Elsewhere monthlies
This is the monthly archive for January 2008.
28 January 2008
Bruce points out the accessibility problems the versioning switch will cause: large amounts of the Web will be stuck in IE7 mode.
Although that's undeniably true, the solution is obvious: if accessibility is really an issue, switch to IE8 mode. Yeah, I know there are plenty of people who don't know how to do that or what to do once they arrive there, but how is that situation worse than today's?
Even today, you need a serious web developer in order to make your site really accessible. Serious web developers know about the switch and can estimate its accessibility impact, once enough research has been done.
Contrarily, unaware developers won't be aware of the switch, but they won't be aware of the latest and greatest in accessibility, either, so their not knowing about the switch or the difference between IE7 and IE8 mode pales into insignificance.
All in all I'm not convinced that accessibility will suffer. Even if IE8 had been a normal browser with a normally upgraded engine, we'd still have to cater to IE7 and its accessibility problems for the foreseeable future.
IE, Theory
| Permalink
Jeremy's reaction to Zeldman. He hasn't changed his position substantially.
IE, Theory
| Permalink
Molly disagrees not so much with the versioning switch, but rather with the secrecy in which the talks were held.
IE, WaSP
| Permalink
David Storey disagrees with the versioning switch. In fact, he attributes it to a alleged Microsoft "Divide and Conquer" strategy; something I cannot agree with.
While I understand Chris Wilson's position on "Don't break the Web", and very much understand that this is an incredibly hard nut to crack (I work on similar issues every day), it feels like this proposal is more "break the web for others". There must be a better solution.
I'd like some more information on this. Exactly how will this break the Web for non-IE browsers? I've heard this argument now and then, but have never yet gotten a good reply.
IE, Theory
| Permalink
Mike Davies studies empty links in screen readers. Conclusion:
Not using proper link text forces the browser and screen reader to fallback to heuristics in an attempt to determine what the link text should be.
Not surprisingly, all browsers have their own take on this.
Accessibility, Screen readers
| Permalink
The controversy acted out be lemurs.
Fun, IE
| Permalink
Eric offers an older example from the time he still worked at Netscape. Webmasters simply weren't interested in the "you're doing it wrong so that's why your site breaks in our new version" argument—and they're right, from their point of view.
IE, Theory
| Permalink
Drew McLellan takes a look at JavaScript libraries and the versioning switch.
With version targeting, IE7 will never go away. Just as browsers are born, they must also die and make way for the next generation.
That's an interesting thought I haven't yet seen anywhere else. Good argument.
IE, Libraries, Theory
| Permalink
Liorean agrees with the versioning switch.
Microsoft had the choices of either having some opt-in to improved standards support, or not improving standards at all. An opt-out wasn’t an option, because live content isn’t using that opt-out. Now, every angry and disappointed web developer out there, if you look at it that way, which one would you prefer? Standards with an opt-in or no standards at all?
Exactly!
IE, Theory
| Permalink
Eric takes the argument a step further.
I think a lot of people are discounting the fact that version targeting is absolutely nothing new in the standards world, let alone the web development world. Conditional comments, CSS hacks, and the DOCTYPE switch itself are all examples of version targeting.
That's certainly something I feel people are generally disregarding. The difference between the doctype switch (and CSS hacks) on the one hand, and the new version switch on the other, is, of course, that the first is widely known and the second one is new. Conservatism?
IE, Theory
| Permalink
Mark Pilgrim disagrees with the versioning switch, but at least he's funny.
Fun, IE, Theory
| Permalink
23 January 2008
Peter Nederlof disagrees with the versioning switch.
[The other browsers] don't need a meta tag in addition to a proper doctype, why would IE?
IE has a specific problem the other browsers don't have: Intranets maintained by people who are not (and maybe do not want to become) web developers, let alone standards-aware web developers.
And what's to prevent us from automatically adding meta tags anyway?
