Please use one of these browsers



  • @Monomelodies said:

    The answer is always simple; pixel-perfect simply does not exist in HTML/CSS. Get used to it, or go back to DTP of Flash or whatever. Describing CSS as "trying to add pixel-perfect layout" is a fallacy in itself; there's a reason you can use percentages. Once you let go of the idea of pixel-perfection, things start to get much clearer.

    The problem is that a lot of people want pixel-perfect layout and CSS actually does aim to give that level of control, but obviously falls short.  And I'm all for dumping HTML/CSS and using something like Flash or Silverlight, but then you same standards freaks would throw a fit about that, too.

     

    @Monomelodies said:

    I'd genuinely be interested to see some of the sites you've worked on. I don't mean this as a flame; they really sound like they could be insightful! Personally, I VERY rarely have beef with Opera, Safari or Konqueror once FF works.

    Once again, I'm not making a blog or a simple "about me" or "about the company" site.  These are web apps and unless you want to shell out $100k per-year subscription fees it is unlikely you will ever see them.

     

    @Monomelodies said:

    Well, like I said, it's less of a deal than IE6, but it's still a deal. And some rendering bugs just still exist. I worked around a few today. Unless you consider otherwise-perfectly-fitting floats needing 10px of extra space not a bug, of course.

    I see those kind of problems with floats in Opera and older versions of Safari from time-to-time.  It's a pain, but you just add a little tweak to your browser-specific styles and move on.

     

    @Monomelodies said:

    "7 years old" is IMHO no excuse for "screwing up". Do you use floats at all? Aren't your hairs grey from IE6's seemingly random interpretation of them? Do you think hasLayout is something to be coveted rather than despised? Think the double-margin bug actually makes sense? C'mon.

    Screwing up?  My point was that when IE5 came out it was the best browser, period.  IE remained the best until Firefox started seeing wide use in 2004 or so.   Meanwhile, I've had problems with floats in all browsers.  Usually it takes a very complex UI layout to trip up FF but it happens.

     

    @Monomelodies said:

    No, it doesn't. The word "sniffing" implies as much. CC's /could/ be a way to reliably achieve this. Of course, in a perfect world they wouldn't be needed, but that's another matter. As long as UA behaviour doesn't change between versions they'd be great. Of course, a problem ensues when, I don't know, IE for Mac shows different behaviour then IE/Win. But that's really a vendor problem.

    The two arguments against using the UA to modify the layout are: 1) newer browser versions may be identified incorrectly and not receive the proper tweaks from your software and 2) technically-inclined end-users can change their UA.  Both of these would apply to conditional comments as well.  If a new version of IE comes out and your page does not have conditional comments to handle it properly, your shit will break, same as if you were identifying based on UA string.  See, ultimately it comes down to UA detection which suffers the same problems if it is done server-side or client-side.

     

    @Monomelodies said:

    But only because your head would be rendering what all the other browsers do, which makes this a bit of a moot argument.

    Don't you see that it's your argument that is moot?  No matter what you do, all the browsers have different behavior.  A sufficiently experienced web developer will be able to anticipate and work around these differences, no matter what browser they start from.

     

    @Monomelodies said:

    The words "one browser" and "IE 6/7/8" in one sentence sort of made me laugh.

    It was obvious that I meant one browser product.  When someone says "I develop for FF" I don't consider that 4 different browsers although the differences between 1.0, 1.5, 2.0 and 3.0 were there.

     

    @Monomelodies said:

    FWIW, my sites(sic) are also compatible with any software released between now and 200 years. As long as the standards don't change, or stay backwards compatible...

    Now it's my chance to laugh.  We'll be lucky if software in 20 years will have any clue how to interpret HTML/CSS.  No, wait, we'd be extremely unlucky because it would mean HTML and CSS are still around!

     

    @Monomelodies said:

    Orly? So your apps only work client-side? No server-side fallback? No concerns at all what happens when some script-kiddie with greasemonkey comes by?

    Really, it's almost always better to write out your stuff in plain HTML and make sure the server handles whatever's thrown at it correctly, and then /maybe/ tart it up for JS/AJAX-enabled users. You'll have the best of both worlds.

    What the hell are you talking about?  How did you conclude I said anything close to this?  My point is that the interfaces I've worked on are complicated enough that worrying about older or non-mainstream browsers isn't worth the effort.  If your interface is mostly plain HTML with just a bit of JS eyecandy, more power to you.  Some of us are working with significantly more complicated and demanding requirements and there is no possible way to design an interface that will work in older browsers.

     

    @Monomelodies said:

    GoogleBot says he disagrees.

    What?

     

    @Monomelodies said:

    Actually, I disagree whenever I'm in the datacentre and am stuck to a terminal. Ok, there's a difference between internet and intranet, and sure, you can bend the rules if you know the right parameters. But as a rule of thumb, sir, I must vehemently disagree with you.

    Seriously, get a laptop with a GUI.  Or, hell, drive on the freeway with a shopping cart powered by Erector set motors, just don't be surprised when people give you the finger as they pass. 



  • @morbiuswilters said:

    And I'm all for dumping HTML/CSS and using something like Flash or Silverlight, but then you same standards freaks would throw a fit about that, too.

    My grievance with Flash is the complete lack of control and access the user has over content. (I could even forgive Flash's appalling performance if we had more control, as at least then we could alter the page with something like Stylish to remove all the parts that play irritating background audio and tie up the processor.)

    I've noticed that Flash links at least now have "open link in new window" (or wording to that effect) in their context menu, but what of middle-click and new tabs? How about control of minimum font size? Forcing pages to have a colour for visited links? Page history and bookmarks? Search engine crawling?

    Flash is good for games, but makes for a horrible page presentation system. Flash is maintained by developers who think that IE 6 is a good browser UI and have never ventured outside that safe little shell of theirs; it's to these idiots that we owe XP's automatic taskbar button grouping.

