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]
Browser version at least 5, please.

By static in Op-Ed
Tue Jun 12, 2001 at 08:20:08 AM EST
Tags: Internet (all tags)
Internet

Amongst webmasters and web developers, it is widely known that if you want to do more than basic formatting with CSS (Cascading Style Sheets), then either forget about supporting Netscape 4, or plan to spend a lot of time creating compromises for it. But is Netscape 4 still used all that much?


Actually, Netscape 4 is still used by a not-insignificant number of die-hards. Netscape 3 is, in fact, still being used, as is IE 2 and IE 3! Browser Watch has a collection of statistics on Agent strings. The numbers, while admittedly probably quite slanted, still make for interesting reading.

But is CSS in Netscape 4 really so bad? Umm... well, a site called Web Review seem to think so. In fact, IE 3 (on the Mac, no less) is at least as buggy, and conveniently in different areas.

I've been building a web site based around Scoop. After digging into Scoop's setup, I decided it would be better if I grafted a semblance of CSS onto it, testing in Opera 5.11 as I went. The CSS I put together is actually very tame, but still Netscape 4 gives trouble. (This is only going to get worse when Scoop itself gets full CSS support in the not-too-distant future.)

So. Is it finally time to stop trying to support CSS in Netscape 4 or IE3 on the web in general?

Sponsors

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

Login

Related Links
o Scoop
o Browser Watch
o Web Review
o Also by static


Display: Sort:
Browser version at least 5, please. | 101 comments (101 topical, editorial, 0 hidden)
My honest opinion. (3.50 / 10) (#1)
by kwsNI on Tue Jun 12, 2001 at 02:08:59 AM EST

Browser upgrades are free. Stop worrying about how older browsers implement newer standards and write good, solid code that works given the current standards.

On the other end of the spectrum, if you're using an older browser, either put up with crappy rendering of web pages using newer standards or download a newer update for your browser. I hate it when I get e-mails from people telling me that some page I made doesn't work in IE2.

kwsNI
I can picture in my mind a world without war, a world without hate. And I can picture us attacking that world, because they'd never expect it. -Jack Handy

Browser upgrades are free. New computers aren't. (3.57 / 7) (#7)
by Estanislao Martínez on Tue Jun 12, 2001 at 02:40:07 AM EST

Are you really about to suggest that somebody running Win95 in a 24MB PC run Mozilla?

--em
[ Parent ]

Quite. (4.28 / 7) (#9)
by static on Tue Jun 12, 2001 at 02:59:17 AM EST

I had a relative start using the web with a 386DX40 running with 20MB of RAM. With Windows 95. I made a point of not installing NN or IE. I gave him Opera instead.

Wade.

[ Parent ]

Browser requirements (3.50 / 6) (#11)
by hal0802 on Tue Jun 12, 2001 at 03:12:45 AM EST

Are you really about to suggest that somebody running Win95 in a 24MB PC run Mozilla?
... Are you suggesting that you can run IE5 on the same PC ? According to this M$ joke, you can (didn't know booting W95 with 16Mb was possible BTW).
- You shall never find the necronomicon -


[ Parent ]
Yes, I am (3.00 / 3) (#19)
by ti dave on Tue Jun 12, 2001 at 03:53:55 AM EST

"... Are you suggesting that you can run IE5 on the same PC ?"

I set up my daughter's computer (win98/IE5.0/P120/24Mb RAM) for her to do her school work with, and it ran fine. I later "upgraded" it to a P166/32Mb RAM and FreeBSD. Ran much smoother.

ti_dave

"If you dial," Iran said, eyes open and watching, "for greater venom, then I'll dial the same."

[ Parent ]
why only 24 megs? (3.00 / 1) (#67)
by delmoi on Tue Jun 12, 2001 at 01:30:59 PM EST

The server that powers picture-rate.com is running an old p200.. but I've got like 198 megs of ram. It's so cheap, and windows 98 (in my experience) Is choppy and annoying even with 64 megs of ram. 128 is a good number for '98

(windows 2000 is actualy more memory efficent then 98, and can run OK with just 64 megs). But I mean, memory is so cheap now, why not get more for your daughter.
--
"'argumentation' is not a word, idiot." -- thelizman
[ Parent ]
Because that is... (3.00 / 1) (#72)
by ti dave on Tue Jun 12, 2001 at 04:07:11 PM EST

the sum of all the functional SIMMs I had when I assembled the box. Around these parts, used SIMMs cost more than new DIMMs. Builiding the box to those specs was a stop-gap measure until tax refund time.

I agree about Win 98 though, 96 Mb is pretty good overall performance for it, 128 would be the bare minimum I'd use to run Office in a semi-snappy manner.

Cheers,

ti_dave
"If you dial," Iran said, eyes open and watching, "for greater venom, then I'll dial the same."

[ Parent ]
Not Mozilla, but IE5 is fine ... (2.33 / 3) (#21)
by FantasticSam on Tue Jun 12, 2001 at 04:41:32 AM EST

I wouldn't recommend mozilla to anyone with a slow machine. IE5 is nice and fast though. Netscape or IE4 users especially will notice a big performance increase.

[ Parent ]
Win95 on 16MB? (3.40 / 5) (#29)
by AdamJ on Tue Jun 12, 2001 at 06:27:30 AM EST

Very possible. I used it with 8MB for at least a year. Pretty sure I ran IE4 for part of that time, too.

Now, things went much nicer with 32MB, and are certainly much nicer now with 384MB, but it's certainly possible - and dare I say it, at the time, it wasn't all that painful.

[ Parent ]

Win95 on 512MB ?! (3.60 / 5) (#44)
by hal0802 on Tue Jun 12, 2001 at 09:29:10 AM EST

Adding too much memory isn't a good thing (tm) either since windows doesn't handle it so well
oh my ...
- You shall never find the Necronomicon -

[ Parent ]
I did it in 4 (3.50 / 4) (#50)
by mauftarkie on Tue Jun 12, 2001 at 12:18:38 PM EST

When I first used Win95 on my 486/66, I only had 4MB. I wouldn't call it usable, because every time you right-clicked you had to wait 10-20 seconds for the menu to swap into memory (and then 10-20 seconds when you made your selection for it to swap back out), but it did work. Needless to say, I promptly upgraded to 20MB and had no further problems.

This was right after 95 came out; before IE was even out. I don't think I was ever able to run Netscape 1.0N (or any of the 0.9x series) on it, though. I was pretty much limited to Mosaic. Which was fine with me at the time.


--
Without you I'm one step closer to happiness without violence.
Without you I'm one step closer to innocence without consequence.


[ Parent ]
I ran it for years in 8. (3.00 / 1) (#75)
by static on Tue Jun 12, 2001 at 07:12:54 PM EST

It was a 486DX4/400 with 8Mb of RAM and a VLB video card. It was also the most stable Windows 95 install I've ever had. Never had to re-install it and boy did I do some horrible things to it! I would have kept using it except that I couldn't add more memory to it.

It made me suspect that Windows 95 (original) was optimised and tested most thoroughly for a VLB machine, perhaps because that's what they developed on. And that PCI support was therefore not tested properly.

Wade.

[ Parent ]

400 MHz? (2.00 / 1) (#80)
by kwsNI on Tue Jun 12, 2001 at 08:47:50 PM EST

That is one blazing fast 486.

kwsNI
I can picture in my mind a world without war, a world without hate. And I can picture us attacking that world, because they'd never expect it. -Jack Handy
[ Parent ]
D'oh! (2.00 / 1) (#83)
by static on Tue Jun 12, 2001 at 11:19:49 PM EST

I meant 100 MHz. Actually, it was 33MHz - 486DX4 CPUs had a 3x multiplier in-core.

