Elsewhere monthlies

This is the monthly archive for October 2007.

28 October 2007

JavaScript Library Overview

John Resig's slides from his presentation at the Ajax Experiences. This sort of overview is SO useful.

Libraries | Permalink

Wir unterbrechen das laufende Programm ...

For once a German blog post: Einfach für Alle warns against the Firefox update, since it seems to contain a few CSS bugs. I haven't seen this warning anywhere in the English blogosphere, and all documentation is in German.

The problem seems to be that floats with negative margins aren't cleared properly, and aren't properly contained in their parent element, either. (Source)

Good to know.

Mozilla | Permalink

Cascading Style Sheets (CSS) Snapshot 2007

Back in May, Andy Budd proposed CSS 2.2 as a way of filling the gap between the 2.1 specification and the current browser implementations.

The CSS Working Group has taken him up on this by publishing the CSS Snapshot, which gives an overview of the CSS 2 and 3 modules that browsers currently (mostly) support.

Any himself indicated that this is not exactly the CSS 2.2 he envisioned. He'd hoped for

a list of all the selectors, properties and values that the working group felt were stable and ready for implementation. That way, browser manufacturers could start implementing and testing new features, under the knowledge that they weren’t going to change. Similarly, us web developers could start playing with these features and baking them into our more avant guarde projects.

He's right in that the Snapshot doesn't exactly provide such a list, but I'm afraid it's the best we're going to get right now.

CSS, Standards/W3C | Permalink

25 October 2007

Mac OS X Web browser rundown 2007

Roger gives a useful overview of where Mac browsers stand nowadays.

Browsers | Permalink

If It Looks Like a Cow, Swims Like a Dolphin and Quacks Like a Duck, It Must Be Enterprise Software

About enterprise software:

what is it about the world of enterprise software that routinely produces such inelegant user experiences?

Good question. I absolutely hate enterprise software, especially the parts that are supposed to maintain a website.

Business | Permalink

23 October 2007

YUI Theater — PPK on the Professionalization of Frontend Engineering

Back in June 2006, Nate Koechley invited me to come to Yahoo! one day, and last Friday I was finally able to take him up on that offer during my first ever visit to San Francisco. He also invited me to give a presentation.

For various reasons I was unable to create a unique presentation for Yahoo!, and therefore I translated my Dutch slides about Fronteers, the Guild of Front End Developers we're setting up in Holland, and repeated the presentation I've given three or four times back home, though without the really tricky subjects such as finance.

Fortunately the Yahoo! people were very much interested in this subject (in fact, Nate's working on professionalisation, too).

Of course Yahoo! recorded the event, and the record has now gone live. Since I absolutely hate seeing and hearing myself on any kind of screen, I haven't watched the video myself yet. Eventually I will, but only in the privacy of my own house. (Right now I'm at the VTM conference.)

Nonetheless, I assume that some of my readers are interested in what I had to say. Please excuse the accent and the slightly ... well ... quirky presentation style. I still have a depressing amount to learn.

Professionalism | Permalink

20 October 2007

Evaluate Low Level JavaScript Performance Characteristics

Benchmark tests for run times of low-level JavaScript functionality in the four major browsers. Overall ranking:

  1. Opera
  2. Safari
  3. Firefox
  4. IE

The results of my own tests tend to swap Safari and Firefox, but they're different sorts of tests. In any case we're one step closer to understanding performance issues in the four major browsers.

Benchmarks, JavaScript | Permalink

Why do DOM interfaces suck so much?

Dustin takes issue with ugly DOM interfaces; especially those that require endless lists of parameters nobody ever uses; for instance window.find or initMouseEvent.

The most striking example remains the third argument of addEventListener: people rarely use event capturing, so why not make this argument optional with the default set to false (bubbling). I agree completely.

DOM | Permalink

17 October 2007

Dealing with the Flexibility of JavaScript

Neil Roberts takes a stab at defining the thin line between well-modularised and obfuscated JavaScript.

a good rule of thumb is that each parameter should always do the same thing. [...] If you have a parameter named url, go ahead and allow it to be a variety of object types, as long as they all indicate URLs. But once you start letting a parameter be a URL or a date, and you figure out which it should be based on variable analysis, you should start worrying.

No doubt there are many more such rules; I'm glad somebody is attempting to describe them.

JavaScript | Permalink

WebDirections North 2008

They're going to do it again: Web Directions North 2008 was just announced. I'm afraid I can't afford to visit it, but I'll be with the speakers in my thoughts.

Incidentally, there are quite a few speakers from outside the regular web dev circuit; excellent choice!

Conferences | Permalink

Findings From the Web Design Survey

At last! Interesting read. I don't entirely agree with all the extrapolations (for instance, "Hispanic" has no meaning in a European context), but that's quibbling about details. The main thing is that Zeldman cs. have done an astonishing job that very badly needed to be done.


