Jul 21, 2000 at 11:39:38 AM EST

Could you e-mail me...?

Certainly. Probably shouldn't have posted the email unobfuscated (oops!) hope you don't get pounced on by spammers because of it. Mail is in the pipe now. Looking forward to continuing the discussion there.

[ Parent ]
Website ADA liability (5.00 / 1) (#55)
by Anonymous Hero on Thu Jul 20, 2000 at 05:31:16 PM EST

Accessible and browser-neutral web design may be more than just good sense, it might be the law depending on where you live. For instance, in the United States, ADA accessibility requirements apply to government and company web pages. If your employer's website does not render correctly for persons with disabilities, they could be fined, or even sued. Few of you may have heard of Avanti, Emacsspeak, or Marco Polo, but compatibility with those is far more important to the disabled community than cute bitmaps in either IE or Netscape.

For more information on internet usability issues for the disabled, visit the W3C's Web Accessibilty Initiative website at http://www.w3.org/WAI/

Features giveth, features taketh away (none / 0) (#57)
by shadowspar on Thu Jul 20, 2000 at 09:44:28 PM EST

I can't believe it took me so long to remember this. It's not really a terribly constructive way to design web pages, but those of you who hate browser-specific web pages simply must take a look at The HTML Terrorist's Handbook. It was originally at http://www.zikzak.net/~acb/hacks/htmlth.html, but the site has gone 404. (I have a copy, but no website or internet connection currently; however, try searching gnutella.)

It's an awesome treatise on the use of HTML as an offensive weapon. My favourite is the one about putting your whole page in NOFRAME tags, and suitable alternative text in the frames, so my fellow frameless peons see my page, but everyone else gets:

Your browser does not support NOFRAMES.
Please upgrade to a NOFRAMES compliant browser, such as Lynx or Mosaic.

Another goodie is the assumption that anyone of importance is running Windoze, and so has all the fonts for FONT FACE="Helvetica" (or whatever).

Of course, anyone who has Windows also has a number of other fonts.... FONT FACE="Wingdings"

Again, not terribly creative or mature, but great fun.
-- Drink Canada Dry! You might not succeed, but you'll have fun trying.

Re: Features giveth, features taketh away (5.00 / 1) (#62)
by Paul Dunne on Fri Jul 21, 2000 at 08:37:11 AM EST

Here it is. And a good read it is too -- I must try some of that!
[ Parent ]
Another dirty trick re \ vs / as url delimiters (none / 0) (#65)
by bee on Fri Jul 21, 2000 at 02:31:38 PM EST

Here's a way to maliciously use the \ vs / issue:


where foo is a directory containing bar.html which is some evil nasty malicious page, and foo\bar.html is the real page.

[ Parent ]
'foo\bar.html' and other ideas... (none / 0) (#67)
by Anonymous Hero on Sun Jul 23, 2000 at 07:47:45 PM EST

This is beautiful.

I think I'll just move my index page to a meta-refresh start\index.html

It does mess with my standard directories start with Capital letters, but I can let that go...

I've been using javascript to close down IE, but Redmond put a stop to that... it now asks the user. I was very happy with that, becuase I figured anyone who used javascript deserved what they got...

var uagent=navigator.userAgent;
if (uagent.indexOf("MSIE") == 25)

I like the noframes idea, but I do use a graphical browser, and I am stuck with Netscape -- I really wish there were an option to turn off frames, so I could surf like that - I've been desiring that for a damn long time. I finally had to give up running N1.1, but I still miss it's simplicity.

-- Ender, Duke_of_URL

[ Parent ]
Not too worried (none / 0) (#61)
by pdcruze on Fri Jul 21, 2000 at 06:32:13 AM EST

As a recent subscriber to K5, I missed your earlier article. No matter. I agree with most of your comments. Yes, it does bug me that I can't view some sites correctly using Netscape under Linux.

Having said that, there's something those Windows users don't realise. Us Linux users sleep a WHOLE lot better at night. We don't have to worry about the virus/security hole of the week like they do. We don't have to constantly face the threat of having all our data wiped. Our computers don't crash every few hours. In short, computing under Linux is a godsend. Its stress-free computing.

Personally, I think those few web sites I can't access properly under Netscape is a small price to pay for this piece of mind. So Windows users have access to a few more sites than I do. Well, they're also growing ulcers a hell of a lot quicker than I am. :-)

Pride and Prejudice: The Platform War Comes to the Web | 69 comments (62 topical, 7 editorial, 0 hidden)
Display: Sort:


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!