Wade

[ Parent ]

windows 95 (4.00 / 3) (#66)
by delmoi on Tue Jun 12, 2001 at 01:28:03 PM EST

Windows 95 was actualy designed to run with just 8 megs of ram (remember, when it came out a 4mb simm was $200..) It could even boot with just 4 but you would run into some problems

I actualy saw a computer lab decked out with 386/16s with 4 megs of ram running 95. They would last about 5 minutes before bluescreening....
--
"'argumentation' is not a word, idiot." -- thelizman
[ Parent ]
No. (3.00 / 5) (#12)
by pwhysall on Tue Jun 12, 2001 at 03:17:51 AM EST

I'm about to suggest that someone running a 24MB PC either buy some more memory, ditch Windows for a more efficient OS, or upgrade.

I don't really think it's reasonable to expect to be able to run with a 5 or 6 year old PC and expect to fully enjoy the benefits of current software, which IS bigger because it does a lot more.


--
Peter
K5 Editors
I'm going to wager that the story keeps getting dumped because it is a steaming pile of badly formatted fool-meme.
CheeseBurgerBrown
[ Parent ]

No, but (3.33 / 3) (#65)
by delmoi on Tue Jun 12, 2001 at 01:26:00 PM EST

They could run IE.

Now, a 24meg linux box... then we have a problem...

Actualy, they shouldn't even be running 95 with just 24 megs of ram, but memory is so cheap nowadays, you can get 64 megs of ram for like $9 or something.
--
"'argumentation' is not a word, idiot." -- thelizman
[ Parent ]
Yes (2.38 / 13) (#2)
by delmoi on Tue Jun 12, 2001 at 02:10:33 AM EST

Why the hell should I spend hours of my time making my sites work for people who can't be bothered to spend 20 minutes or whatever to get a non-sucky browser?
--
"'argumentation' is not a word, idiot." -- thelizman
I have (3.66 / 6) (#20)
by i on Tue Jun 12, 2001 at 04:01:42 AM EST

SunOS 5.6, and the only non-sucky browser I can get in 20 minutes (or 20 days, for that matter) is Lynx. So please support Lynx.

and we have a contradicton according to our assumptions and the factor theorem

[ Parent ]
links (3.50 / 2) (#55)
by Nimey on Tue Jun 12, 2001 at 12:48:45 PM EST

And, by extension, links(1), which is $DEITY's own browser.
--
Never mind, it was just the dog cumming -- jandev
You Sir, are an Ignorant Motherfucker. -- Crawford
I am arguably too manic to do that. -- Crawford
I already fuck my mother -- trane
Nimey is right -- Blastard
i am in complete agreement with Nimey -- i am a pretty big deal

[ Parent ]
Lynx (3.00 / 1) (#62)
by delmoi on Tue Jun 12, 2001 at 01:17:24 PM EST

I'm not really sure that there's much of a point in a picture-rating site supporting Lynx :P But, if you want this should work
--
"'argumentation' is not a word, idiot." -- thelizman
[ Parent ]
Well, lets see... (3.00 / 4) (#22)
by squigly on Tue Jun 12, 2001 at 04:58:04 AM EST

Hours of time - presumably this would be less than 3 hours. You have to do this once. If 10 people have to spend 20 minutes downloading a new web browser, (plus the time to install, disconnect, reboot the OS, and find the page again), then its much more efficient for you to do it.

Of course, a well designed page shouldn't take any extra time. The very least you could do is force it to make all text black, and all backgrounds white as a fallback. Doesn't look pretty, but it will have the information there.

[ Parent ]

Still overdesigned (4.00 / 4) (#24)
by AndrewH on Tue Jun 12, 2001 at 05:10:23 AM EST

The very least you could do is force it to make all text black, and all backgrounds white as a fallback.
What is so wrong with the defaults set in the browser?
John Wilkes Booth, Lee Harvey Oswald, John Hinckley Jr — where are you now that we need you?
[ Parent ]
For picture-rate.com it would have (3.00 / 1) (#63)
by delmoi on Tue Jun 12, 2001 at 01:22:08 PM EST

I know ns4 would be capable of doing the little 'slider' thing on picture-rate.com which was my first 'web application' in a while. I don't know if I'll do another one though, it really sucked :P Getting it to work in IE and Mozilla was a pain, since they used slightly diffrent object models.

I do have a totaly non DHTML version that dosn't use anything other then a few very basic tags (hr, br, img, form, a), and making that didn't take long. But I really wouldn't wanted to have done the slider thing with NS4's broken-ass DHTML.
--
"'argumentation' is not a word, idiot." -- thelizman
[ Parent ]
Because (4.00 / 6) (#28)
by Vulch on Tue Jun 12, 2001 at 06:22:11 AM EST

You are the one that wants people to visit your site maybe? I don't care. If I can't see your site in a browser that doesn't support CSS or javascript then, unless you are the only place with the information I'm after, I'm not going to bother.

No matter whether your site is a personal one or a business site, if you want me to visit it then I expect you to make it easy for me to do so. A stockbroker lost my business last year because their HTML defaulted to using white on white. If you had CSS available the text was turned blue, but it wasn't visible on the browser on my handheld away from my desk. I suggested they change the defaults and got an answer that basically boiled down to "We can't be bothered because 90% of our customers use X". Turned out that wasn't quite right, 90% of their hits might have been using that, but they were from people just keeping an eye on portfolios. It was the 10% using other browsers were doing most of the trading, and going elsewhere to do business.

[ Parent ]

Hrm? (3.50 / 2) (#64)
by delmoi on Tue Jun 12, 2001 at 01:24:13 PM EST

We can't be bothered because 90% of our customers use X". Turned out that wasn't quite right, 90% of their hits might have been using that, but they were from people just keeping an eye on portfolios. It was the 10% using other browsers were doing most of the trading, and going elsewhere to do business.

How do you know this?

Anyway, my freetime is worth more to me then wether or not a small but vocal minority can few pages I put up for fun.
--
"'argumentation' is not a word, idiot." -- thelizman
[ Parent ]
Sources (3.00 / 2) (#88)
by Vulch on Wed Jun 13, 2001 at 05:29:05 AM EST

How do you know this?

A friend now works there in the IT department, and has commented on the change in traffic to the site since the changes were made. Their share of the trading market has declined much more than the market average, and I've recommended the broker I switched to using to several other people who had the same problems I did.

For a private set of pages it's your decision, but for a business site you'd better be damn sure that the visitors you lock out don't contain a high proportion of your potential customers.

[ Parent ]

Why be standards compliant? (4.00 / 2) (#86)
by epcraig on Wed Jun 13, 2001 at 12:49:03 AM EST

Because some dial-up users (me, for instance) tend to use text browsers for efficiency.
Because I can do without access to a site which won't cater to my preferred disabilaties.
Because it's not necessary for me to upgrade at your convenience just to see your site.
There is no EugeneFreeNet.org, there is an efn.org
[ Parent ]
Because.... (none / 0) (#93)
by Tezcatlipoca on Thu Jun 14, 2001 at 11:27:05 AM EST

... not everybody can choose its own browser.

I work in a big company where all the sysadmin have Sun machines only. The best alternative for a modern browser is Netscape, I wish I could make Mozilla, but does not work in Solaris 2.6

Others have pointed out blind people that require special browsers.

And surely one coud find many reason whay somebody can't install an specific browser.



Might is right
Freedom? Which freedom?
[ Parent ]
What Solaris are you using? (1.00 / 1) (#96)
by AndrewH on Fri Jun 15, 2001 at 05:23:31 AM EST

Obviously not the one that lets you choose between IE and Konqueror.

