Great Moments in CSS
-
Have you tried Wolfram Alpha? It is the greatest system ever! It is a total revolution in whatever the heck it does. The theory and coding behind it are so great that there is no need to make it look usable. Because it's all done with CSS!
-
Try using a browser
that isn't 8 years old.
-
Eek. Hard to believe someone who enjoys computing-based WTFs can in all honesty not only still be using IE 6, but actually consider it a WTF when something standards based does not work in it.
Well, there ia a WTF, but it's in the poster using such an archaic PoS, and expecting it to work.
-
We stopped supporting IE6 after IE8 came out. You want it => pay more monies
-
@newfweiler said:
The theory and coding behind it are so great that there is no need to make it look usable. Because it's all done with CSS!
zOMG! IT NO WERK N LYNX 2!
-
Doesn't render in Mosaic or Gopher, either. What could they be thinking!!!
-
@heltoupee said:
Doesn't render in Mosaic or Gopher, either. What could they be thinking!!!
Looks great in emacs!
-
@obediah said:
Looks great in emacs!
No it doesn't.
You're just used to how everything else looks in emacs, which is even crappier.
-
@dhromed said:
@obediah said:
Looks great in emacs!
No it doesn't.
You're just used to how everything else looks in emacs, which is even crappier.
I guess adding a few hundred worthless buttons to beef up the pixel count could pretty things up considerably.
You are right about image editing however, emacs could use some help there.
-
TRWTF is you using IE6.
use a standards compliant browser, at least compliant to the point of IE7/8...
i actually specificly won't support IE6 anymore - not even in the best case...
-
If you were running Windows 2000, I [b]might[/b] understand you using IE6, since that's the highest version if IE it supports, but you're running Windows XP so your only possible excuse could be if you're browsing from work and your IT department refuses to upgrade to a non-ancient version of IE (possibly due to an ancient and broken intranet app that only works in IE6).
Interestingly, Wolfram appears to use XHTML 1.0 Strict, and it even validates without any errors. It's nice to see others actively try to comply to standards.
-
@Quietust said:
Interestingly, Wolfram appears to use XHTML 1.0 Strict, and it even validates without any errors. It's nice to see others actively try to comply to standards.
It's served as text/html, though, which means any mainstream browser is just going to treat it like broken HTML and not XML. Fail.
-
Herein lies the inherent disconnect between web developers and system administrators. Web developers don't understand it, so they assume the system administrators have it taken care of. System administrators don't want to trust the web developers to stick to XHTML, so even if they think about it, they don't do it. Wee
By the way, thanks for reminding to reconfigure my new server.
-
According to W3Schools browser stats, 15% of access are still from IE6 -- and that's a site that caters to developers.
(This reminds me of the time I posted a story concerning an IBM 370/145 in 1974, and one of you complained "That's bogus! The 370/195 was available by 1974. Nobody had a 370/145 any more.")
It would be a wonderful world for developers if we could just ignore environments we didn't like.
Well, I think the real WTF is developing applications to run in a browser, period! ("Full stop" if you're British.) Native applications are the only ones that are pleasant to use.
-
@newfweiler said:
According to W3Schools browser stats, 15% of access are still from IE6 -- and that's a site that caters to developers.
As someone that is currently in the process of trying to hire someone, I can tell you that there are a lot of unbelievably shitty and unqualified "Web developers" out there. So, I am not sure it's fair to say "and those are developers".
(Also, caveats, AVG scans ahead in search results and masquerades as IE6, there are a bunch of web crawlers that pretend to be IE6, people set their User-Agent to get around stupid browser detection and forget to change it back, etc.)
-
@newfweiler said:
Well, I think the real WTF is developing applications to run in a browser, period! ("Full stop" if you're British.) Native applications are the only ones that are pleasant to use.
It's a trade-off, clearly. Web apps aren't as pretty as desktop apps and they don't usually have great UIs. However, web apps are inherently centralized which makes them more convenient from an access, backup and distribution standpoint. Web apps can also make use of large server-side resources on otherwise underpowered platforms. Not every app should be made to run in a browser, but the increasing popularity of web apps isn't for nothing.
-
@snover said:
@newfweiler said:
According to W3Schools browser stats, 15% of access are still from IE6 -- and that's a site that caters to developers.