    Now, I'm fully with you on how horrible the Web standards are. The idea that we should rally around IE's implementation is a fallacy however, because IE has been trapped in feature poverty for so long. It's only Microsoft's grudging acceptance of CSS2 that makes Internet Explorer even tolerable, because otherwise we'd still be using nested tables, and you remember how tedious and painful that was. CSS2 itself was a pitiful attempt to capture the needs of the Web, but at least it's forced everyone to make a lot of presentation control available that simply didn't exist in HTML at all. It's not great, but it's a great step up from HTML 3.

    [Edit: What's with the tag box giving me "Macustust" when I type "Must"..??]



  • @Daniel Beardsmore said:

    My grievance with Flash is the complete lack of control and access the user has over content. (I could even forgive Flash's appalling performance if we had more control, as at least then we could alter the page with something like Stylish to remove all the parts that play irritating background audio and tie up the processor.)

    I've noticed that Flash links at least now have "open link in new window" (or wording to that effect) in their context menu, but what of middle-click and new tabs? How about control of minimum font size? Forcing pages to have a colour for visited links? Page history and bookmarks? Search engine crawling?

    I agree for the most part about Flash being pretty mediocre.  What we have to realize is that there are several different uses for the web and that HTML/CSS is not particularly good for most.  Sure, it's fine for simple sites that are text-heavy with a little bit of "pretty".  However, when it comes to complex applications, media-rich sites or sites that aim for a very specific artistic layout, HTML and CSS fall flat.  Of course, we've had the ability to do all of these for awhile, but unfortunately real application and media technologies for the web like Java, ActiveX, Flash, etc.. ended up being technological abortions and now we're stuck in some kind of purgatory where we have to construct our own drag-and-drop, rich GUI interfaces using divs and CSS hacks all bound into a messy ball with Javascript.

     

    @Daniel Beardsmore said:

    it's to these idiots that we owe XP's automatic taskbar button grouping.

    I never particularly liked that feature and always turned it off, but I don't think it was all that bad and most people seemed quite fine with it.

     

    @Daniel Beardsmore said:

    The idea that we should rally around IE's implementation is a fallacy however, because IE has been trapped in feature poverty for so long.

    I think this is because Microsoft tried to push alternate technologies instead of giving in to the demand for a mediocre development platform based on HTML.  In some ways, IE6 may turn out to be a surprise blessing as the browser's relative lack of sophistication has made complex web apps less feasible, perhaps giving alternative technologies a chance to gain footing.  Also, I must point out that it was IE that gave us XMLHttpRequest, without which most of this Web 2.0 stuff wouldn't even be possible.  I'm not actually convinced that it was a good thing, but I think claiming a lack of innovation on the part of MS is a tad unfair.  Other than being better at adhering to an arbitrary standard, what amazing advances has Firefox given us?  Tabs?  Extensions?  Cool stuff, but hardly life-changing.

     

    @Daniel Beardsmore said:

    [Edit: What's with the tag box giving me "Macustust" when I type "Must"..??]

    <span wisdom="pstorer">Welcome to Community Server.  Get used to it.</span>



  • @tster said:

    You do realize that was a joke, right?

    Maybe... but it's also reality, sadly. My sarcasm-meter must've been put off by all this nonsense about IE.

    @tster said:

    Also, why do you put slashes around words that you want to emphasize.  Why not italicize them or something.
     

    /em/
    *strong*
    _underline_

    Sorry, too used to writing plain-text e-mails.



  • @dhromed said:

    What situation is that? I've not had an issue in IE7 with that sort of thing. After IE6 double-margin-bug and 3px-hard-space-bug were fixed, it's all been hunkydory in IE7.
     

    Something along the lines of:

    <style type="text/css"><!--
    h1 {
      width: 100px;
      float: left;
      }
    ul {
      list-style: none;
      margin: 0;
      padding: 0;
      text-indent: 0;
      float: right;
      }
    ul li {
      float: left;
      width: 100px;
      margin-left: 2px;
      }
    ul li.first-child { /* fake pseudo-class for non-supporting UAs */ 
      margin: 0;
      }
    //--></style>

    ...

    <!-- assuming a 406px canvas -->
    <h1>my header</h1>
    <ul>
      <li class="first-child"><img src="someimg1" alt="1"/></li>
      <li><img src="someimg2" alt="2"/></li>
      <li><img src="someimg3" alt="3"/></li>
    </ul>

    For some reason unfathomable to myself, IE (all versions) wouldn't fit all the LIs next to each other, rather dropping the last one down a line. Oddly enough, it positioned right next to the H1 so it /should/ have fitted.

    DISCLAIMER: this example isn't the real world code where it went wrong; you might get this to work properly.



  • @morbiuswilters said:

    The problem is that a lot of people want pixel-perfect layout and CSS actually does aim to give that level of control, but obviously falls short.

    That's why it's our job to explain to them why they can't have it, period. You can't have both pixel-perfect and a UI which allows the user to arbitrarily resize their fonts, to give an example.

    @morbiuswilters said:

    And I'm all for dumping HTML/CSS and using something like Flash or Silverlight, but then you same standards freaks would throw a fit about that, too.

    Yup, you would, but that's got nothing to do with standards (see below).

    @morbiuswilters said:

    Once again, I'm not making a blog or a simple "about me" or "about the company" site.  These are web apps and unless you want to shell out $100k per-year subscription fees it is unlikely you will ever see them.

    Gee, I'm shaking. I'll just ignore the part where you try to assert your seniority by resorting to mentioning $$$, mkay?

    @morbiuswilters said:

    I see those kind of problems with floats in Opera and older versions of Safari from time-to-time.  It's a pain, but you just add a little tweak to your browser-specific styles and move on.

    Of course older versions may have bugs now fixed. Everyone has that "from time-to-time". But that's something else than claiming "FF has as many rendering bugs as IE6", which was basically what you were doing.

    @morbiuswilters said:

    Screwing up?  My point was that when IE5 came out it was the best browser, period.  IE remained the best until Firefox started seeing wide use in 2004 or so.   Meanwhile, I've had problems with floats in all browsers.  Usually it takes a very complex UI layout to trip up FF but it happens.