Others have pointed out blind people that require special browsers.

And how exactly do Netscape-friendly web pages help them?


John Wilkes Booth, Lee Harvey Oswald, John Hinckley Jr — where are you now that we need you?
[ Parent ]
Problems with IE and konqueror in Solaris. (none / 0) (#98)
by Tezcatlipoca on Fri Jun 15, 2001 at 09:47:53 AM EST

IE had serious conflicts with the graphic cards in our workstations (not to mention it is not the company's standard). We decided not to put something else like gnome or kde because all our users have to use CDE, so it has clear benefits for us to stick to that.

Now if you tell me I can run konqueror in Solaris 2.6 without installing all KDE then that could be explored. ;-)

Regards.

Might is right
Freedom? Which freedom?
[ Parent ]
HTML 1.0 (3.83 / 6) (#3)
by eLuddite on Tue Jun 12, 2001 at 02:13:46 AM EST

Is overkill. All you need is <p>.

---
God hates human rights.

Forget <p> (4.00 / 4) (#5)
by dram on Tue Jun 12, 2001 at 02:20:43 AM EST

All you need is <pre>

-dram
[grant.henninger.name]

[ Parent ]
Forget <pre> (4.00 / 3) (#25)
by Estanislao Martínez on Tue Jun 12, 2001 at 05:25:18 AM EST

Plain text files can offer the same fetures, with 11 bytes less per file. Cool, eh?

--em
[ Parent ]

10 Bytes (4.00 / 3) (#31)
by moshez on Tue Jun 12, 2001 at 07:13:12 AM EST

Compensate for the text/html -> text/plain

[T]he k5 troll HOWTO has been updated ... This update is dedicated to moshez, and other bitter anti-trolls.
[ Parent ]
True (2.80 / 5) (#33)
by dram on Tue Jun 12, 2001 at 07:18:55 AM EST

But then you get an ugly grey backgroud on most peoples browsers. I would sacrafice that 10 bytes for the white background. But it is a trade off between style and size, and I for one would much prefer style. Size just doesn't matter that much, at least thats what she said.

-dram
[grant.henninger.name]

[ Parent ]
actualy, no (3.00 / 1) (#68)
by delmoi on Tue Jun 12, 2001 at 01:42:56 PM EST

IE 5 puts a white background up by default. It uses "windows colors" meaning whatever you have selected as the defualt "window background" color. I personaly hate it, and set IE to display grey backgrounds the way they are *supposed* to be :P

Anyway, I like grey backgrounds.
--
"'argumentation' is not a word, idiot." -- thelizman
[ Parent ]
Graceful degradation (4.31 / 16) (#4)
by Estanislao Martínez on Tue Jun 12, 2001 at 02:15:32 AM EST

One of the design goals for HTML and CSS is to allow for "graceful degradation", i.e., the fact that a document uses CSS or other newer standards should not stop the document from displaying on browsers which can't support it, either because of age or other limitations (e.g. text-only browsers such as Lynx, or browsers designed for blind people).

I'd say make your site meaningful and fairly usable when browsed in Lynx. If Lynx can do your site, the older graphical browsers should.

--em

Re: Graceful degradation (4.16 / 6) (#26)
by MugginsM on Tue Jun 12, 2001 at 05:43:35 AM EST

The problem being talked about is not so much that the older browsers don't support the newer standards. It's more that certain specific browsers (netscape 4.x, and I hear, IE3) *claim* to support
CSS, and will have an attempt at rendering it. But do such a terrible job that the user will assume your page is broken.

Netscape 3 and lynx will work fine, since they don't know about CSS and will degrade well. The problem is browsers that think they support it and do a shocking job of trying to render it.

That isn't "old". that's "very buggy".

Catering for old browsers is easy - HTML is designed that way. Catering for very buggy, but common, browsers is a different story...

- Colin

[ Parent ]
Not on Netscape 4 (3.00 / 1) (#82)
by IntlHarvester on Tue Jun 12, 2001 at 11:05:16 PM EST

One of the main problems with Netscape 4.x is that it doesn't degrade gracefully at all. Mixing CSS and legacy HTML style tags seems to be a sure way to crash the browser.

In the spirit of the discussion, I should point out that the native stylesheet engine in Communicator is a proprietary thing called JavaScript Style Sheets. CSS is supported by a translation engine.

[ Parent ]
the way I see it (4.57 / 7) (#6)
by jacob on Tue Jun 12, 2001 at 02:36:38 AM EST

The way I see it, you owe your audience two things:

One, conform to standards. HTML and CSS are the big ones here. Don't ever do anything that isn't allowed by those specifications, and always do everything required by them. Following the specs can be a true pain, but it's absolutely necessary.

Two, pitch to your audience. If the people who are going to be visiting your web site are using browsers that only support a subset of the standards, then by all means program only in that subset. If people who will be visiting your site will be using three different browsers that support three different HTML subsets, then you're going to have to figure out something that works for all of them.

True, you can hang up an "use a browser I prefer or go away" sign, but that's just plain rude. Unless you're Jesus's webmaster and you're setting up www.secondcoming.christ, nobody owes it to you to visit your site. If they decide to visit it, then you ought to treat them like guests rather than beggars at the doorstep. Try your best to accomodate them, and if you can't, then at least have the decency to be apologetic about it instead of telling them they're not worth your time.

Of course, my view pertains mostly to people who are doing web design professionally. If you're doing web design for fun, then hey, do whatever makes you happy.



--
"it's not rocket science" right right insofar as rocket science is boring

--Iced_Up

That's the sort of reason I used Opera 5 to test. (3.20 / 5) (#8)
by static on Tue Jun 12, 2001 at 02:56:57 AM EST

Because if it works in Opera 5, then IE 5 and Netscape 6 are also very likely to get it right.

Only someone using Netscape 4 said there was a problem. One other person was using Netscape 3, but he was already aware he wouldn't see the pretty CSS colours. I even had someone say it looked good in Konquerer!

Wade.

[ Parent ]

Opera is great for testing (3.40 / 5) (#16)
by ubernostrum on Tue Jun 12, 2001 at 03:36:35 AM EST

Not only is it standards-compliant, not only does it render things the way they *should* look, but it also has that wonderful "Validate HTML" that's only a right-click away...I don't know how many times it's found tiny errors I'd overlooked.

Typically I also view in IE 5.5 and Netscape 6, but Opera's my main test bed...IE is...well, it's IE, and Netscape 6 still doesn't handle CSS entirely right...I've had effects not show up, or render improperly, even though every other browser on the face of the earth has no problem. Konqueror seems to do a good job, too.




--
You cooin' with my bird?
[ Parent ]

Opera is okay for testing, IMO (4.00 / 3) (#56)
by mauftarkie on Tue Jun 12, 2001 at 12:52:56 PM EST

I do check under Opera (admittedly, under Linux), even though I only have a few clients that use it. I tend to find the opposite, though -- that Opera for Linux doesn't always follow the spec. Having only used Opera under Windows in the early days (and not recently), I have no idea if they use the same rendering engine or not. My understanding is that they're similiar, but different. YMMV.

I personally try and find the middle ground -- something that looks great in Mozilla, IE 5, and Opera, which is actually fairly easy to do as long as you don't try anything super fancy. Presentation in NN4, for me, is a secondary requirement. I consider it antiquated technology at this point, and if people want to stick with it that's their choice. If it doesn't look great, I don't care. As long as my content is mostly functional, I'm happy. I don't rely heavily on Javascript (and when I do, it's simple form checking code), which is probably why I've been so successful with my approach. Something I used for testing before my Mac laptop was stolen was iCab. It doesn't support CSS (yet), but it gives a visual smiley face for when you write spec'd code. They even have a Browser Test that tests some of the more exotic HTML4 features. I have yet to see IE, Mozilla, or Opera render this page correctly.

FWIW, I still use NN4 90% of the time myself. I'm not fond of Opera's GUI interface (I know people swear by it). Mozilla still has a few annoying bugs (although I'm using it more and more each passing week). IE isn't available for my platform. I refuse to run KDE/Konq because KDE refuses to allow me to run my system the way I want and because the QT widget set doesn't appeal to my tastes. If/when Mozilla gets stable enough, and I can compile something like Galeon (which always seems to bomb on my system), I'll use it. Until then...

Of course, everyone is different. What doesn't work for me will probably work for you. Just my 0.02$ American.


--
Without you I'm one step closer to happiness without violence.
Without you I'm one step closer to innocence without consequence.


[ Parent ]
Guest (4.00 / 6) (#30)
by strumco on Tue Jun 12, 2001 at 06:54:04 AM EST

nobody owes it to you to visit your site
Absolutely. With very few exceptions, web publishers need visitors more than the vistors need them.

then you ought to treat them like guests rather than beggars at the doorstep
Actually, I would argue that it's the other way round; that the web publisher is using the reader's screen, computer, modem, phone-line, internet account and, above all, his time. Under such circumstances, one ought to be, at least, polite.

DC
http://www.strum.co.uk
[ Parent ]

The answer: (3.40 / 5) (#10)
by Inoshiro on Tue Jun 12, 2001 at 03:09:48 AM EST

is yes.



--
[ イノシロ ]
IE6 + test (3.66 / 3) (#34)
by Highlander on Tue Jun 12, 2001 at 07:19:32 AM EST

Seems it doesn't pass the test .. is the test correct ?

Moderation in moderation is a good thing.
[ Parent ]
Yes, it's very correct. (5.00 / 1) (#95)
by Inoshiro on Fri Jun 15, 2001 at 03:26:48 AM EST

Feel free to click the obvious w3 validator links, or compare against the reference images.

In order of usefullness are Mozilla, Mac IE, and Opera 5. Opera 5 does not support some of the CSS2 used.



--
[ イノシロ ]
[ Parent ]
Heh (3.66 / 3) (#60)
by delmoi on Tue Jun 12, 2001 at 01:04:49 PM EST

Dosn't seem to work right in IE5.5 :P
--
"'argumentation' is not a word, idiot." -- thelizman
[ Parent ]
That's pretty tough. (3.00 / 1) (#76)
by static on Tue Jun 12, 2001 at 07:23:47 PM EST

Nice to see it displays okay in Opera 5.11, though.

Wade.

[ Parent ]

There's precedent (3.83 / 6) (#13)
by Pseudonym on Tue Jun 12, 2001 at 03:23:31 AM EST

The Web Standards Project certainly thinks this is a good idea.



sub f{($f)=@_;print"$f(q{$f});";}f(q{sub f{($f)=@_;print"$f(q{$f});";}f});
Funny thing is... (3.25 / 4) (#35)
by iGrrrl on Tue Jun 12, 2001 at 07:23:15 AM EST

...the page you pointed to in the webstandards site has one of those annoying little "features": Using the 'back' button merely reloaded the page. OT? Maybe not, since I'm on Netscape 4.6.

--
You cannot have a reasonable conversation with someone who regards other people as toys to be played with. localroger
remove apostrophe for email.
[ Parent ]

Does any browser handle it correctly? (3.33 / 3) (#36)
by i on Tue Jun 12, 2001 at 07:32:31 AM EST

I mean, pages that do redirection should be skipped by "back" (but not "forward"!) button. Is there any browser that does it?

and we have a contradicton according to our assumptions and the factor theorem

[ Parent ]
IE does. (3.00 / 3) (#47)
by Holloway on Tue Jun 12, 2001 at 10:15:28 AM EST

Internet Explorer 4+ does. Opera does.


== Human's wear pants, if they don't wear pants they stand out in a crowd. But if a monkey didn't wear pants it would be anonymous

[ Parent ]
IE5 (3.00 / 2) (#59)
by delmoi on Tue Jun 12, 2001 at 01:03:34 PM EST

IE 5 dosn't work if you use javascript to jump the page forward.
--
"'argumentation' is not a word, idiot." -- thelizman
[ Parent ]
Yeah, but... (4.00 / 4) (#45)
by Pseudonym on Tue Jun 12, 2001 at 09:30:06 AM EST

But, you see, that's not a standard. We can be antisocial, so long as we don't violate any standards. :-)



sub f{($f)=@_;print"$f(q{$f});";}f(q{sub f{($f)=@_;print"$f(q{$f});";}f});
[ Parent ]
Rather plainly, WaSP are poopheads (4.00 / 3) (#48)
by Holloway on Tue Jun 12, 2001 at 10:21:52 AM EST

That's not the only thing.

A WaSP page would be white text and white BGCOLOR but they'd validate, damnit!


== Human's wear pants, if they don't wear pants they stand out in a crowd. But if a monkey didn't wear pants it would be anonymous

[ Parent ]

Hrm (4.00 / 3) (#58)
by delmoi on Tue Jun 12, 2001 at 01:02:48 PM EST

Works fine on my machine. What causes that (in all the cases I know of) is that the browser gets redirected to another page. When you try to go 'back' it'll take you back to the redirect page, and you get, well, redirected. I didn't have that problem with IE5, so it might have been that you got a special redirect because you were using Netscape.

I find that clicking the back button twice really fast fixes it. On all my 'redirect' pages I have a delay so that you can still hit the back button again without any problems.
--
"'argumentation' is not a word, idiot." -- thelizman
[ Parent ]
Netscape 4 == annoyance (3.28 / 7) (#14)
by ubernostrum on Tue Jun 12, 2001 at 03:26:15 AM EST

I recently did a major overhaul of my personal site, and finally got around to moving the bulk of the page formatting into style sheets, for easy overhauls the next time around. The single hardest part of it was making sure everything I wanted to do displayed properly in Netscape 4.x...keeping in mind that I'm on a (from August to June - during summer I host somewhere else) student web server at a college where Netscape 4.76 is the standard (IS installs it on every computer if it's not already there), that was fairly annoying. I think one big step to solving this would be convincing IT and IS folks at places like this to upgrade regularly - the average user will keep using what he's got until it doesn't work anymore, but the guys with the authority (read: "the geeks who fix my computer when it breaks") can probably jumpstart that a little...

For my part, I've applied to work in IS at my school, and I try to convince people to upgrade whenever possible...usually I just show them some of my "fun" test pages and say "See what you can have if you upgrade?" But I think that a combined peer pressure/IT/IS campaign for upgrading browsers could help a lot.




--
You cooin' with my bird?

Version 5 Browsers (3.00 / 9) (#15)
by pwhysall on Tue Jun 12, 2001 at 03:27:19 AM EST

There are many alternatives.
--
Peter
K5 Editors
I'm going to wager that the story keeps getting dumped because it is a steaming pile of badly formatted fool-meme.
CheeseBurgerBrown
And for the Mac... (4.00 / 3) (#91)
by tfrayner on Wed Jun 13, 2001 at 12:05:49 PM EST

iCab, a little standards-compliant browser which is coming along nicely.



[ Parent ]

Amen to that (2.16 / 12) (#17)
by byoon on Tue Jun 12, 2001 at 03:38:01 AM EST

Could someone write a virus that searches out Netscape 4 (or less)/IE 3, uninstalls them and leaves Opera 5, Mozilla or IE 5 in their place?

