mpt News

AOL client to use Mozilla’s rendering engine, finally

NewsForge: “The Gecko rendering engine at the heart of the Mozilla Web browser is scheduled to replace Microsoft’s Internet Explorer as AOL’s default browser — the one in the millions of free AOL CDs distributed every year — in the 8.0 version of AOL’s client software.”

For years, many of us have been waiting for one of two things to happen:

Microsoft being forced to unbundle Internet Explorer from Windows;
AOL basing their client software on Mozilla rather than on Internet Explorer.

Given Microsoft’s lawyers and the USDOJ’s cold feet, the first might not happen for some decades. The second, however, appears to be just around the corner.

In the next couple of years I think we can expect AOL/Time-Warner sites to start validating, but Microsoft aren’t going to give up the standards that easily. Remember how monopolies work. In the board game, owning all properties in a group gives you the ability to charge higher rent on all of them. Same thing applies here.

Microsoft used their Windows monopoly to establish the dominance of Internet Explorer, and currently they’re using the dominance of Internet Explorer to maintain their Windows monopoly (it’s hard to switch to Linux or whatever when some Web sites won’t work in Linux browsers). Now, what’s the next step? Simple. Use their Web properties to protect the dominance of Internet Explorer.

Microsoft controls several high-traffic Web properties now — including Hotmail, MSN, MSNBC, and Expedia. We’ve already seen them dip their toes in the water, experimenting with blocking standards-compliant browsers from their new MSN site. Now I can easily see them going further. Imagine one day needing to use MSIE to check your Hotmail. Or to use Encarta. Or to get a Passport. Then it would become a pitched battle. “Sick of Web sites not working in AOL? Switch to MSN, and get 30 days free!”

I think the next few years are going to be rather messy.

Pixels have feelings, too

Steven Garrity: “What struck me most about the new visual style in [Windows] XP is that it is full of little glitches and holes. None are particularly significant, and this may seem nitpicky, but as Blink 182 puts it so well, it's All the Small Things.”

Death to file pickers

I rant about file selection dialogs on the LinuxStep mailing list: “Why should I have to save a document if I want the computer to remember it? That’s completely retarded. In the real world, if I write a letter, it stays written, until someone shreds it. If I bake a cake, it stays baked, until someone eats it. If I paint a picture, it stays painted, until someone burns it. In the real world, whenever we create something new, we don't have to jump to the nearest telephone and holler ‘Ahoy there! File, Save As! Thankyou!’.”

So funny it hurts

James Lileks: “This is some curious glimpse into someone’s fantasy — a world where men regularly happen across women whose undergarments have fluttered to their ankles.” I have pains in my stomach now — I’ve been laughing at this so hard for so many minutes.

And now for something completely Zimbabwean

If you’re in the USA (or any country which imports computers or laws from the USA), and feeling glum about the gradual erosion of your democracy and your civil liberties, be thankful you don’t live in Zimbabwe at the moment. The day before national elections, their democracy is teetering between anarchy and fascism.

It wasn’t me who wrote my essay so badly

I’ve just spent an hour or so doing something which, in all modesty, I’m very good at: proof-reading an essay. Well, in this case it was an academic paper — something which is much the same as an essay, but with lots more equations.

After trying to explain a particular detail of what makes great essays great, I was browsing the Web for corroborating advice, and happened on the “Cyber Essays” site. Cyber Essays purports to be “your one-stop source for free, high-quality term papers, essays, and reports on all subjects”.

For what purpose? The only reason I can think of for providing a “source” of papers and essays is so that you can plagiarize them. Doing this, of course, is a pretty dumb idea — hopefully most essay markers are smart enough to do a Web search on random phrases from your essay if it seems better than (or even different from) what was expected.

However, I’m wondering if this site is actually a “honeypot” to discourage plagiarism. In computer circles, a honeypot is a dummy computer set up to attract electronic intruders, so that systems administrators can study the cracking techniques used and learn to guard against similar attacks on the important computers in their organization.

Similarly, using “Cyber Essays” as the source for your next essay or term paper won’t get you very far, because — judging by my sampling of the essays — most of them are utter bilge.

The public must not find out about our new browser

Dan Mosedale: “Steve, you just marked this [bug report] Netscape-confidentail [sic] without any comment. Why is there a need for this to be confidental [sic]?”

Stephen P. Morse: “Chris Nalls [Netscape product manager] was concerned because this report mentions 6.5 explicitly.”

Thwap, thwap, thwap.

In the frame for locking up the Web

When I enrolled at the University of Canterbury in 1996, a few people were still browsing the Web with NCSA Mosaic, but most were using Netscape Navigator 1.1 or 1.2. Soon Netscape introduced version 2.0, and with it the most grievous usability torpedo in HTML: frames.

The usability problems with frames are well-documented: Web page tasks which are normally in the vicinity of “Easy” on the pain-o-meter — like saving, printing, bookmarking, and e-mailing — leap towards “Impossible” once frames are involved. So, it was nice for me to discover recently that over six months ago, in their XHTML 1.1 recommendation, the W3C had abolished frames.