As someone that is currently in the process of trying to hire another Web developer, I can tell you that there are a lot of unbelievably shitty and unqualified people out there that think they know what they are doing. OP might be a good example.
Any web developer who is currently using IE6 for anything other than compatibility testing is TRWTF. I still think it's a good idea to support IE6 (perhaps with slightly degraded functionality) but using it as part of your development platform is retarded.
-
@morbiuswilters said:
I still think it's a good idea to support IE6 (perhaps with slightly degraded functionality) but using it as part of your development platform is retarded.
I don’t know. One the one hand, I certainly understand not wanting to alienate some visitors. On the other hand, at some point, you really have to say "enough is enough" and drop support in order to get people to finally upgrade their shit. (It’s like the digital TV transition in the US. Some people will procrastinate until the end of time unless you break things for them.) Given that I can choose to spend 20-30% less time on a project by not trying to support IE6 (beyond loading Dean Edwards’s IE7 and throwing up a big message that reminds them their Web browser is older than 9/11), I choose to do that. Especially now that there are not two newer major versions available.
(Coincidentally, the reason I’m hiring someone is so that the company can stop outsourcing the rest of their work to a company that basically consists of one guy that writes ColdFusion and one guy that cuts and pastes JavaScript. And yes, they both still use IE6.)
-
I still use IE6. Why? Because it works. I have yet to see a website (with one exception) that doesn't work. The one website? Some stupid webstie that throws up an ENORMOUS banner telling me the website isn't compatible with IE6 and that I should upgrade, despite the fact that the website works just fine with IE6.
-
@chrismcb said:
I still use IE6. Why? Because it works.
You mean, apart from those unimportant little security concerns and the billions of web designers cursing it for not working like it should? I agree.
-
I might understand you using IE6, since that's the highest version if IE it supports
We run Firefox in Windows 2000. There ARE alternatives.
(Remember: the "I" is for "Idiot")
-
You only get that experience because of the millions of development hours wasted trying to support it. Have some compassion, and stop.
-
So ... does CSS suck or is it just CSS implementations that suck?
If CSS is a standard, why do different browsers mess up your page in different ways?
-
@newfweiler said:
Generally speaking, they don't. (barring different browsers having different support levels)So ... does CSS suck or is it just CSS implementations that suck?
If CSS is a standard, why do different browsers mess up your page in different ways?
The main problem is that IE has somewhat insane default styles under certain circumstances, which you have to override by manually setting whatever sane browsers do by default. Oh, and IE's CSS breaks inunexpected ways if you have invalid HTML (particularly missing td's - which Firefox kindly assumes are supposed to be right before that /tr).
-
@Weng said:
@newfweiler said:
Generally speaking, they don't. (barring different browsers having different support levels)So ... does CSS suck or is it just CSS implementations that suck?
If CSS is a standard, why do different browsers mess up your page in different ways?
The main problem is that IE has somewhat insane default styles under certain circumstances, which you have to override by manually setting whatever sane browsers do by default. Oh, and IE's CSS breaks inunexpected ways if you have invalid HTML (particularly missing td's - which Firefox kindly assumes are supposed to be right before that /tr).
I remember back in the olden days Netscape barfed on missing </td>s. The tragedy is that IE6 barfs on pretty much everything these days...
-
@newfweiler said:
Have you tried Wolfram Alpha? It is the greatest system ever! It is a total revolution in whatever the heck it does. The theory and coding behind it are so great that there is no need to make it look usable. Because it's all done with CSS!
There must be something else wrong with your browser. I have IE6 on this computer (to test stuff in, I have IE7 on my laptop and IE8 in a VM) and the site renders almost exactly like in FireFox, except for some minor bugs. It's not nearly as FUBARed as the screenshot you posted.
-
@newfweiler said:
So ... does CSS suck or is it just CSS implementations that suck?
If CSS is a standard, why do different browsers mess up your page in different ways?
Both.
The original CSS 2 specification was actually so full of errors that they ended up replacing it with CSS 2.1 (the list of changes is 81 pages long). That fact doesn’t excuse Microsoft’s complete inability to write a working rendering engine, though, since MSIE (until version 8) didn’t even work consistently with basic stuff that was defined in CSS 1 (like the float and list-style properties), and often broke in interesting and inexplicable ways due to the two different rendering methods they used internally (hasLayout).