Testing things and kludging them to look OK in old browsers takes up at least 50% of development time in my experience, esp. if you do want to use CSS.

One thing I keep in mind is my intended audience. When I was designing a site for my little PC repair/upgrade business I figured anyone who needed my business would be using IE 5, and for the most part I was right. A few still used NS 4x but those were few and far between and my half-assed workarounds looked OK in those.


"I'm a going to break you down into the little cubes." -Picasso
No. (2.00 / 5) (#32)
by Highlander on Tue Jun 12, 2001 at 07:14:49 AM EST

Mozilla crashes as much as Netscape 3 did on unix.

In addition, it takes about 30 seconds to start up.

Enough said.

Moderation in moderation is a good thing.
[ Parent ]
30 seconds? (3.33 / 3) (#57)
by delmoi on Tue Jun 12, 2001 at 12:58:09 PM EST

Hrm on my machine after loading mozila -turbo (which takes about 16 megs of ram) Mozilla is loaded by the next screen refresh. Faster then I can move my mouse to it.

Maybe you should try this on something other then your 486 with 32 megs of ram.
--
"'argumentation' is not a word, idiot." -- thelizman
[ Parent ]
more numbers .. (3.00 / 2) (#89)
by Highlander on Wed Jun 13, 2001 at 06:21:34 AM EST

Well, the beonex release needs only about 3 seconds, but the official Seamonkey7 release needs about 20 seconds( with stripped executable ). Any why isn't it installed with turbo as default :-)

Moderation in moderation is a good thing.
[ Parent ]
486 ? (3.00 / 2) (#90)
by Highlander on Wed Jun 13, 2001 at 06:25:22 AM EST

And no, it is not a 486 - the comp I use goes 800MHz, lots of memory - still needs 20 seconds to start seamonkey 7 official release default install.

However, using turbo reduces time to 7 seconds which is almost OJ. But who knows about the "turbo" option ?

Moderation in moderation is a good thing.
[ Parent ]

Netscape 4, IE 3 and CSS (4.07 / 14) (#18)
by driptray on Tue Jun 12, 2001 at 03:42:02 AM EST

CSS is designed to be optional. Its part of the spec. As a web designer you must realise that users may not have a CSS browser, or may disable CSS. Therefore your site must still work OK without CSS.

The other side of this is that users who have browsers with a badly broken CSS implementation (Netscape 4, IE 3) should disable CSS in their browser, or get a new browser.

So it's:

  1. Web designers have a responsibility to adhere to standards, and
  2. Users shouldn't expect that their buggy browsers will be catered to.

But note that I said buggy browsers. Lynx is not buggy AFAIK, and its CSS implementation is flawless.


--
We brought the disasters. The alcohol. We committed the murders. - Paul Keating
Don't exploit software bugs. (4.00 / 5) (#40)
by Holloway on Tue Jun 12, 2001 at 08:36:26 AM EST

2. I don't think it's reasonable to expect users to know that it's their buggy browser, they'll just think your site is ugly and difficult to use (and they're right). It takes two to render a page and they're both responsible.

When people think only one is responsible they're political and are not providing what's best for the user. They often talk about what's the best for the future (standards, standards, standards), or that new standards are broken and need to be changed. They use this to leverage their side: 'I make code that only works in Mozilla - die M$ Internet Exploiter!!' vs 'I have 70% of the market share so I can force MS-HTML'.

I hate both of them. I [heart] users.

The coding style change between TABLEs and DIVs (any full CSS implementation should be DIVs) means a lot of interface elements are lost. Most colour, the columned layout, font sizes and families - gone. That's not good hacker, that's not good.

One thing I've noticed about the transition to CSS is that there's the misunderstanding that this is an end to hacks. Instead of that nested table image mouseover problem in Netscape 2 we have the align:justify bug in Opera. So it's new hacks to get around bugs - but we have standards on our side. Hurrah or something?

Finally, in the last year, I trust that standards have won the browser war (so far as webpages go - plugins are another matter).

The best page for a user will always be a mishmash of official and unofficial standards (and software will always have exploitable bugs - so don't exploit them).


== Human's wear pants, if they don't wear pants they stand out in a crowd. But if a monkey didn't wear pants it would be anonymous

[ Parent ]

Correction (2.33 / 3) (#41)
by Holloway on Tue Jun 12, 2001 at 08:42:36 AM EST

Netscape 2 = Netscape 3.


== Human's wear pants, if they don't wear pants they stand out in a crowd. But if a monkey didn't wear pants it would be anonymous

[ Parent ]
Why not? (4.00 / 3) (#43)
by AndrewH on Tue Jun 12, 2001 at 09:10:49 AM EST

I don't think it's reasonable to expect users to know that it's their buggy browser
But if their browser is buggy, especially if it is (excuse the oxymoron) Netscape4.X-quality, they are really better off getting one that works.
John Wilkes Booth, Lee Harvey Oswald, John Hinckley Jr — where are you now that we need you?
[ Parent ]
I mostly agree. (3.50 / 4) (#46)
by Holloway on Tue Jun 12, 2001 at 10:09:52 AM EST

But if their browser is buggy, especially if it is (excuse the oxymoron) Netscape4.X-quality, they are really better off getting one that works.
Oh you're completely right, it is buggy. They would probably be better off getting a modern browser. But expecting users to understand or even care about that is unreasonable. Users just want to use a website and get in and out. Downloading a browser and installing it is beyond most people - and that's if they have that option. They may be using a system they can't administer, at a hospital's network for patients (as a case I know), or they may not have the system requirements for a modern browser (moores law: software gets twice as slow every 18 months).

PS: as a preemptive strike, nope, I don't think that it's the users fault and they should accept a layout far below what their browser can do because of a new standard. Their browser can do columned layouts with bgcolors, coloured text, different fonts sizes and families. These make a website attractive, more comprehendable and usable (every interface book I've read says that users respond to these elements).

But then in most cases it (hopefully) shouldn't be about choosing a side. Options are best, and browser detection with server-side templates is where the effort should go - not in forcing a mostly foreign page down the browsers throat and to make using the website more difficult for the user.


== Human's wear pants, if they don't wear pants they stand out in a crowd. But if a monkey didn't wear pants it would be anonymous

[ Parent ]

just tell those whom it's broken for? (3.00 / 1) (#73)
by coffee17 on Tue Jun 12, 2001 at 05:27:48 PM EST

I don't know jack about CSS, but let me ask, would it be possible to set things up to just put up a message if someone is viewing it with NN 4 or IE 3 *with* CSS enabled, and have the message be about how CSS is buggy in those browsers with a link to some major up to date browsers (opera, NN, IE, mozilla, lynx), and instructions on how to turn off CSS if the user doesn't feel like upgrading?

-coffee


[ Parent ]

Not really, sorry. (4.00 / 2) (#85)
by Holloway on Wed Jun 13, 2001 at 12:03:39 AM EST

CSS doesn't have a way of querying the browser and showing an upgrade messages. It does have a display:none attribute that some silly buggers suggest using so that CSS incompatible browsers won't know to hide the upgrade message... but that's bad for a few good reasons. So no, CSS doesn't have an appropriate way of including an upgrade message. Natch.

Using Javascript is good. Server-side detection of the browser is better still. However these can always fail (I use junkbuster which apparently reports my browser as something other than what it is - and I don't have javascript turned on) so redirecting the user to an upgrade page is IMO is a bad idea. For example, NZoom keeps telling me to upgrade despite the fact I can render their stupid pages (and I'll never return to that site - stuff is much kinder). I think it's better to have a little upgrade message at the top of each page and not be rude about the request.


