Elsewhere monthlies
This is the monthly archive for August 2005.
31 August 2005
A bit harsh maybe, but essentially understandable. Opera is the browser that is supposed to do everything right, but usually doesn't quite. JavaScript support, especially, has sometimes been problematic (right now the load/unload nonsense bears witness).
Andy's wrong at one point: 'Opera 4 was without doubt more advanced that both IE4 and Netscape 4'. It wasn't! It was a disastrous browser that should never have been released. See my old Opera 4 page in the Browsers section.
Opera
| Permalink
30 August 2005
Roger Johansson explains the difference between elements, tags and attributes one more time. He's right: many people use "tag" for everything, even ALT.
HTML
| Permalink
Eric Meyer discovered that the use of a few ID values that have special meaning in the old document.all DOM may cause IE to crash.
DOM
| Permalink
29 August 2005
Is Google a platform? No, it isn't, and it's not becoming one, either. Is Google an Evil Empire? Not yet.
Insightful article.
Society
| Permalink
Useful overview of the problem with the load event handler and possible solutions, including their drawbacks.
onload
| Permalink
28 August 2005
Very useful overview of the hasLayout property and the problems it causes.
CSS, IE
| Permalink
27 August 2005
Stuart Langridge takes a look at a proposed way of removing CSS hacks from your code, and finds the method wanting.
Although CSS hacks are definitely a Bad Thing, and I find the proposed solution (using <html class='firefox' >, <html class='ie'> etc.) interesting, I agree with Stuart that it calls for server side browser detects. In additon to Stuart's criticism, I feel that browser detection is generally too subtle an art for server side coders who don't know their browsers. Besides, we all agreed we wouldn't use them any more, didn't we?
CSS
| Permalink
Davaid Flanagan doubts whether the lack of Back and Forward functionality within an AJAX application will make a difference to users.
Usability
| Permalink
A sneak peek into Jeremy's upcoming book. This chapter explains why JavaScript has such a bad name, and what we can do about it. Needless to say, I completely agree.
JavaScript
| Permalink
Jeremy bravely attempts to define the difference between 'American' and 'European' design. I suspect that 'European' design doesn't (yet) exist: we should talk about British, Dutch, Swedish, French, etc. design, since all of these countries have their own design traditions.
Nonetheless the question remains whether we can discern such 'national' design identities at all.
Society
| Permalink
Scott Andrew discovered that Firefox does support document.all, which it's supposed to ignore. So Firefox now executes supposedly IE-only code branches.
Annoying, but since we haven't needed document.all at all since IE 4 went down, it shouldn't be too big a problem for standards aware JavaScript sites.
DOM, Mozilla
| Permalink
ALA returns. Excellent articles, as usual.
Society
| Permalink
Derek summarizes the recent accessibility discussions, and introduces a highly useful distinction between accessibility and availability. I agree with him, with the caveat that I haven't yet read any of the other recent accessibility posts and articles.
Accessibility
| Permalink
Very useful CSS trick, and it's so damn obvious once you understand it.
CSS
| Permalink
19 August 2005
Tim Bray explains why 'Web 2.0' is largely a buzzword.
'Open Source means something quite specific that you can explain to anyone in a sentence or two. Web 2.0 means so little, and is so vulnerable to corruption from the hypemeisters, that I suspect that if it’s a meme at all, it’ll be short-lived.'
Society
| Permalink
1 August 2005
Some tests for determining JavaScript execution speed. Please do the tests; the results are automatically added to the database.
Benchmarks, JavaScript
| Permalink
A reaction from the Safari team to my benchmark test of the same name. Although my conclusion seems not to have been correct, the test case allowed the Safari programmers to solve a bug in their getElementsByTagName implementation.
CSS modification, Safari, Tests
| Permalink
On IE problems with pseudo-elements and pseudo-classes.
CSS, IE, Tests
| Permalink
Older
See the July 2005 archive.