The last fun WTF that I dealt with in regards to the CSS specification itself was an incompatible change between CSS 2 and CSS 2.1. In CSS 2, you were allowed to set the overflow property on block-level and replaced elements; in CSS 2.1, they changed the spec so that overflow only works on non-replaced block-level elements… which wouldn’t really be a problem, except that suddenly there is no way for the parent document to control the appearance of scrollbars on embedded content when using any strict DTD or HTML 5*. Except in Firefox, which still implements the CSS 2 way of doing overflow, even in the still-upcoming version 3.1.
Of course, none of this would be much of a problem if ad serving companies were smart enough to set html, body { overflow : hidden; } on their generated pages, but they aren’t. So now we’re serving almost-valid XHTML Strict because the W3C decided that implementing overflow on replaced elements was too hard (or something; I can’t find an actual mention of the reason for the change on lists.w3.org), and because ad companies suck (but we knew that already).
I might be a little bitter.
* In yet another amazing contradiction by the W3C, the HTML 5 draft spec removes the “scrolling” attribute from <iframe> because that behaviour is “better handled by CSS”. Err, right.
-
@pbean said:
@newfweiler said:
Have you tried Wolfram Alpha? It is the greatest system ever! It is a total revolution in whatever the heck it does. The theory and coding behind it are so great that there is no need to make it look usable. Because it's all done with CSS!
There must be something else wrong with your browser. I have IE6 on this computer (to test stuff in, I have IE7 on my laptop and IE8 in a VM) and the site renders almost exactly like in FireFox, except for some minor bugs. It's not nearly as FUBARed as the screenshot you posted.
Maybe there is. Many three-column CSS layouts render strangely, with column 3 under columns 1 and 2. http://onstartups.com/ is a two-column layout that renders with column two under column 1.
It seems to me that this three-column thing is a hack. There's nothing in CSS that specifically means three-column layout, but there's ways of using "float" that sometimes work. Isn't "float" really for something else, viz. wrapping text around a picture?
Given three div's, how do you say in CSS "put this one on the left, this one in the middle, and this one on the right?" How about with 4? i.e. how do you do exactly what <table> does, but without using tables?
-
@newfweiler said:
It seems to me that this three-column thing is a hack. There's nothing in CSS that specifically means three-column layout, but there's ways of using "float" that sometimes work. Isn't "float" really for something else, viz. wrapping text around a picture?
Nothing at all. :)
@newfweiler said:
Given three div's, how do you say in CSS "put this one on the left, this one in the middle, and this one on the right?" How about with 4? i.e. how do you do exactly what <table> does, but without using tables?
- float : left + margin
- display : table
- display : inline-block
- position : absolute
- css3 multi-col
There are probably some other options that I haven’t thought of off the top of my head. Of course, don’t expect any of these to work properly in IE6 without needing to hack around its complete inability to do anything properly. #1 will fail because IE6 has a bug that causes margins on floats to become doubled. #2 will fail because IE<8 doesn’t support display : table. #3 will fail because IE<8 only supports inline-block on elements that are normally already inline. #4 will fail because IE6 doesn’t know how to calculate the width of an element when given both left and right properties. #5 will fail because IE doesn’t support any CSS3.
Are you maybe starting to understand why you shouldn’t use IE6 yet?
-
Personally I'm quite fond of method #6: Shoot the fucking designer that thinks columns are a good idea.
-
@morbiuswilters said:
Any web developer who is currently using IE6 for anything other than compatibility testing is TRWTF. I still think it's a good idea to support IE6 (perhaps with slightly degraded functionality) but using it as part of your development platform is retarded.
@snover said:
I don’t know. One the one hand, I certainly understand not wanting to alienate some visitors. On the other hand, at some point, you really have to say "enough is enough" and drop support in order to get people to finally upgrade their shit.
That wasn't Morb's point.
Morb referred to development platform, not the final user base. And he's fully, completely and absolutely right. You do not use IE6 as your primary development browser. You use either Firefox+Firebug+[Webdevtoolbar]+[whathaveyou]. I suppose Webkit's dev thing is also adequate for the task. At least they all support console.log, unlike IE8; wtf Microsoft.