== Human's wear pants, if they don't wear pants they stand out in a crowd. But if a monkey didn't wear pants it would be anonymous

[ Parent ]

Web design is political (4.00 / 1) (#79)
by driptray on Tue Jun 12, 2001 at 08:20:10 PM EST

When people think only one is responsible they're political and are not providing what's best for the user.

Web design *is* political. Whose side are you on, the users or the deezignars?

As for providing "what's best for the user", doesn't that phrase seem a little condescending? Who really knows what's best for the user? If it's anybody it's the the users themselves.

Now I'm not a total purist on this issue. Pragmatism rules, so of course as a web designer you will have to make concessions to various browser bugs that are "out there". But there comes a time when coddling one group of users negatively impacts another group of users. Which users should be rewarded - the ones persisting with buggy implementations or the ones that have made better choices?

The coding style change between TABLEs and DIVs (any full CSS implementation should be DIVs) means a lot of interface elements are lost. Most colour, the columned layout, font sizes and families - gone.

They were never there for a lot of users, or if they were, they constituted more of a problem than a feature. These had serious usability issues for the most part, and caused a hell of a lot of failure. Font sizes? How about letting the user choose that. Font families? Mostly harmless 'cos they could be overriden by the user. Font and Table background colours? No way (in Netscape) for users to override these, so clashes with user colour schemes occurred. Columned layout? OK in some circumstances, but typically used with fixed widths that take no consideration of browser window width or user font size.

I'm a big believer in user control. That's both the reality and the philosophy of the web, although the reality is constantly under threat by ever more web designers trying to rigidly control the presentation of their pages.

So let's not condescend to users. Let's recognise their power and control over the web. But let's also recognise the responsibilities they have in exercising that power and control.


--
We brought the disasters. The alcohol. We committed the murders. - Paul Keating
[ Parent ]
Nossa condescention. (4.00 / 1) (#81)
by Holloway on Tue Jun 12, 2001 at 11:03:46 PM EST

I think your idea about 'condescending to users' hinges on my bluntly put 'provide what users want'. Their wants and need are shown throughout every interface/usability book. I don't think providing for these is condescending at all. There are good and bad interfaces, how people use things is a science.

Any product/service involves making use of your expertise and providing what you think is best for the user - there's nothing condescending about it. I think dictating for the user to have a certain browser to get a more usable page is condescending (standards or not - the user is of utmost importance to me).

I agree that user choice is always best.

Which users should be rewarded - the ones persisting with buggy implementations or the ones that have made better choices?
The larger and growing market share. But your idea of "better choices"... now that sounds condescending (not really, I'm just playing wit ja). But better choices? I sez:
"Users just want to use a website and get in and out. Downloading a browser and installing it is beyond most people - and that's if they have that option. They may be using a system they can't administer, at a hospital's network for patients (as a case I know), or they may not have the system requirements for a modern browser (moores law: software gets twice as slow every 18 months)."

There are many good reasons why someone may be using an older browser. That older browser can do many interface elements that users respond to - but only if you cater to them.

I'll continue to take the time to provide for these 8% of users. I won't be a lazy sod.

ps. When talking about breaking standards (mish-mash of official and unofficial) I don't mean doing stupid things. I mean doing things that work and are unofficial. I'm not among the among the BLINK/MARQUEE/fixed-width-table tag crowd.




== Human's wear pants, if they don't wear pants they stand out in a crowd. But if a monkey didn't wear pants it would be anonymous

[ Parent ]
Ooooh! Shiny new feature! (3.75 / 12) (#23)
by Arkady on Tue Jun 12, 2001 at 05:01:23 AM EST

Lookie! A new feature in [x] browser! Let us immediately rush to use this!

If interoperability still means anything today, and I rather doubt that, then you should be looking at the "any browser" pages.

It's absurdly arrogant to produce a site which only functions on your own browser of choice, as though you content were so manifestly marvelous(even to folks who can't see it yet) that everyone would feel impelled to go out and intall your favorite browser just to earn the honor of viewing it.

So, get a grip on yourself there, laddie, and think about why people would want to look at your content and whether it's really necessary for you to have that Flash animation or that style-sheet to get your point across.

-robinĺ


Turning and turning in the widening gyre
The falcon cannot hear the falconer;
Things fall apart; the centre cannot hold;
Mere Anarchy is loosed upon the world.


Not about shiny features (3.62 / 8) (#27)
by pwhysall on Tue Jun 12, 2001 at 06:03:40 AM EST

This isn't about supporting "shiny new features". It's about supporting web standards.

The fact that Netscape 4 has such horribly broken CSS shouldn't really figure in the mind of a web designer, in an internet landscape that is home to Mozilla, IE5, Opera and Konqueror, all of whom do web standards much better than the old, broken Netscape.

(I know there's prolly a bunch of browsers on things like the Mac and OS/2 that do CSS properly, too, but I'm not aware of them - so don't flame me :)
--
Peter
K5 Editors
I'm going to wager that the story keeps getting dumped because it is a steaming pile of badly formatted fool-meme.
CheeseBurgerBrown
[ Parent ]

Not necessarilly. (3.80 / 5) (#37)
by gcmillwood on Tue Jun 12, 2001 at 08:05:20 AM EST

The question should really be, "Who do I want to view my website?" If you don't make the appropriate adjustments to make NS4 render the pages correctly, then don't expect NS4 users to use the site. For most site owners, missing out on NS4 traffic is not going to be a big deal.

In some situation, the audience for your web-site can be single browser (i.e. corporate intranets) and in this situation you can use browser specific 'features'. Even in this situation I'd still try to keep to the standards though - who knows when the corporate browser may change, and what it may change to.

[ Parent ]
I agree (4.50 / 4) (#61)
by pwhysall on Tue Jun 12, 2001 at 01:10:17 PM EST

I don't think webmasters should code to any particular browser. They should code to whatever version of (X|D)HTML they claim to have in the DOCTYPE declaration, and then rely on the browsers to correctly display this.

Time spent working around the broken-ness of something like NS4, for example, is time that could be better spent elsewhere; worse, it's time spent not encouraging people to get a proper browser.

I think that it's only the willingness of webmasters to put up with the sack of cack that is NS4 and its CSS that meant that Netscape never actually bothered fixing the damn thing because hey, everyone's coding for us.
--
Peter
K5 Editors
I'm going to wager that the story keeps getting dumped because it is a steaming pile of badly formatted fool-meme.
CheeseBurgerBrown
[ Parent ]

Reality Check (2.33 / 3) (#39)
by sventhatcher on Tue Jun 12, 2001 at 08:30:00 AM EST

It's not about the newest neatest features.

It's not about promoting your favorite browser.

It's not even about being arrogant or presumptious (sp?).

It's about wanting to format and wanting to control the way your web page looks. I know HTML wasn't really designed to do any real formatting, but it's too late now to go back and revise the standards.



[ Parent ]
it is viewable in any browser (3.33 / 3) (#53)
by delmoi on Tue Jun 12, 2001 at 12:31:41 PM EST

Any browser that doesn't suck. I mean really. CSS, JavaScript, etc. All of those are open standards published by the W3C. It's not my fault that CSS is horribly broken in NS4, and I'm not going to take responsibility for it anymore.
--
"'argumentation' is not a word, idiot." -- thelizman
[ Parent ]
That is a solution, yes. (3.00 / 1) (#77)
by static on Tue Jun 12, 2001 at 07:40:13 PM EST

But what prompted me to post this story in the first place is that such "graceful degradation" gets easily derailed when trying to implement CSS and faced with NS4 and IE3. One way around this is to minimize using the parts of CSS that are broken in NS4 and IE4. That's a lot of effort! The other way is simply draw a line saying that anything older than a version 5 browser is not tested thoroughly and the user should try to upgrade.