    Eh? Who mentioned IE5? I've been using good ole' Mozilla since, I dunno, 2002 or so. NN6 before that, though it was horribly bloated. Whether or not IE5 was better than NS4 is a matter of debate (I'd be inclined to agree with you on that).

    @morbiuswilters said:

    The two arguments against using the UA to modify the layout are: 1) newer browser versions may be identified incorrectly and not receive the proper tweaks from your software and 2) technically-inclined end-users can change their UA.  Both of these would apply to conditional comments as well.  If a new version of IE comes out and your page does not have conditional comments to handle it properly, your shit will break, same as if you were identifying based on UA string.  See, ultimately it comes down to UA detection which suffers the same problems if it is done server-side or client-side.

    Ultimately it shouldn't be necessary, that's what standards are for. Anyway, conditional comments allow (for IE) some pretty strict control over versions. If I target my IE7-fixes at IE7 and IE8 doesn't need them, well, all will be well (in theory). Of course, that's assuming MS doesn't break it by parsing IE7 comments "so the web won't break".

    @morbiuswilters said:

    Don't you see that it's your argument that is moot?  No matter what you do, all the browsers have different behavior.  A sufficiently experienced web developer will be able to anticipate and work around these differences, no matter what browser they start from.

    So much is true, but if you need more fixes for Firefox than for IE something weird might be going on. My alarm bells would be tripped, at least.

    @morbiuswilters said:

    It was obvious that I meant one browser product.  When someone says "I develop for FF" I don't consider that 4 different browsers although the differences between 1.0, 1.5, 2.0 and 3.0 were there.

    I know that's what you meant, but the difference between IE6, 7 and shortly 8 is a lot bigger than between FF 1 and 3, which are more evolutions of an already pretty good rendering engine. Web-devs tend to consider IE6 and 7 as separate products needing separate fixes - generally not so with different FF-versions.

    @morbiuswilters said:

    Now it's my chance to laugh.  We'll be lucky if software in 20 years will have any clue how to interpret HTML/CSS.  No, wait, we'd be extremely unlucky because it would mean HTML and CSS are still around!

    Of course I jested.

    @morbiuswilters said:

    What the hell are you talking about?  How did you conclude I said anything close to this?  My point is that the interfaces I've worked on are complicated enough that worrying about older or non-mainstream browsers isn't worth the effort.  If your interface is mostly plain HTML with just a bit of JS eyecandy, more power to you.  Some of us are working with significantly more complicated and demanding requirements and there is no possible way to design an interface that will work in older browsers.

    I'm not worrying about older browsers. What I am saying is that when properly coded, it should degrade gracefully. Of course, if you're building a MS Word-clone in your browser, well, that sort of stuff goes out the window. Though you might want to consider whether or not you've chosen the right platform in that case.

    @morbiuswilters said:

    @Monomelodies said:
    GoogleBot says he disagrees.

    What?

    Or disabled users?

     @morbiuswilters said:

    Seriously, get a laptop with a GUI.  Or, hell, drive on the freeway with a shopping cart powered by Erector set motors, just don't be surprised when people give you the finger as they pass.
     

    There are many reasons why some people prefer text-interfaces. Screen reader software prefers them. Maybe I'm blind or extremely short-sighted? Or maybe I prefer as-is content to sucky marckup generated by people who think Comic Sans is a cool font and pink on red makes for a nice colour-scheme.



  • @Monomelodies said:

    That's why it's our job to explain to them why they can't have it, period. You can't have both pixel-perfect and a UI which allows the user to arbitrarily resize their fonts, to give an example.

    Or we could just provide them the ability to do what they want and stop trying to cram religious preferences down their throats.

     

    @Monomelodies said:

    Gee, I'm shaking. I'll just ignore the part where you try to assert your seniority by resorting to mentioning $$$, mkay?

    I find it increasingly difficult to take you seriously or care what you have to say.  You said you'd be interested in seeing "sites" I've worked on and I explained that 1) I don't work on "sites" and 2) you have to pay-to-play.  I fail to see why you think any of that is bragging or invalidates my claims in any way.

     

    @Monomelodies said:

    But that's something else than claiming "FF has as many rendering bugs as IE6", which was basically what you were doing.

    Seriously, stop making shit up.  I never said that, I merely said that you have to work around bugs in all browsers and that people who whine about IE are usually lazy or not particularly good at web development.

     

    @Monomelodies said:

    Who mentioned IE5?

    I should have said "IE4", probably.  Clearly I was saying that IE was the best browser for many years and didn't have any real competition until FF in 2004.

     

    @Monomelodies said:

    I've been using good ole' Mozilla since, I dunno, 2002 or so.

    Mozilla was not in wide enough use by 2002 to count for anything.  Additionally, it barfed on so many pages that I find it pretty ridiculous to claim that it was a better browser than IE.

     

    @Monomelodies said:

    Ultimately it shouldn't be necessary, that's what standards are for. Anyway, conditional comments allow (for IE) some pretty strict control over versions. If I target my IE7-fixes at IE7 and IE8 doesn't need them, well, all will be well (in theory). Of course, that's assuming MS doesn't break it by parsing IE7 comments "so the web won't break".

    Why are you incapable of listening to what I am saying?  Standards do not provide for this wonderful level of compatibility that you seem to think they do.  What's more, they do not result in any sort of innovation.  The conditional comments are fine, so long as future versions don't have their own quirks.  Of course, future versions of any browser inevitably will, so the only choice is to test with new versions before they come out and make adjustments then.  I have not argued against this methodology at all -- you have.  Now you've taken a different tack and are claiming this is the way to go.  My point all along has been that conditional comments and UA detection both suffer the same limitations and that I would prefer to do my modifications server-side.  Additionally, it doesn't require any new functionality in non-IE browsers.

     

    @Monomelodies said:

    What I am saying is that when properly coded, it should degrade gracefully.

    Simple stuff will, complex stuff won't.  Anything else is a ridiculous delusion.  I've said this many times and you keep reiterating your point without providing any reason for it.  Saying "if it's done right it should degrade gracefully" 100 times doesn't mean that a complex site can actually be done that way.