For the record, Snover, you are also right. Testing for IE6 is desirable, mostly because of people like
chrismcb</b>, who do not care about how much extra time we need to make out shit work in IE6.</p><p>Nor do they, apparently, care about (or fully realise the import of) the extra better features that newer browsers offer them (tabs. yay tabs). That's to be forgiven, of course, because it's just preference. I can't live without Photoshop CS3, but surely my mom has enough with, say, 6 or 7. <br></p>
-
@newfweiler said:
So ... does CSS suck or is it just CSS implementations that suck?
CSS is pretty good, but like any language it has its wtfs. CSS's wtfs are inspired by the ivory tower syndrome of the W3C.
@newfweiler said:
If CSS is a standard, why do different browsers mess up your page in different ways?
Because it's hard to write an engine, and parts of the specification are actually up for subjective interpretation, and the spec also leaves some things up to the browser maker. Also, there's a long tradition of "engine feature war", rather than one of "browser feature war with a standard engine" as is growing today.
But it's coming along okay. Stuff is maturing.
Web app philosophy is another can of worms entirely. If we really want online apps (such as gmail today, and in the future, something like online Photoshop) , we need something other than a browser to do it. The current heap of languages and systems that make up the web today is not a pretty synergy of technologies -- it's in fact a cancerous blob of a browser that was originally a document viewer.
-
So what is the best platform to use right now for client-side apps? Flash / Flex? Silverlight? A Java applet in the browser?
-
HTTP POST forms.
-
@snover said:
@newfweiler said:
It seems to me that this three-column thing is a hack. There's nothing in CSS that specifically means three-column layout, but there's ways of using "float" that sometimes work. Isn't "float" really for something else, viz. wrapping text around a picture?
Nothing at all. :)
Oh wonderful! And when pray tell will all browsers be CSS-3 compliant? What CSS-3 browser do YOU use?
-
@dhromed said:
We stopped supporting IE6 after IE8 came out. You want it => pay more monies
lucky you ... some of our customers (corporate) stilll have it installed ... with custom security settings ... and they want their pages to look nice ...
-
@dhromed said:
something like online Photoshop
something like :
Pixlr Editor: Free Advanced Photo Editor & Image Editing Tool
Free online photo editor with advanced image editing features. Perfect for both beginner & professional. Start editing with Pixlr editor now!
or
-
@Nelle said:
@dhromed said:
something like online Photoshop
something like :
Pixlr Editor: Free Advanced Photo Editor & Image Editing Tool
Free online photo editor with advanced image editing features. Perfect for both beginner & professional. Start editing with Pixlr editor now!
or
That's pretty good stuff.
-
@newfweiler said:
Oh wonderful! And when pray tell will all browsers be CSS-3 compliant? What CSS-3 browser do YOU use?
Gecko has supported a majority of css3-multicol since version 1.8 (Firefox 1.5), and it looks like they will have a complete implementation in the upcoming Gecko 1.9.1 (Firefox 3.5). WebKit has supported it since version 522.11 (Safari 3).
Also your comment about a “CSS-3 [sic] compliant” browser seems to indicate a lack of understanding of the underlying architecture of the CSS3 specification. CSS3 is not one large monolithic specification; it consists of multiple, discrete modules that each describe different functionality. So, all browsers may have support for CSS Multi-column Layout, but none may have support for CSS Speech. It doesn’t mean they aren’t CSS3 compliant, it just means that they did not choose to implement certain modules.
(Frankly, the real solution is probably CSS Template Layout, but it’ll be at least 5 years before there’s even a chance of using it. :\)
-
@chrismcb said:
I still use IE6. Why? Because it works.
Wow, your mind is going to be blown in the year 2016 when you discover tabbed browsing. That is, if there's anything left after you discover IDEs in the year 2015 because you refused to stop using vim "because it works."
-
Now reread all the posts above that use "old", "new", "recent", etc. as criteria for using browsers, only for "IE6" substitute "Unix" (much older than IE6 by the way -- how many of you are still using this 40-year-old operating system?) and for the most recent browser substitute "Microsoft Windows Vista Ultimate Edition". Still think new things are better than old things?