Wade.

[ Parent ]

Not time for CSS (3.00 / 8) (#38)
by Tachys on Tue Jun 12, 2001 at 08:15:23 AM EST

Because every browser that supports them sucks.

They always have tons of crappy features you don't want.

Mozilla was supposed to save us from this because open source was supposed to be immune to bloatware.

But that was false

I just hope kmeleon can deliver a small fast browser that doesn't suck



Whatever (4.00 / 3) (#52)
by delmoi on Tue Jun 12, 2001 at 12:27:39 PM EST

At least 95% of the world uses a good CSS capable browser (IE or mozilla, even opera). Why should I optimize for a small chunk of that 5% made up of techno elitists who think that Mozilla (which loads faster then IE5.5 on my machine now with -turbo) is 'too bloated'.

CSS makes designing web pages a lot easier, and lets you make them a lot cooler. and I'm sorry if you think mozilla is to bloated to use, but I really don't care.
--
"'argumentation' is not a word, idiot." -- thelizman
[ Parent ]
Seven years are not enough ? (3.00 / 1) (#71)
by camadas on Tue Jun 12, 2001 at 04:05:59 PM EST

Yes, CSS were first presented in 1994. There was plenty of time to implement them since. Even Microsoft did it reasonbly well, first in the Mac Version of IE5 and now in IE6. Opera with less resources managed to implement it (with some quirks). Netscape didn't, they just patched it on in a way that breaks valid CSS, instead of ignoring it. And that is not acceptable.

[ Parent ]
Seven years? Bah. (3.00 / 1) (#84)
by Holloway on Tue Jun 12, 2001 at 11:22:32 PM EST

Seven years has nothing to do with it (unless you like coding for software that doesn't yet exist... you know, for fun). CSS isn't ready because not enough browsers implement enough of it.

Of course what's enough is debateable.


== Human's wear pants, if they don't wear pants they stand out in a crowd. But if a monkey didn't wear pants it would be anonymous

[ Parent ]

Supporting NS4 without going braindead (4.36 / 11) (#42)
by Dion on Tue Jun 12, 2001 at 09:07:33 AM EST

There is one easy way to support NS4: Don't.

Now I'm not saying that you should only support more popular browsers, what I'm saying is: Don't support any browser at all, support standards!

Build everything using sane guidelines, read some Jacob Nielsen for hints

Personally I use HTML that is as close to HTML4.01 strict as I can make it (sadly there is no form formatting available that can validate, so I use tables for formatting my forms) and CSS2, with absolutely no client side scripting at all.

Now everyone who has ever seen the broken old NS4 attempt CSS will tell me that this doesn't work, well, you are right, so my server detects NS4 and simply leaves out the inclusion of the CSS, this way the site still works perfectly in NS4, even if it does look kind of boring.

One *should* put some time into degrading a bit more gracefully, you could give the NS4 clients a different, simpler (and probably broken) CSS that would allow it to render the page somewhat correctly.

Trying to get pixelperfect rendering from a webbrowser is wrong and should not be attempted, so I'm not really bothered by having pages show up looking differently in different browsers, so my sites support everything from Lynx and NS to IE5, Mozilla, Opera and Konqueror, it just looks very nice in the good browsers.

Note: No my personal webpage is *not* one of those sites, it was built for show, not for function.



The Narrow Minded Web (3.75 / 8) (#49)
by PresJPolk on Tue Jun 12, 2001 at 10:24:12 AM EST

One thing that always disturbs me in discussions of the WWW is the common, misguided focus on presentation, browsers, and anything other than the goals.

Just use standard HTML, with CSS, and present your information. That's it. That is your goal, after all, right? You want to present information to your reader. Make it easy for your reader. Never depend on anything that may be an obstacle for a potential reader, including java, graphics, cookies, or scripting.

The value of the WWW is its accessability, far beyond that of any other network of information. By creating HTML documents that rely upon anything that locks your reader into one class of HTML reader or internet connection, you restrict the class of people who can read your information, and create inconveniences for those who can.

In short, follow the standards, people.

Hrm? (2.33 / 3) (#51)
by delmoi on Tue Jun 12, 2001 at 12:22:23 PM EST

Never depend on anything that may be an obstacle for a potential reader, including java, graphics, cookies, or scripting.

Asside from Java, all of those things *are standard* And when you're doing something other then presenting text you might need to use some of those.

And why shouldn't we present something that might be 'an obsticle' to our readers? Do they have some god given right to view our content?
--
"'argumentation' is not a word, idiot." -- thelizman
[ Parent ]
Americans with Disabilities Act (3.50 / 4) (#69)
by pin0cchio on Tue Jun 12, 2001 at 01:45:11 PM EST

And why shouldn't we present something that might be 'an obsticle' to our readers? Do they have some god given right to view our content?

If you do business in the U.S., and you have more than 15 employees, you are subject to the Americans with Disabilities Act. Users behind screenreader-friendly browsers (lynx, w3m, and Links) have trouble with java, graphics, frames, scripts, etc. In fact, the W3C recommends that you add textual captions of both the video and audio track of every video on a web site. For example, if you provide the video, the accessibility guidelines recommend that you also provide a screenplay, synchronized to the video with SMIL.


lj65
[ Parent ]
Incremental standards (4.87 / 8) (#70)
by fluffy grue on Tue Jun 12, 2001 at 03:00:48 PM EST

The other thing about the standards is that they are incremental. No website should rely solely on any feature which isn't in base HTML. A site which is extra-flashy and wonderful-good in MSIE 5 should be at least usable in Lynx (which is about as minimal of an HTML processor as you can get). It doesn't have to be pretty and flashy and such, but it needs to at least be navigable!

And why are there so many websites out there which reimplement HYPERLINKS in JavaScript?! It takes a lot less code to do an <A HREF="foo"> than it does to do some of the fugly Javascript hacks I've seen for the same goddamned functionality!

I'm fine with sites breaking visually on my browser of choice (Netscape 4.77 with Javascript normally turned off), as long as I can still get to the information. None of the sites which become totally inaccessible on Netscape have any good reason to be; they've become inaccessible because they don't have an alternate browsing mechanism on top of their stupid flashy floaty-everywhere-impossible-to-fucking-click hyperlinks.

Also, not everyone uses a mouse. HTML doesn't specify a mouse. Not everyone even uses a monitor. I find it disgusting that Korg, a fucking synthesizer manufacturer, deems it necessary to have so much visual flash that even with a browser which doesn't choke on their site that blind people wouldn't have any chance of finding out about their products. (Yes, blind people use synthesizers too. Ever heard of Stevie Wonder?) And anyway, their site is ultra-flashy for no good reason. A few days ago I just wanted to find out something about their new Karma synth workstation, and even after I bent over backwards to get the site working (i.e. navigable) I still couldn't actually find any of the information I wanted on it because the site was so fucking confusing to navigate!

Here's my acid test for whether a website is broken or not: If I can navigate and read it in Lynx (not w3m; w3m's a nicer browser, which goes against the point of this test), it's not broken. If it works in Lynx, it'll work in any browser for any situation. If it doesn't work in Lynx, it's either because it's hostile about using Javascript for its navigation, or it doesn't have proper ALT attributes on its images.

You will notice that every single website I have designed works perfectly in lynx, while still being relatively attractive (and still usable, which is something most webdesigners have apparently no clue about) in more modern browsers. I could easily add in Javascript for mouseovers and other eyecandy and it'd work just fine. I could migrate over from tables (which don't work in lynx, which is why lynx is the acid test and not w3m) to CSS and it'd still work just fine (it'd be butt-ugly in older browsers, but at least it would work).

