Table of Contents

Articles Posted During January 2003

se_dedit_flamini

I know nothing of Latin, so I can't vouch for its accuracy, but this translation of Blowin' in the Wind is aesthically very pleasing to my mind's ear:

Quot viae quae homo peragranda,
Donec grandem anima fit.
Sic, et quot maria columba velanda
Donec quiescit acta,
Sic et quot sagitt'ab arcibus volanda,
Ante pac'aeternam fore?

Respons'amici,
Se dedit flamini,
Responsum se dedit flamini.

Of course, my mind is most likely butchering the pronunciation, so what do I know? Thanks, Stuart.

2003-01-28 08:55:59

enhanced_google_favelet

I combined my two most oft-used favelets for searching Google. Activating this one searches for the currently-selected text (if any), or prompts the user for a query:

Google [selected]

(You won't be able to test properly without dragging this to the browser's toolbar because activating the link will wipe out your current selection.)

For those interested, the underlying Javascript for this functionality is:


var s = '';
if(window.getSelection)
    s = window.getSelection();
else if(document.getSelection)
    s = document.getSelection();
if(s == '') {
    p = prompt('Search Google:','');
    if(p)
        location.href = 'http://www.google.com/search?q='+escape(p);
}
else
    window.location.href = 'http://www.google.com/search?q='+escape(s);

I've not tested on Windows browsers, but it works in IE5 & Gecko on OS X.

2003-01-26 23:06:15

data_mine

The city of Houston has signed a contract to replace Microsoft Office with a Web-based service from local startup SimDesk for the city's 13,000 PCs. More than just an attempt to save money on software licensing & hardware by replacing conventional workstations with inexpensive 'Internet Appliances', city officials say the program is intended to give every Houston resident access to the application:

[T]he goal is to ensure that every Houston resident has access to basic PC functions, whether they're using a public PC at the library, an old-hand-me-down laptop or a $50 Internet appliance.

That's really what's driving this--our mayor (Lee P. Brown) is going to be the first mayor on the planet to really bridge the digital divide in a major city, Lewis said. It just makes sense to make sure all sectors of our community have access to these kind of productivity tools.

Sounds like an admirable goal, but if the federal government was pushing for a centralized repository of all the information citzens enter into their 'productivity tools', there would be a tremendous outcry from privacy advocates. Should such proposals from local governments be treated any differently?

2003-01-24 08:55:57

multipage_browsing

An alternative for folks who think that tabbed browsing is the only way to surf.

2003-01-21 14:34:30

DENG

DENG is a Flash MX XML/CSS/XHTML/XForms/XFrames rendering engine:

DENG works like a browser (in a browser). Fed with the URL of an XML document, it will load that document, parse it, load and parse user defined stylesheets if present, apply default and user defined stylesheets to each element (CSS 2.1) and finally paint the styled document.

2003-01-21 14:21:53

chimeras_future

Disturbing comments from Mike Pinkerton, Chimera's lead developer:

I'm torn about what to do with Chimera. It's obvious it will only ever be a marginal product on a even more marginal platform. AOL and Netscape have no interest in supporting it. Who aspires to be number two in an already over-commoditized space? Working my ass off for 3% just isn't any fun any more. Safari has already won, the rest is just to see by how much.

I sure would hate to see the only truly open source browser designed for OS X fail due to lack of interest.

2003-01-20 08:53:03

edit_attribute

There's been quite a bit of discussion about XHTML 2.0 lately, but Hixie points out an interesting feature that I missed:

Any element can be marked with an edit attribute to indicate that it was inserted, deleted, changed or moved.

This makes much more sense than wrapping existing elements with INS or DEL tags, and because elements with edit="deleted" would have a default presentation of display:none, I think authors might actually use it. Of course, there's nothing to stop us from adding del{display:none} to our style sheets now.

2003-01-16 21:34:42

spot_the_gif-erence

New b3ta quiz: Guess whether each kitten photo is a GIF or JPEG file. Much to my dismay, I only got 11 correct out of a possible 16.

2003-01-10 11:56:21

extending_safari

Mark Pilgrim has suggested that the Safari developers should help authors embrace the new browser by including some way to target it with specific CSS to work around its various bugs, either by intentionally disabling its support of some obscure CSS syntax, or building in support for some form of invalid, non-standard CSS that only Safari would recognize.

Let me just say straight away that intentionally disabling support of valid CSS selectors is a horrible idea, and I hope the Safari team doesn't even consider it.

Adrian Holovaty suggests using standardized comments to hide rules found between two comments:


/* Begin Safari v1.0 hide */
styles_to_be_hidden {
...
}
/* End Safari v1.0 hide */

I think this is a reasonable approach, similar to Microsoft's conditional comments, and could be utilized to allow Safari-specific rules to be added to the style sheet:


/* show-safari-1.0

...Safari-specific CSS...

*/

Scott Reynen's idea of using 'fake' selectors could be problematic, as I don't know how UAs are supposed to deal with invalid selectors. However, the CSS Spec is clear as to how unknown properties are handled: the declaration is to be ignored. The Mozilla team exploited this when implementing the -moz-* series of proprietary CSS properties.

I've no idea if this is feasible, but I can see where the ability to prepend any valid property with -saf1.0- to tweak a ruleset for Safari would be fantastic for authors.

Another possibility would be the inclusion of a Safari-specific at-keyword, as UAs are supposed to ignore the entire block when encountering unknown at-rules. So, authors could add something like this so their style sheet:


div { color: blue; }
@safari {
    #foo { color: red; }
    #bar { color: red; }
    }
p { color: blue; }

And, DIVs & Ps would be blue in all browsers, while the declarations for #foo & #bar would only be honored by Safari.

Allegedly, Dave Hyatt is still undecided as to whether the idea of crippling (or extending) Safari is a Good Thing™, so we'll just have to wait & see how this plays out. Me? I'm all for extending the browser, at least until it's out of beta.

2003-01-10 01:06:36

safari_vs_tfoot

Mark Pilgrim notes that Safari incorrectly renders the TFOOT element between the THEAD & TBODY elements (as it appears in the document's markup), citing my CSS Filters table as an example of this behavior. This surprised me, because in my experience, Safari rendered the table nearly identically to Gecko-based browsers.

I confirmed that Safari was handling the document correctly, then proceeded to run Safari through some of the test pages for hiding CSS. When I returned to the table, sure enough, the TFOOT was misplaced. Reloading the document produced erratic results, with the TFOOT's placement seeming to randomly cycle between its correct position at the bottom of the table, a point within the TBODY's content (second cell of the twelfth row), to the point Mark described, between the THEAD & the TBODY.

I built a stripped-down demo document in an attempt to track down the bug's trigger, but after rigorous testing, I've been wholly unsuccessful in my attempts to coerce Safari into rendering the test table correctly. I added a simple style sheet to the document, which demonstrates that, unlike Netscape 4.x, Safari does indeed recognize the row group elements, and tests also show that at least some of the element's DOM events fire correctly.

Curiouser & curiouser...

2003-01-07 22:41:45

safari_changelog

Here's an email outlining the changes Apple made to KDE's source when developing Safari. [Hat tip: Brian Foley on the css-discuss mailing list.]

2003-01-07 15:22:27

safari

Apple released a beta browser today called Safari, based on KHTML, the Open Source rendering engine used in KDE's Konqueror browser.

I've played with it a bit & while there are certainly problems, it looks promising. I've given a quick testdrive & found that Safari appears to correctly support the following CSS:

Rest assured, I'll be doing more thorough testing.

2003-01-07 14:36:09

vision_deficiency

Perhaps I need to adjust the ambient lighting in my office, but I couldn't make it past Plate 6 of the Color Vision Deficiency Test. After taking the test, you may want to visit your doctor.

2003-01-07 07:51:47

gone

</2002>

2003-01-01 00:01:53