     

    @Monomelodies said:

    Though you might want to consider whether or not you've chosen the right platform in that case.

    Agreed.  However, this is the ball of shit we are stuck with.  HTML and CSS are good for basic, text-rich documents and nothing else.  The problem is that we've been stuck building more complicated interactive software with this mess.  It doesn't help that there are many apologists like yourself.

     

    @Monomelodies said:

    Or disabled users?

    @Monomelodies said:

    There are many reasons why some people prefer text-interfaces. Screen reader software prefers them. Maybe I'm blind or extremely short-sighted? Or maybe I prefer as-is content to sucky marckup generated by people who think Comic Sans is a cool font and pink on red makes for a nice colour-scheme.

    How about a more powerful development platform that allows for the creation of accessible software?  This is a solved problem.  Windows has had it for years and nobody had to write their whole interface in markup language and then hack at it with CSS and JS to get it to work.



  • @morbiuswilters said:

    My point all along has been that conditional comments and UA detection both suffer the same limitations and that I would prefer to do my modifications server-side. 

    Which is almost, but not entirely correct. If, for example, I use Firefox and alter my UA string to look like IE7 to see some stupid webpage that refuses to let me in otherwise, then surf somewhere else without remembering to change it back, conditional comments on that page won't trigger just because my UA says I'm an IE browser. In other words, a conditional comment won't be handled unless the condition set really is met, because nothing can be done (short of hacking the rendering engine, that is) to have the browser act to conditional comments specifying a different version.

    You're right that you still don't avoid having to check and likely change stuff when newer browser versions come out, but at least you know that whenever a given browser+version sees your site, it will render the same regardless of any settings. With UA sniffing, you can't know this, but apparently, IE is the only browser willing to admit this may be a problem. It is, of course, also superior to CSS hacks (Holly hack, etc.), since it removes the imaginary link between parsing bugs and rendering bugs - but I believe you already know this.

     Someone mentioned IE/Mac: I'm not sure if that one supported conditional comments of any kind, but that browser isn't really one you'd be supporting these days anyway.



  • @Pidgeot said:

    Someone mentioned IE/Mac: I'm not sure if that one supported conditional comments of any kind, but that browser isn't really one you'd be supporting these days anyway.

    It doesn't, and I agree on not supporting it :) Having said that, I know a place stuffed with graphic designers working on old Macs (yeah, TRWTF). Guess what I had to support... ;)



  • @morbiuswilters said:

    Or we could just provide them the ability to do what they want and stop trying to cram religious preferences down their throats.

    *Sigh* There's nothing religious about it - the web is UA-independent. If you can't deal with that, move to another platform.

    @morbiuswilters said:

    I find it increasingly difficult to take you seriously or care what you have to say.  You said you'd be interested in seeing "sites" I've worked on and I explained that 1) I don't work on "sites" and 2) you have to pay-to-play.  I fail to see why you think any of that is bragging or invalidates my claims in any way.

    I agree it invalidates nothing. In fact, it's completely irrelevant as far as I'm concerned. Also, I apologise for assuming (obviously falsly) that you were somehow involved in web development.

    @morbiuswilters said:

    Seriously, stop making shit up.  I never said that, I merely said that you have to work around bugs in all browsers and that people who whine about IE are usually lazy or not particularly good at web development.

    Allow me to quote you from an earlier post:

    @morbiuswilters said:

    IE6 is a bit more quirky than new versions of FF

    See, it might just be my imagination here, but the phrasing "a bit" implies you think there's hardly any difference. Of course, you're by all means allowed to think that, but I'm pretty sure I'm not the only one who'll disagree with you.

    @morbiuswilters said:

    Mozilla was not in wide enough use by 2002 to count for anything.  Additionally, it barfed on so many pages that I find it pretty ridiculous to claim that it was a better browser than IE.

    We're running in circles. You say IE4 or 5 was the best, I point to something else that was indeed (at least in my opinion) a lot better already way back when, and you counter by saying its market share was small. So true, but we're not talking market share, we're talking quality. On a side note, the only pages it generally barfed on were those relying on IE-specific rendering bugs/features.

    @morbiuswilters said:

    Why are you incapable of listening to what I am saying?  Standards do not provide for this wonderful level of compatibility that you seem to think they do.

    That is your opinion, and I disagree. I hear you alright, only saying something often enough doesn't somehow validate it.

    @morbiuswilters said:

    Simple stuff will, complex stuff won't.  Anything else is a ridiculous delusion.  I've said this many times and you keep reiterating your point without providing any reason for it.  Saying "if it's done right it should degrade gracefully" 100 times doesn't mean that a complex site can actually be done that way.

    Anything "can" be done, whether or not it's worth the hassle is something you should decide on a per-project basis. E.g., if you're developing for the government over here accessibility is actually a requirement. OTOH if you're building some web-based photo application you might decide not to cater for the visually impaired for, well, obvious reasons. (You probably still want to make sure spiders can access public parts of it, though.)

    @morbiuswilters said:

    Agreed.  However, this is the ball of shit we are stuck with.  HTML and CSS are good for basic, text-rich documents and nothing else.  The problem is that we've been stuck building more complicated interactive software with this mess.  It doesn't help that there are many apologists like yourself.

    Ah, well - now we're really getting somewhere. HTML and CSS are actually *shock* meant for basic, text-rich documents. If you're building complicated interactive software, use something else. Like Java, that was sort of invented for that.

    @morbiuswilters said:

    How about a more powerful development platform that allows for the creation of accessible software?  This is a solved problem.  Windows has had it for years and nobody had to write their whole interface in markup language and then hack at it with CSS and JS to get it to work.

    I think we can end this discussion by agreeing that TRWTF is people building complex applications in HTML/CSS/Javascript. It's like writing a 3D-shooter in Visual Basic - you'd probably be able to pull it off, but there are tools and platforms much better suited to that task.