Nobody's going to prevent us, but why would you (unless you know what you're doing and WANT IE8 mode)? Doctypes were also necessary to create valid (X)HTML; the meta tag does not have such an extra function. Therefore it will only be added because web developers want to switch IE to a certain mode.
The argument about expected rendering behavior is even more absurd. No serious developer still believes that what IE renders is actually correct.
This is the key point where I feel the No camp is making a serious mistake.
We automatically define "web developers" as "standards-aware web developers". Unfortunately there are plenty of other web developers around. Feel free to denounce them as incompetent, but do not deny that they're an important factor, especially for Microsoft. They are the ones who maintain Intranets created specifically for IE, and any implementation of the standards in IE must take their needs into account, or fail miserably for business reasons.
IE, Theory
| Permalink
Roger Johansson disagrees with the versioning switch—for now. He keeps open the option of changing his mind in the future, though.
IE, Theory
| Permalink
Alex Russell agrees with the versioning switch, though he has quite a few important questions about details.
IE, Theory
| Permalink
Dean Edwards disagrees with the versioning switch, and also objects to using WaSP as a front. That last criticism rings true: WaSP as a whole wasn't involved in this decision, and it shouldn't be pretended that it was.
IE, Theory
| Permalink
Rachel Andrew disagrees with the versioning switch, quoting (again) the default behaviour.
IE, Theory
| Permalink
The Safari team (Maciej Stachowiak speaking) explains why the versioning switch will not be implemented in Safari any time soon, though they do not pass judgement on IE for implementing it.
IE, Theory
| Permalink
Ian Hickson disagrees with the versioning switch. I can't resist the temptation to reply to some of his arguments.
If Web authors actually use this feature, and if IE doesn't keep losing market share, then eventually this will cause serious problems for IE's competitors — instead of just having to contend with reverse-engineering IE's quirks mode and making the specs compatible with IE's standards mode, the other browser vendors are going to have to reverse engineer every major IE browser version, and end up implementing these same bug modes themselves. It might actually be quite an effective way of dramatically increasing the costs of entering or competing in the browser market. (This is what we call "anti-competitive", or "evil".)
I believe Ian paints the current situation too black. True, other browsers will have to take some IE bugs into account when updating their rendering. However, one of the points of the versioning switch is to allow IE to become more standards compliant, so the difference between IE's higher rendering modes and the other browsers would become smaller and smaller as time progresses—which means less and less implementations of IE bugs are necessary.
An ideal situation, true, but not an impossible one.
It will also increase the complexity of authoring by an order of magnitude. Big sites will become locked in to particular IE version numbers, unable to upgrade their content for fear of it breaking.
Unless they decide to redo their sites and hire competent web developers, who'll switch to IE's highest feasible mode. OK, that won't happen everywhere, but in general those sites that won't be updated to a higher IE mode won't be updated today, either, so the situation will not get worse.
Besides, this is the other point of the versioning switch: if you don't want to update, you don't need to. Eventually this will go for Intranet sites more than for Internet, where the existence of other browsers will have a mitigating effect.
Imagine in 18 years — only twice the current lifetime of the Web! — designers will not have to learn just HTML, they'll have to learn 4, 5, maybe 10 different versions of HTML, DOM, CSS, and JS, just to be able to maintain the various different pages that people have written, as they move from job to job.
Or they could decide to upgrade to the highest feasible IE version. In fact, I feel that the situation Ian sketches could increase the standard-awareness of web developers.
- Web developers will have to know about several IE modes; Ian's right about that.
- Therefore they have to study web standards as well as the various modes, since without knowledge of web standards the modes are incomprehensible.
- Once they gain enough insight in web standards, they'll want to work with the highest feasible IE mode, and will switch the sites they maintain to that mode.
The alternative is that they don't want to learn anything and will stay locked in IE7 mode permanently. That would be a pity, but again we're not worse off than we are today. Today, plenty of web developers don't want to learn the standards, either.
All in all, I think Ian may be too negative.