Websites must be designed to gracefully degrade; no single feature which isn't in base HTML should be assumed to be present! I can't reiterate this enough - I am perfectly fine, fine, FINE with people using advanced features - but don't rely on them!
--
"Is not a quine" is not a quine.
I have a master's degree in science!

[ Hug Your Trikuare ]
[ Parent ]

yes yes yes yes (3.00 / 1) (#74)
by PresJPolk on Tue Jun 12, 2001 at 07:06:30 PM EST

That's a much better version. Not "Don't rely upon X", but "Make sure your HTML degrades well in the absence of X".

[ Parent ]
I must add this. (3.50 / 2) (#78)
by static on Tue Jun 12, 2001 at 07:44:53 PM EST

I surfed Scoop the other day from Lynx. Even posted a few comments. I was surprised how navigable it was.

Wade.

[ Parent ]

It's not limited to CSS (3.00 / 8) (#54)
by silentz on Tue Jun 12, 2001 at 12:46:01 PM EST

IMHO, CSS is reasonably well implemented in Netscrape compared to most of its unpardonable horrors...
  • Take using bgcolor on a table or cell to give it a nice background colour, then putting a table on top... where's the bgcolour gone? To get that nice hairline effect, I have to use bloody 1 pixel graphics... That is toss. And it's just HTML - no "new features" here!
  • What about the fact that if i set my <input> box to "15 characters wide" Netshite makes it at least twice that... Unforgivable!
It seems that Netscape actually want no one to use their browser.

And the worst thing of all is that I have to check all my pages in NN and spend hours getting them right for some scrote who "doesnt like bill gates"

Grow up.

What the hell is wrong with coders these days? (4.30 / 10) (#87)
by Waldo on Wed Jun 13, 2001 at 01:22:38 AM EST

This is so damned easy. I have been developing websites for...oh...going on eight years now, I guess. All along, my clients have been complaining to me before I even get started. In 1994, they wanted to know why they couldn't have audio on their pages. In 1996, they wanted to know why they couldn't have a full-screen movie as a splash screen. In 1998, they wanted to know why their site didn't take advantage of all of IE's stupid new image-morphing functions. And now, they want to know why their site can't take advantage of any number of DHTML / CSS2 / WhizzBang 3.5 features.

And I'll tell you the same thing that I told them then: you damned well don't need it. There are very few HTML widgets that one would ever need to construct a site that won't work in v3 browsers. If I had a buck for every time that I fixed some whiner's crappy code to make it work in every browser under the sun, I'd be...well, not rich, but I'd have a good $30, for sure.

Note that I said "make it work." Not "make it look identical," or "create a photo-reproduction of how it looks on their machine." Just make it work. Make it not ugly. Make it looks something like what they intended. But you know what? It's HTML. It's a suggestion of an appearance. You want something identical? Start making yourself 580px GIFs and start imagemapping. Short of that, just stop creating layouts that aren't feasible in earlier browsers. I'm sure that we can all come up with designs that won't work in 3.0 browsers. This is like, in the print world, coming up with designs that involve an array of spot colors, five-color printing and a custom, detailed die-cut, and then bitching when you're told that it will involve a special press across the state. Don't do that.

Do all of my sites work for every browser ever? Hell no. With some of my sites, I just don't care if they work for everybody. Do most of them? Absolutely. If your design isn't working in earlier browsers, face up to the two most likely facts: you've either designed poorly, or your code sucks.

Not that I'm bitter. I'm just sick of this discussion. I've had it about once a week since 1993. People don't know much more now than they did then, it seems.

-Waldo

So... are you advocating sticking with HTML+? (3.00 / 1) (#92)
by static on Wed Jun 13, 2001 at 10:19:10 PM EST

Because that wasn't my point. I want to use current technology - XHTML 1 and CSS2 - as much as possible on the web sites I have control over. But Netscape 4 (and IE 3) break under that. XHTML 1 and CSS2 degrades fairly okay under NS3 because it has no clue about CSS. It is Netscape 4's badly broken CSS that is my issue here. I don't want to have to support that.

Wade.

[ Parent ]

whose code sucks? (none / 0) (#94)
by rfsayre on Thu Jun 14, 2001 at 09:08:14 PM EST

One of the most objective ways to rate the quality of someone's code is to run it through a validator. Validation is not the last word on a site, sometimes it may be in your audience's best interests to ignore some specs. However, a quick test of some of the "professional" sites you've done reveals that none of your pages validate, which is rather shocking given the spartan look of you pages. The validator also reveals that you probably never ran them through a validator either, since there are typos in the tags. Hint: that's one of the best reasons to use a validator.

The issue of whether you're designing at all is debatable. I would use the term "marking up text files". I can understand why many people think that's the best way to go, but I wouldn't call it design. Telling every single one of your clients that they need the same thing has got to be wrong. Think about what consumers equate web sites with: magazines and catalogs, which are generally layed out well and generally don't have black backgrounds. I would say your code is crap if it doesn't deliver what the client needs.

A lot of sites require a very polished look and feel (which isn't possible without precise positioning), just to stay within the guidelines of the corporate style guide. With these types of sites , the client should determine how much of your time ( and client $$) is spent making the site really nice vs. time versioning out the site for different browsers. Some clients will go for maximum backward compatibility, and others will decide that they can't communicate what they want to in older browsers (or without Flash 5). They probably know their audience better than you do, so you should do what they say. You should try and educate them enough to make informed decisions.

One day the web won't be broken. But sites won't work in 3.0 browsers then, either.

[ Parent ]

Agreed, in Parts (none / 0) (#101)
by Waldo on Mon Jun 18, 2001 at 09:18:24 PM EST

I agree that the most objective method of rating code is running it through a validator, but I don't think that it holds a lot of water. OK, your code meets a spec -- but how does it work? I think that we can all agree that few browsers properly implement any HTML standard. And we can also all agree that it would be grand if they all did. But they don't, and so there's no single HTML spec that, upon being met, will result in code that renders correctly in every relevant browser. So, no, few of the sites that I work on will validate. And that's how I like it -- otherwise, they wouldn't work on enough browsers.

And, yes, some sites don't have to work in every browser. I certainly wouldn't argue against that. There have been a number of occasions when I've made sites (usually for an intranet, or an Internet-based service that uses kiosks, or things of that nature) that only work on a particular browser. In most cases, I could make them work in every browser, but there's no point in billing out that time.

-Waldo

[ Parent ]
Didn't you say you used XHTML ? (none / 0) (#97)
by jneves on Fri Jun 15, 2001 at 07:26:39 AM EST

My last web site has been a new experience. I'm finally rid of this kind of problems. My solution ? XHTML + XSL. I code the side in XHTML and all browser specific code goes in a stylesheet that converts my XHTML in HTML 4.01/Transitional. This way it works with most browsers out of the box (I do not test them all), even if graceful degradation for browsers like lynx depends more on design than on code.

cool! (none / 0) (#99)
by jacob on Fri Jun 15, 2001 at 02:16:34 PM EST

I had been wondering whether somebody would do that. Out of curiosity: did you find that your XML transformations were general, or were they specific to your particular web pages?

--
"it's not rocket science" right right insofar as rocket science is boring

--Iced_Up

[ Parent ]
Some are... (none / 0) (#100)
by jneves on Fri Jun 15, 2001 at 02:29:22 PM EST

My experience is that some of then are generic (replacing <br/> by <br> or <br />) and some are very specific to the looks of the site (font and colors configuration). I'll try to put the generic and specific parts in different files so I can reuse as much as possible. The specific part is, more or less, what you'd put in a CSS file.


[ Parent ]
Browser version at least 5, please. | 101 comments (101 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!