  • @Pidgeot said:

    Which is almost, but not entirely correct. If, for example, I use Firefox and alter my UA string to look like IE7 to see some stupid webpage that refuses to let me in otherwise, then surf somewhere else without remembering to change it back, conditional comments on that page won't trigger just because my UA says I'm an IE browser. In other words, a conditional comment won't be handled unless the condition set really is met, because nothing can be done (short of hacking the rendering engine, that is) to have the browser act to conditional comments specifying a different version.

    If FF implemented conditional comments today, there would an extension out tomorrow that would let you set the comments FF obeys.  Some people are driven to play with the guts of the system, just for the hell of it.  In and of itself that is not a bad thing at all, but when these technically-inclined users start modifying things and then complaining about how other programmers weren't able to anticipate (or simply did not care about, as in my case) their modifications it gets on my nerves.  I do all kinds of digging through my software but I also realize that I will encounter unexpected behavior and I'm not about to start crying about how the sites I am visiting should not be doing things the way they are.



  • @Monomelodies said:

    *Sigh* There's nothing religious about it - the web is UA-independent. If you can't deal with that, move to another platform.

    HTML is UA-independent and that is fine.  The web encompasses many technologies other than HTML.  I'd be glad to see it all replaced with something better.

     

    @Monomelodies said:

    Also, I apologise for assuming (obviously falsly) that you were somehow involved in web development.

    I consider developing software for the web "web development", but it doesn't really matter.

     

    @Monomelodies said:

    Allow me to quote you from an earlier post:

    @morbiuswilters said:

    IE6 is a bit more quirky than new versions of FF

    See, it might just be my imagination here, but the phrasing "a bit" implies you think there's hardly any difference. Of course, you're by all means allowed to think that, but I'm pretty sure I'm not the only one who'll disagree with you.

    I think "a bit" was appropriate there.  IE6 is definitely more screwy than FF, but in the grand scheme of things I don't consider it all that bad.  Then again, I'm comparing it relative to standards-based technologies in general.  Generally the gap between products that supposedly adhere to a standard is much wider than the gap between IE6 and FF.

     

    @Monomelodies said:

    You say IE4 or 5 was the best, I point to something else that was indeed (at least in my opinion) a lot better already way back when, and you counter by saying its market share was small. So true, but we're not talking market share, we're talking quality.

    Market share matters, though.  The problem is that you assume quality is some absolute that Mozilla possessed more of.  My point is that quality is relative: when you have a defacto standard like IE that 98% of the market is using and 98% of developers are coding for, adherence to an arbitrary standard is not worth much.

     

    @Monomelodies said:

    On a side note, the only pages it generally barfed on were those relying on IE-specific rendering bugs/features.

    I remember it barfing a lot on standards-compliant pages, too.  Mozilla was very buggy prior to mid-2003 and bloated and slow until Firefox came out as an alternative.

     

    @Monomelodies said:

    Ah, well - now we're really getting somewhere. HTML and CSS are actually *shock* meant for basic, text-rich documents.

    I know, except it's not really great at that either.  I mean, it's okay, but it contains a lot of extraneous bullshit like forms that would be better handled by a real software development platform.  What's more, the unfortunate trend is towards web-based UIs, simply because the central control afforded and the ability to use it from virtually anywhere outweigh the cons of developing applications in a markup language.

     

    @Monomelodies said:

    If you're building complicated interactive software, use something else. Like Java, that was sort of invented for that.

    Once again, I'd love to, and hopefully Flash and Silverlight will fill the void.  Java applets never caught on and certainly will not now.

     

    @Monomelodies said:

    I think we can end this discussion by agreeing that TRWTF is people building complex applications in HTML/CSS/Javascript. It's like writing a 3D-shooter in Visual Basic - you'd probably be able to pull it off, but there are tools and platforms much better suited to that task.

    Agreed, but we are stuck with HTML for now because of how history has unfolded.  Personally, I'd rather HTML and CSS be stripped back until they are simply good document formatting languages and having something else fill the void(s) for rich UIs and multimedia-heavy content.  Right now Flash is doing better and better, but it's still a rough patch for the web.



  • @morbiuswilters said:

    I do all kinds of digging through my software but I also realize that I will encounter unexpected behavior and I'm not about to start crying about how the sites I am visiting should not be doing things the way they are.
    Agreed. I'm one of the 1% using Opera and when I have it mask or identify as IE or FF, I know the website may not behave as intended. I'd prefer not having to resort to deceit, though. Instead of denying access with Opera, just warn me and let me in at my own risk.
    And don't get me started on those websites that are targeted at a [i]specific[/i] browser version and deny access because your browser version is [i]too high[/i], most of the time adding insult to injury by claiming your browser is too old.



  • @morbiuswilters said:

    HTML is UA-independent and that is fine.  The web encompasses many technologies other than HTML.  I'd be glad to see it all replaced with something better.

    Well, can't argue with something better. But that something will also have to be properly specified and platform-indepedent, so you're stuck in the same situation basically (apart from the "clean break" advantage).

    @morbiuswilters said:

    Market share matters, though.  The problem is that you assume quality is some absolute that Mozilla possessed more of.  My point is that quality is relative: when you have a defacto standard like IE that 98% of the market is using and 98% of developers are coding for, adherence to an arbitrary standard is not worth much.

    I partly follow you there, but I'm talking more about IE(6, especially) just not doing what it should do. And that's not just question of following standards or making your own instead, but text falling off the screen for no particular reason. That's not a matter of dispute, that's just plain buggy. Of course other products have also been buggy in the past. I can't really recall when Mozilla started shaping up - you're probably right saying that was mid-03. Btw, I quickly switched to Phoenix (FF's predecessor) which was a sort of light-weight Mozilla fork. At least got rid of all the bloat...

    @morbiuswilters said:

    I know, except it's not really great at that either.  I mean, it's okay, but it contains a lot of extraneous bullshit like forms that would be better handled by a real software development platform.  What's more, the unfortunate trend is towards web-based UIs, simply because the central control afforded and the ability to use it from virtually anywhere outweigh the cons of developing applications in a markup language.

    An interesting point. You previously stated the fact that Windows gives you all that interface on a platter as a big plus. But somehow when HTML does that it's a problem? Care to elaborate? (I'm not trying to rattle you BTW, I'm seriously interested in peoples' point of view on these kinds of things, since I'm obviously not all-knowing and might hear a good enough argument to change my own.)

    HTML forms do leave a little to be desired, but in all they're okay for most tasks I think. They are a bit of an odd one out in the standard though, I agree on that.

    @morbiuswilters said:

    Once again, I'd love to, and hopefully Flash and Silverlight will fill the void.  Java applets never caught on and certainly will not now.

    If those are the kinds of applications you're writing Silverlight is looking good. Only downside is it's not platform-independent (and no, Mono/Moonlight doesn't count ;-)). If Microsoft swallows their pride and actually open up (or support other platforms themselves, though I'd leave that to communities personally) they could really have a winner.

    @morbiuswilters said:

    Agreed, but we are stuck with HTML for now because of how history has unfolded.  Personally, I'd rather HTML and CSS be stripped back until they are simply good document formatting languages and having something else fill the void(s) for rich UIs and multimedia-heavy content.  Right now Flash is doing better and better, but it's still a rough patch for the web.

    If SaaS takes off things will prolly speed up in that respect :) I don't think anything needs to be stripped tho; just have your Silverlight or Flex or whatever ends up winning that "fight" and use it for software ;)



  • @Monomelodies said:

    Btw, I quickly switched to Phoenix (FF's predecessor) which was a sort of light-weight Mozilla fork.

    Phoenix is Firefox. Mozilla had to change the name, twice, due to choosing names already used by other organisations. Scarily enough "Firefox" isn't a new word either, for example there's a 1978 novel of the same name and an 80s film based on it. There was also an unrelated UK Web site called Firefox that didn't actually render in Mozilla Firefox...



  • @fatdog said:

    I don't know but all the stats of around 25 sites I have to maintain daily say that more than 80% still use IE. So honestly I won't rank both browser, because stats are doing it for me.

    Are your stats based on the provided User Agent strings, or are you detecting browser fingerprints? If it's the former, I'd say your stats are rather inaccurate.

    Most notable is that it used to be something like 90+% of Opera users apparently have their UA set to IE, due to a number of sites which provided broken CSS files to Opera users. (I don't know if this practice is still used, but I have certainly seen a lot of discussion about it, and years ago I verified that it was at the time being practiced.) In response, Opera made it very easy for its users to change the UA string - but as of four years ago (if memory serves me correctly), they had no reminder to change it back when they went to a different site.



  • @tgape said:

    Are your stats based on the provided User Agent strings, or are you detecting browser fingerprints? If it's the former, I'd say your stats are rather inaccurate.

    Most notable is that it used to be something like 90+% of Opera users apparently have their UA set to IE, due to a number of sites which provided broken CSS files to Opera users. (I don't know if this practice is still used, but I have certainly seen a lot of discussion about it, and years ago I verified that it was at the time being practiced.) In response, Opera made it very easy for its users to change the UA string - but as of four years ago (if memory serves me correctly), they had no reminder to change it back when they went to a different site.

    I've heard this nonsense before.  The number of Opera users is so small that even those masked to IE aren't worth counting. 



  • @morbiuswilters said:

    @anthetos said:

    Or they could just learn how the web works and not have this problem or anything like it...

    How precisely would that be?  Hell, why even bother?  It's not worth the time or effort to take Linux into account for 99% of sites.  You might as well worry about Opera and Lynx while you're at it. 

     

     Well, if you provide basic acessibility, Lynx support will come right along...most of the time, Lynx breaking goes back to complex layouts or  non-alted images that would also break a screenreader. I must say, I do appreciate it whenever someone seems to still pay attention to Lynx users... 



  • @morbiuswilters said:

    I agree for the most part about Flash being pretty mediocre. What we have to realize is that there are several different uses for the web and that HTML/CSS is not particularly good for most. Sure, it's fine for simple sites that are text-heavy with a little bit of "pretty". However, when it comes to complex applications, media-rich sites or sites that aim for a very specific artistic layout, HTML and CSS fall flat. Of course, we've had the ability to do all of these for awhile, but unfortunately real application and media technologies for the web like Java, ActiveX, Flash, etc.. ended up being technological abortions and now we're stuck in some kind of purgatory where we have to construct our own drag-and-drop, rich GUI interfaces using divs and CSS hacks all bound into a messy ball with Javascript.

    QFT. That's a pretty good summary.

    @morbiuswilters said:

    I never particularly liked that feature and always turned it off, but I don't think it was all that bad and most people seemed quite fine with it.

    It's a combination of two features: placing buttons of the same program side by side, and stacking windows into the same button. The former is really useful, the latter is a pain. I've noticed that TweakUI lets you configure the threshold for when buttons are combined, allowing you to have the good feature and stave off the bad feature (even easier if you're one of the few who realise you can extend the taskbar to multiple rows.

    @morbiuswilters said:

    Also, I must point out that it was IE that gave us XMLHttpRequest, without which most of this Web 2.0 stuff wouldn't even be possible. I'm not actually convinced that it was a good thing, but I think claiming a lack of innovation on the part of MS is a tad unfair. Other than being better at adhering to an arbitrary standard, what amazing advances has Firefox given us? Tabs? Extensions? Cool stuff, but hardly life-changing.

    The most obvious one is security. It doesn't matter how clever or not IE is, it's still a menace. For all the more enlightened people forced to use Windows, at least Firefox and Opera mean that they can do so safely!

    The term "life-changing" is a hard one to qualify, as you know, but my perspective on IT is that computers are far too insufferable, which is wonderful since there's virtually no escaping them any more. Anything that we can do to make the experience less painful, is worth the effort. Scarily enough, I used to be happy with IE 5, albeit with my taskbar on three rows to make room for the number of windows I ended up having open. I don't know how I coped! I can't imagine going back to the Web without Firefox, without the ability to take control the way the untiring efforts of extension developers allow us to.

    Actually, the browser that was well ahead of the game for a long time was iCab for Macintosh, which the core Firefox is finally overtaking. Firefox 3's "awesomebar" is the first mainstream Windows browser address bar to actually beat iCab's. Internet Explorer for Macintosh was also ahead with similar features (queuing download manager with resuming downloads and page title search from the address bar) although iCab was far more stable, had address bar query string generation, complex page plugin and content filter control and uncountable little features like per-page and per-image animated GIF stop/start and per-image inline reload. UA switching was of course built-in and, I think, configurable per site.

    It took a long time before I got Firefox anywhere close to being as controllable as iCab and I missed iCab so terribly. However, Firefox is extensible and iCab is not so iCab ultimately cannot compete, and iCab has fallen behind. It was infamous for its proprietary rendering engine but the developer threw in the towel and moved to the lamentable WebKit.

    @morbiuswilters said:

    Welcome to Community Server. Get used to it.

    I really thought I was :)



  • @Daniel Beardsmore said:

    The most obvious one is security. It doesn't matter how clever or not IE is, it's still a menace.

    True, but I have a feeling security on FF and Opera will continue to decline.  The addition of new, complex features, more powerful scripting and the increasing market share of FF will make it a bigger target in the future.

     

    @Daniel Beardsmore said:

    I really thought I was :)

    Community Server's resourcefulness in continuously finding new ways to fuck things up is awe-inspiring.



  • @morbiuswilters said:

    True, but I have a feeling security on FF and Opera will continue to decline.  The addition of new, complex features, more powerful scripting and the increasing market share of FF will make it a bigger target in the future.

    A common and understandable mistake: IE isn't swiss cheese /because/ it's so wide-spread. It's swiss cheese, and becuase it's wide-spread every hole gets exploited. It's a logical fallacy to assume that thus every wide-spread application would also resemble swiss cheese.

    Of course, more bugs will be found if a product is in wide use. But since not every product out there is deeply ingrained in the underlying operating system the risks are usually negligeable.

    @morbiuswilters said:

    Community Server's resourcefulness in continuously finding new ways to fuck things up is awe-inspiring.

    Ah, now that's interesting. You stated earlier that CS was an example of a complex application that couldn't be browser-independent and gracefully degrading [presumably because of its "complexity"].

    Aw, c'mon. CS is bleeding FORUM SOFTWARE. The fact that they can't get it right only goes to show they're amateurs, not that it can't be /done/ right. On a related note, the "real world" you referred to earlier isn't a controlled environment where only people paying mucho $$$ can access your app (in that case it's easy to state browser X as a requirement). The Real World is an app in the wild that Joe N00bAverage can use. Yes, that means catering for IE, obviously. But it also means catering for other clients. Let's say Firefox takes 5% (a *very* conservative estimate). If your userbase is 100, that's 2 people. Ignore them all you like. If your userbase is 10.000, there's 200 people complaining stuff doesn't work. I'm getting tired of answering mails now :) If it's a million, 20.000 people are shut out. When would marketing start to complain about missing revenue?



  • @Monomelodies said:

    A common and understandable mistake: IE isn't swiss cheese /because/ it's so wide-spread. It's swiss cheese, and becuase it's wide-spread every hole gets exploited. It's a logical fallacy to assume that thus every wide-spread application would also resemble swiss cheese.

    Of course, more bugs will be found if a product is in wide use. But since not every product out there is deeply ingrained in the underlying operating system the risks are usually negligeable.

    [citation needed]  IE7 was decoupled from the core OS fairly significantly.  Additionally, last time I checked FF had far more exploits than IE.  IE has had a pretty good security record for about 2 years now.


    @Monomelodies said:

    Let's say Firefox takes 5% (a *very* conservative estimate). If your userbase is 100, that's 2 people. Ignore them all you like. If your userbase is 10.000, there's 200 people complaining stuff doesn't work. I'm getting tired of answering mails now :) If it's a million, 20.000 people are shut out.

    lern2emailfilter

     

    @Monomelodies said:

    When would marketing start to complain about missing revenue?

    When would management do a cost-benefit analysis and determine supporting a negligable part of the market is not profitable.



  • I do use Linux as my desktop OS, and I use Windows at work, and I can say pretty clearly, that Javascript/AJAX/etc. support in Firefox for anything that does not require a plugin is identical (its the same codebase, all the javascript engine code is cross-platform compatible)

     Now, I'm not going to pretend and delude myself that most businesses care about making their page work in Linux based web browsers. I would however think that businesses would care abotu firefox compability, simply because firefox's usage ranges anywhere between 10%-30% of current web browser usage. At worst, one in every tenth person uses firefox, at best, that goes up to every 1/3-ish. If I was a business with a few hundred thousand hits per day, I know I don't want to loose a potential 1/10-1/3rd of my userbase just because they don't use IE.

    Forcibly checking a UA string to exclude based on platform if a site is using just javascript and html, and no platform specific features (which you likely have to go out of your way to use in Firefox; I'm not even aware of any off hand) just shows poor coding or at least, a dangerous unawareness by the developer.

    When you take into account that a vast majority of people using the internet only use IE because that's what comes on the computer, and just thinks IE is "The Internet" and not a program for rendering web documents, I consider those statistics pretty good.

     As for flash, I don't have a problem with flash in general (aside from the fact that the current implementation at times is extremely slow and buggy, although Flash 10 on both Windows and Linux seems to be a vast improve). I do however feel that a single company with a properiety product like flash should be taken with caution. Say what will you will about IE or Firefox, but no one can argue that you have to pay or reverse enginneer binary formats to create a web page. Given the fact that SWF spec files until very recently were closed (and the current "open" SWF specs still lack parts to allow someone else beside adobe from implementing the spec fully), I would only use flash if there was no other realstic alternative available. Unfortanately, for many use cases, there is no good alternative.



     



  • @morbiuswilters said:

    [citation needed]  IE7 was decoupled from the core OS fairly significantly.