IE, Theory
| Permalink
Nicholas Zakas tentatively agrees with the versioning switch.
Microsoft is in a really tough position. They have really been trying to listen to web developers clamoring for better standards support. They are also listening to their partners, big corporations with thousands of Internet and intranet pages that would cost them millions to upgrade. Microsoft got slapped when IE7 broke some pages by fixing layout bugs; they couldn't afford to repeat that mistake. Whether we like to admit it or not, there are tons of web sites that are designed to work specifically for IE6. They were written at a time when IE had over 95% market share and haven't been updated since.
There literally is no good move for Microsoft to make in this vein. Either they do nothing and don't break the Web but garner the ire of web developers everywhere, or they pacify web developers by forcing standards on everyone and break tons of web sites and web applications. This is the literal rock and hard place scenario. Microsoft is trying to satisfy both conditions and I honestly feel that this is a huge step forward for IE.
I tend to agree with this summary of Microsoft's position.
IE, Theory
| Permalink
Andy Budd sees the point of the versioning switch, but disagrees with the default behaviour. Fortunately he offers actual arguments:
Clueless developers won't know about this behaviour so every new site they build will automatically be rendered as IE7. Clued-up developers will use this as an excuse to freeze support for IE and turn their attentions to better browsers. Users will see less benefit from upgrading and will be more likely to turn to other browsers.
- Clueless developers: correct; that's the point of the default behaviour.
- Clued-up developers: this is an interesting scenario that I hadn't thought of yet. It would lead to massive IE7-only style sheets, but that's kind of OK, I suppose.
- Users: I'm afraid this is a common misconception. Users don't care what browser they're using.
Anyway, good to see some arguments in the No camp.
IE, Theory
| Permalink
Sam Ruby takes a practical approach to the versioning switch: he has engaged the edge value. If you trust your coding skills and believe in progressive enhancement, that's exactly what you should do.
IE, Theory
| Permalink
22 January 2008
John disagrees strongly with the versioning switch. He raises a few technical points:
- Cross-window scripting. What if one window uses one version mode and another one another? A very valid point.
- Security issues. I don't know anything about security, so I really can't judge.
Unfortunately he, too, is making insubstantial claims:
The fundamental issue is that Safari, Firefox, and Opera will all be harmed by attempting to implement this.
I'd love someone to explain this to me. What possible harm can a <meta> tag that they're going to ignore anyway do to Safari, Firefox, and Opera? I suspect an underlying philosophical issue, but I'd like to get it into the open, too. What, besides "it's Microsoft so it must be Evil", is the problem?
IE, Theory
| Permalink
Isofarro disagrees strongly with the versioning switch. So strongly, in fact that I'm wondering if he isn't overshooting his target.
Standards compliant pages must always carry a Microsoft passports that identify the holder as a standards compliant page. Sounds like a ghetto to me.
That's what we are being asked to do. We're segregated from the rest of the Web, and forced to carry identity documents that prove we are standards compliant, and earn the dignity to be rendered in a standards compatible way.
Just comparing something you disagree with to something most people feel negative about is too easy—especially when the comparison makes no sense at all. I call this playing for effect, and not serious debate.
IE, Theory
| Permalink
Zeldman replies to Jeremy and explains that we've got precious little choice. It's either versioning, or no more web standards.
IE, Theory
| Permalink
Andrew Dupont mostly agrees with the versioning switch, though he has some doubts.
IE, Theory
| Permalink
Jeremy might agree with the versioning switch, if it weren't for the default behaviour.
Personally I feel that this default behaviour is the whole point of the switch, and besides I wonder whether the situation will become much worse with the versioning switch in place. It may not become better, but I don't see how it will get worse.
IE, Theory
| Permalink
David Baron disagrees with the versioning switch, pointing to this mail he wrote back in April.
IE, Theory
| Permalink
Robert O'Callahan of Mozilla disagrees with the versioning switch, and asks some interesting technical questions.