You read that right. Frames are officially dead.

You see, XHTML 1.0 — like HTML 4 before it — came in three varieties: Strict, Transitional, and Frameset. If you make a page with frames in HTML 4 or XHTML 1.0, your frameset is supposed to use the Frameset DTD, and the frames themselves use either the Strict or (more likely) the Transitional DTD. What XHTML 1.1 does is continue the Strict DTD, and leave the other two behind. Bye bye <font>, bye bye <frameset>.

The question is, will anyone care?

For most Mozillans, the authority and importance of W3C specifications is without question, sometimes absurdly so. However, in the last couple of years it appears the wheels have started wobbling on the W3C wagon.

Five months ago there was a furore over the use of patents in W3C recommendations, such as those in Scalable Vector Graphics (SVG).
CSS2 and CSS3 define system colors and fonts, which Web authors are supposed to be able to use in order to mimic the user interface of the browser. The particular colors and fonts defined just happen to make a lot of sense for Windows 95/98/2000 (with a little bit of iMac crackheadedness thrown in), but map badly onto other platforms. The bigger issue, however, is that allowing Web authors to mimic the browser’s user interface is nothing more or less than a giant security vulnerability.
The Platform for Privacy Preferences (P3P) is hilariously flawed. It requires Web sites to supply a machine-readable statement about what they are going to do with personal information you give them, a statement which you are then supposed to use in order to decide whether you let them have that information. That is, a Web site has to be trustworthy in order for you to be able to specify that you don’t trust it. At the same time, P3P establishes a system of configurability which is practically impossible for user agents (that is, Web browsers) to present in a way understandable to most humans.
Most damningly, it was recently revealed that only a tiny minority of the W3C’s member organizations support the standards which the Consortium is producing.

The original Web language, HTML, is an interesting case study of how the W3C is starting to lose the plot. HTML 2.0 was very basic, and in the absence of a separate style language, browser vendors added all sorts of tacky extensions to it in order to allow graphically versatile (if not necessarily usable) page designs. However, in the late 1990s, the HTML working group did an excellent job of cleaning up the mess — first producing HTML 3.2 as “this is where we are now”, and then HTML 4.0 as “this is where we should be”.

But even here, the obfuscation of the standards had begun. The HTML 3.2 specification was released as a single Web page. (I have fond memories of poring over this page when teaching myself HTML.) The HTML 4.0 specification, however, has thirty-one pages. Certainly HTML 4 has more features than HTML 3.2, but not that many more.

As for the XHTML specification, I can’t tell how many pages it has — it seems to link into specifications for individual modules, which in turn link to raw code of DTDs with little or no explanation of what the features are for. It’s hopeless. You can’t expect Web developers to read this stuff.

And that’s the problem. If Web developers can’t follow the specifications, they’re not going to pack up their tools and go home. They’re going to carry on with what they know: MSHTML, which is much more clearly documented, and which (of course) corresponds exactly to what they see when testing their site in Microsoft Internet Explorer.

Microsoft is a member of the W3C, so it’s easy to blame malice for the decline of the Consortium’s products where perhaps stupidity is more likely. However, if indeed it is malice, the motive is pretty clear. If the W3C becomes irrelevant, then those who code the majority browser get to control the standards. Into the trunk we’ll go.

Unless the W3C cleans up its act, and soon, developers are going to start ignoring it, or treating it as a curiosity. Then we really will be in trouble.

The Scientology Googlebomb

If everybody linked to a site opposing Scientology, that would counteract the Googlejacking of the search term by the members of the cult.

Bleh, this Googlebombing thing must be the second most placeboic form of activism in the world. (The most placeboic form being posting to Slashdot, of course.)

I’m back, and shorter than ever

Ahhhhhhhh … I needed that holiday. Canoeing, playing piano, learning how to play squash, and generally enjoying the Nelson sunshine. (And yes, when using the family computer I did visit some Weblogs, but I did not inhale.)

I’m 23. My three sisters are 21, 16, and 14. A week at home has revealed that all three of them are now taller than I am. Cringeworthy.

Laterz

I’m going to Nelson for a holiday. Back in a week.

Don’t fear the style sheets

Dave Winer ran a survey which basically came down to this: Would you rather your site was better quality (from using CSS), or that it was full of spelling mistakes so that it doesn’t work as well with Google?

Dave didn’t get the result he wanted, so now he’s rephrased the question. “If you had to choose between a plain text weblog that had something to say and one that used all the latest and greatest technology but had nothing to say, which way would you go?”

O, he’s getting desperate now. A Weblog which uses CSS is more plain text than one which is crufted up with tables and pixel graphics and <font> tags. And using the latest time-saving technology (like CSS, or Radio, or Manila) means you have more time to say interesting stuff, not less.

Dave, if you don’t want to use CSS, then don’t. The rest of us will have sites which are faster, more readable, more accessible, and cheaper to run. And we’ll just mutter to ourselves about how long Scripting News takes to load.