    Eh? No it hasn't been, whoever told you that? Can you deinstall IE7 from a Vista box? I must admit I've managed to ignore Vista like the plague, so perhaps its Explorer app doesn't actually use IE anymore. Then again, who uses Vista? Then again, you apparently can't install IE6 on Vista (way to cater for web developers!) so something's fishy at the very least.

    @morbiuswilters said:

    Additionally, last time I checked FF had far more exploits than IE.  IE has had a pretty good security record for about 2 years now.

    Maybe you're checking in the wrong place? I'd have to check if it's still possible, but I remember IE7 "supporting" cross-site Ajax scripting as a "feature". Things like that don't exactly instill me with trust in the product. Then there's ActiveX with regular gaping holes. Sure, the fact that IE runs sort-of sandboxed on Vista helps a lot, but it's a little sad it took MS about 10 years to get this sort-of right, no?

    @morbiuswilters said:

    lern2emailfilter

    Say what?

    @morbiuswilters said:

    When would management do a cost-benefit analysis and determine supporting a negligable part of the market is not profitable.
     

    Management won't consider 20.000 potential paying customers as "negligable", even though you might. If Opera were vastly different than everything else out there then yes, supporting it would be madness except for very large user-bases or very specific cases (e.g., the government or large banks). Only Opera isn't vastly different, of course, so supporting it for 99.99% isn't an issue for any web-dev worth his/her salt.



  • @morbiuswilters said:

    @anthetos said:

    Or they could just learn how the web works and not have this problem or anything like it...

    How precisely would that be?  Hell, why even bother?  It's not worth the time or effort to take Linux into account for 99% of sites.  You might as well worry about Opera and Lynx while you're at it. 

    My personal website is designed to work on Opera and Lynx on Linux. Guess what? As a side effect of targeting those two browsers, it also works with everything else.



  • @Monomelodies said:

    Let's say Firefox takes 5% (a *very* conservative estimate). If your userbase is 100, that's 2 people.
     

    Lern2DoMaths 



  • Ehm... they're fine, thank you...your point being?



  • @Carnildo said:

    My personal website is designed to work on Opera and Lynx on Linux. Guess what? As a side effect of targeting those two browsers, it also works with everything else.

    That's great for you.  In the real world, this is frequently not the case.  Also, designing to work on Lynx is pretty pointless. 



  • @Monomelodies said:

    Ehm... they're fine, thank you...your point being?

    5% of 100 is 5.  That's what he was pointing out.

     

    TRWTF is that upsidedowncreature got the meme wrong.  It should have been "lern2math".  All lower-case and verbs are always omitted. 



  •  Ah wait, I somehow got 5% vs. 1/5 mixed up in my head. Prolly cause firefox does closer to 1/5th than 5% ;-) I stand corrected! It only strengthens my point, though...



  • @morbiuswilters said:

    That's great for you.  In the real world, this is frequently not the case.  Also, designing to work on Lynx is pretty pointless. 

    Well, as I stated before, in the real world it usually pays to cater for as many people as possible, and designing cross-browser accomplishes just that most of the time. That's his (and my) point. Your example of an expensive - presumably more-or-less intranet - application simply isn't "the real world". The real world is random users, and they use weird shit. I had a support ticket from a Win98 user today, for instance.



  • @Monomelodies said:


    Let's say Firefox takes 5% (a *very* conservative estimate). If your userbase is 100, that's 2 people.

    @morbiuswilters said:

    5% of 100 is 5.

    @Monomelodies said:

     Ah wait, I somehow got 5% vs. 1/5 mixed up in my head. Prolly cause firefox does closer to 1/5th than 5% ;-) I stand corrected! It only strengthens my point, though...

    1/5 of how many users makes 2 people?



  • @fatdog said:

    @Monomelodies said:


    Let's say Firefox takes 5% (a *very* conservative estimate). If your userbase is 100, that's 2 people.

    @morbiuswilters said:

    5% of 100 is 5.

    @Monomelodies said:

     Ah wait, I somehow got 5% vs. 1/5 mixed up in my head. Prolly cause firefox does closer to 1/5th than 5% ;-) I stand corrected! It only strengthens my point, though...

    1/5 of how many users makes 2 people?

    Ok, who's using the RNG as a calculator?



  • @danixdefcon5 said:

    Ok, who's using the RNG as a calculator?

    Probably mistook the OMGWTF calculators for serious software.



  • @Daniel Beardsmore said:

    @danixdefcon5 said:

    Ok, who's using the RNG as a calculator?

    Probably mistook the OMGWTF calculators for serious software.

    Funny, but inaccurate.  The OMGWTF calculators had to actually return the right results, they just had to work in a very convoluted, WTFy way.

     

    Besides, if you need a calculator to find 5% or 1/5th of 100 then you are TRWTF.



  • @morbiuswilters said:

    Funny, but inaccurate.  The OMGWTF calculators had to actually return the right results, they just had to work in a very convoluted, WTFy way.

    From TFA:

    Note that [the above list of specific calculations] will be the only test cases run against your calculator ... If it passes all of these, then congratulations, you’ve built a fully-functioning calculator so far as we’re concerned.

    None of the tests actually check whether the calculator is able to calculate either 5% of 100, or 15 × 100. As such, they're quite free to return 2 if they so desire.



  • @Daniel Beardsmore said:

    From TFA:

    Note that [the above list of specific calculations] will be the only test cases run against your calculator ... If it passes all of these, then congratulations, you’ve built a fully-functioning calculator so far as we’re concerned.

    None of the tests actually check whether the calculator is able to calculate either 5% of 100, or 15 × 100. As such, they're quite free to return 2 if they so desire.

    Good point, I forgot that the contest did not use open-ended test cases. 



  • @Carnildo said:

    My personal website is designed to work on Opera and Lynx on Linux. Guess what? As a side effect of targeting those two browsers, it also works with everything else.
     

    Except IE. At least use links rather than lynx :) but it doesn't support much interesting (fonts, images, javascript) anyway.

    I used to develop for Opera on Linux, (been using Opera since v3.21 on my 486) but have been swayed by Firefox in the last couple of years. All the extra addons and features make it better than anything else.


Log in to reply