IE, Theory
| Permalink
Jonathan Snook agrees with the versioning switch.
IE, Theory
| Permalink
Anne van Kesteren disagrees with the versioning switch. (Note: disagreeing is fine, but assigning ideological blame to people who happen to agree with the plan is not.)
IE, Theory
| Permalink
Chris Wilson talks about the problems the new versioning switch has to solve. Basically, IE6 was broken, but when IE7 was released and solved a lot of CSS bugs, everybody (except for a few standardistas) still expected it to work exactly as IE6.
Of course, it didn't work that way. Sites optimised for IE6 broke in IE7. That was a major problem for Microsoft, and it's the reason they've added the versioning switch to IE8.
IE, Theory
| Permalink
Eric discusses his initial and later reaction to the new IE versioning switch. I agree on the basics: when I first heard of it I wondered if it was such a good idea, but on thinking it over I realised this is probably the best way of handling multiple browser versions now and in the future.
IE, Theory
| Permalink
Aaron Gustafson unveils the new versioning switch in IE8, and discusses why the doctype switch is broken.
IE, Theory
| Permalink
John talks about rounding problems with percentual widths. The problem is a nasty one, and I decided long ago that even a perfect browser would have serious problems with this. Therefore I'm not particularly surprised that browsers differ on the exact treatment of percentual widths.
Just make sure you test everything in all browsers, and don't try to make the elements fit perfectly—that won't work.
In fact, when I use percentual widths my standard routine right now is to make very sure that, when all widths are added up, about 2 to 4 % of the total width is left for the browser to divide as it sees fit. (OK, usually my elements also have a padding, and in the "W3C box model" padding make everything much more complicated.)
CSS
| Permalink
Eric has created a historical table for browser releases. It's always useful to see the gaps between releases in a graphic way.
BTW, the creation process wasn't exactly easy.
History
| Permalink
14 January 2008
John is preparing for his next book and is wondering if he should treat memory leaks. IE6 was infamous for its memory leaks, but the problem has been solved in IE7 and IE6 is on the way out. Should a description of this problem still take valuable book real estate?
Personally I think that memory leaks should still be treated, although a detailed description of the IE6 problems is not necessary any more by the time the book is released. A short summary might be useful, though.
Besides, other browsers are rumoured to have memory leaks, too. I'd love to have a solid description of these problems—or a statement that despite rumours these browsers have in fact no problems.
So my answer is Yes, treat them. Keep the treatment short, except for problems that creep up very frequently.
Browsers, JavaScript
| Permalink
Ian Hickson asks for our help in filling the last 16 questions of the Acid 3 (JS) test. If you have time on your hands and access to Firefox and Safari trunk builds, participate.
Browsers, JavaScript, Tests
| Permalink
10 January 2008
Thomas Vanderwal tries to define the elements that social software has to enable. An interesting read.
I couldn't resist applying his ideas to myself and my blog, and in general it seems they fit quite nicely. It seems that this has a broader application than just social software (unless you count blogging software as social; something I don't do).
However I'm wondering about Relationships, since I feel they can be asymmetric. A person comments on my blog, and tries to establish a relationship, but I consider him a bore or worse and decline the offered relationship. Does that fall within Thomas' definition of Relationship? Probably, but I feel that there's a loose end here. (On the other hand, I could be totally wrong. I'm not really used to thinking in these terms.)
Society
| Permalink
Funny idea. Maybe this is the solution to the low comment quality problem.
Via Andrew Dupont.
Society
| Permalink
John Resig explains cross-site XMLHttp. The principle is simple: someone offering data may opt to add a header that says which other sites are allowed to download and play with the data.
A few months ago I mused about something like this, but I didn't post anything about it since I couldn't (and can't) judge the security consequences. Now the Firefox team says there aren't any (which seems to be implicit in them incorporating this feature in FF3), which is good enough for me.
Data Retrieval
| Permalink
Older
See the December 2007 archive.