Anyway, it sucks to have to use a "standard" with so many versions and so many different implementations. Should Wolfram Alpha support a browser used by only 15% of the population? (Or whatever the figure is -- what cutoff would you accept to justify not supporting IE6?) The way CSS renders differently in different browsers is like having different models of computers where 2 + 2 = 4 in 75% of them, 2 + 2 = 3 in 20% of them, and 2 + 2 = FILE_NOT_FOUND in 5% of them.
-
@snover said:
It doesn’t mean they aren’t CSS3 compliant, it just means that they did not choose to implement certain modules.
Wonderful. Now you have to make a chart for all the versions of all the browsers with entries for what modules they chose to implement and what modules they chose not to implement if you want your page to render correctly for all your users.
-
@newfweiler said:
Now reread all the posts above that use "old", "new", "recent", etc. as criteria for using browsers, only for "IE6" substitute "Unix" (much older than IE6 by the way -- how many of you are still using this 40-year-old operating system?) and for the most recent browser substitute "Microsoft Windows Vista Ultimate Edition". Still think new things are better than old things?
You're right. This is not at all a moronic anology. When I fire up Unix, I'm still running the orginal source on a PDP-7.
-
@newfweiler said:
Wonderful. Now you have to make a chart for all the versions of all the browsers with entries for what modules they chose to implement and what modules they chose not to implement if you want your page to render correctly for all your users.
Huh? How is this any different than what happens today? Do you really think that all parts of CSS 2.1 are implemented fully and correctly by all browsers that claim they are CSS 2.1-compliant? The only difference now is that in CSS3 there are more useful boundaries between the various parts of the specification, which makes it easier to get working implementations and to get final specs out faster. The CSS WG discusses the reasoning behind making the spec more modular, and frankly, it’s one of the few things that seems genuinely well thought out.
-
@snover said:
Huh? How is this any different than what happens today?
Well it's certainly not getting any better, apparently.
I revers my earlier statement. The future browsers, it seems, will require the same vast knowledge of quirks that today's browsers require, and the browsers that came before those.@snover said:
Do you really think that all parts of CSS 2.1 are implemented fully and correctly by all browsers that claim they are CSS 2.1-compliant?
Yes, damnit. Or at least, I must hope that they put the bar much higher than it is today. Browser crap significantly reduces the joy I have in coding websites.
@snover said:
The only difference now is that in CSS3 there are more useful boundaries between the various parts of the specification, which makes it easier to get working implementations and to get final specs out faster. The CSS WG discusses the reasoning behind making the spec more modular, and frankly, it’s one of the few things that seems genuinely well thought out.
I agree with that. It makes it easier for vendors to tick off "complete!" on a bit, and developers can trust that it is good & true.
-
@newfweiler said:
Now reread all the posts above that use "old", "new", "recent", etc. as criteria for using browsers, only for "IE6" substitute "Unix" (much older than IE6 by the way -- how many of you are still using this 40-year-old operating system?) and for the most recent browser substitute "Microsoft Windows Vista Ultimate Edition". Still think new things are better than old things?
Actually, yes. I use Vista Business on my personal machine (and it's a damn sight better than XP Pro *EVER* was - not least because I can actually multitask thanks to the fact that Vista is a 64bit system and has drivers, unlike XP64), and I have exactly 1 vintage Unixen... Because I can't crowbar a Linux onto it.Furthermore, your analogy is flawed because "Unix" is a product family. "IE6" is a version of a product. "How many of you are still using System V Unix?" would be more appropriate. That answer? FUCKING NOBODY. Seriously, I doubt there's a running production copy anywhere in the entire fucking world. Everybody uses Linux or BSD or HP-UX these days.
-
@newfweiler said:
Now reread all the posts above that use "old", "new", "recent", etc. as criteria for using browsers, only for "IE6" substitute "Unix" (much older than IE6 by the way -- how many of you are still using this 40-year-old operating system?) and for the most recent browser substitute "Microsoft Windows Vista Ultimate Edition". Still think new things are better than old things?
Two people already proved that you're anology gravely sucks. Furthermore, I'm pretty sure there are UNIX versions that are more recent than Vista: HP-UX 11i v3 for example. And if you count BSD an Linux there are hunderds...