Surveys | Permalink

13 October 2007

Screen Readers and display:none

Gez Lemon discusses the difficult relation between screen readers and display: none. The latest news was that screen readers honoured display: none, but Gez has found circumstances in which they nonetheless announce hidden content.

Screen readers | Permalink

11 October 2007

CSS Eleven: Are You In Or Out?

David Storey of Opera has a few critical notes about the CSS Eleven initiative. Although some of these are unavoidable (some people are not invited; big deal; I wasn't and I don't care), his other points may be more serious.

For instance, David feels that the members should be "elected"; and that brings us smack bang back to the elitism vs. democracy debate, in which I stand firmly on the side of the elitists (more on that in March in Austin).

The CSS Eleven consist of people who've decided to sacrifice part of their spare time for helping out countless other web designers/developers. As a general rule I feel that whoever does the actual work decides exactly how it is to be done.

Who would elect the CSS Eleven? Why would they have to be elected, anyway? Why should we let people who are not willing to spend a lot of time on these problems to have a vote in deciding who will be allowed to spend their time on it?

Further, David talks about the predominance of white males. Although this is undeniably true, I refuse to see it as a problem. After all, our web development world is predominantly white and male. (David also said there were no women in the group; and that's not true: Jina Bolton appears quite female to me.)

What David doesn't mention is the under-representation of non-native speakers of English (Roger Johansson is the single one).

Again, I don't really mind. I'm going to speak at two conferences in the next six weeks, and in both I'll be the single non-native speaker. No big deal, but when we bemoan the lack of women and non-whites in our web development world, non-native speakers of English should at least get an honourable mention.

Anyway, this, too, is something that won't be decided in a hurry. I wish the CSS Eleven all luck in the world; they'll need it.

CSS | Permalink

CSS Eleven

Andy Clarke has announced CSS Eleven,

an international group of visual web designers and developers who are committed to helping the W3C's CSS Working Group to better deliver the tools that are needed to design tomorrow's web.

Let's see what they come up with.

CSS | Permalink

Custom Attributes And Class Names

There is some discussion going on about using custom attributes. Alex Russell uses them in Dojo (and I myself have been using and promoting them since 2003). First Aaron Gustafson criticized him mildly, Aaron's main argument being that you can use a custom DTD.

Now Dan Webb weighs in. He goes further than Aaron: he doesn't want custom attributes at all. Instead, he points to the possibility of using good old class to store the data the scripts need. Possible, but when you've got a lot of data to store, it can become ugly in a hurry.

All in all, right now I tend to side with Alex, but I'll read the arguments of the opposite camp with care.

Maybe we can get around the problem by officially specifying a bunch of attributes that are expressly meant for storing script data. I'd like more than one; ideal would be something like:

<element scriptData1="retrospect" scriptData2="opaque" />

I don't think "scriptData" is the best name, but the idea would be that there's an infinite set of attributes to hold script data; and yes, that would mean that scriptData999999 would also be allowed.

Anyway, this discussion has hardly started up yet. I'm curious how it will proceed.

JavaScript, Standards/W3C | Permalink


Vitally necessary.

& Ceiling Cat sawed teh light, to sees stufs, & speraratered the light form dark & stufs but taht wuz ok cuz cats can seez in teh dark & not tripz ovr nethin.

Via Andrew Dupont.

Fun | Permalink

Less Than Perfect

Nate has started working on a kind of checklist for ascertaining site quality. Each site initially gets 100 points, but points are subtracted for, among others, missing doctypes, href="#", and other problems.

The list still has to grow a lot, but it's a very interesting concept that neatly ties in with some things we're going to try over here, even though we want to check web developers, and not sites.

In any case, I'll be following Nate's initiative with interest.

Theory | Permalink

WebKit now supports CSS @font-face rules

Now that, besides Opera, Safari will also start to support @font-face, it's becoming more and more interesting. IE still has its proprietary, but not wholly unworkable, .eot system. So that means that right now only Firefox is missing this particular boat.

Richard Rutter gives an overview of the pros and cons of downloadable fonts, and he finds the cons not very convincing:

  1. Some people will use horrible fonts.
    Sure, but when colours were introduced on the web, some people turned out to use horrible colours. Is that a reason to abolish colours?
  2. Intellectual property. Font foundries will gradually find themselves in the same spot as the music industry (though with three zeroes less of bottom line). The fact that there will be intellectual property problems does not mean we can't use this technique. In fact, the first font foundry that will devise a workable compromise will see a lot of growth.

Welcome, @font-face.

CSS | Permalink


See the September 2007 archive.

This is the linklog of Peter-Paul Koch, mobile platform strategist, consultant, and trainer. You can also visit his QuirksBlog, or you can follow him on Twitter.

Atom RSS

Category rchives: