When
· Naughties
· · 2005
· · · April
· · · · 01 (7 entries)

Technology · A good way to be­come a gen­er­al­ly hap­py per­son is to find your tal­ent and a way to live by it. My tal­ent is soft­ware; and writ­ing about soft­ware ...
 
Truth · What I Believe I don't be­lieve in God, and I don't be­lieve in Adam Smith or Karl Marx, and I don't be­lieve in the ab­strac­tions of Cap­i­tal or Labour or GAAP, and I don't be­lieve in the Na­tion or the Fam­i­ly or the Race or the Tribe ...
 
Not An April Fool’s Joke · Just fool­ish­ness. The XML Bi­nary Char­ac­ter­i­za­tion Work­ing Group has is­sued their fi­nal re­port which rec­om­mends (sur­prise, sur­prise) that the W3C pro­duce a “Binary XML” spec­i­fi­ca­tion. El­liotte Rusty Harold nails it. I don’t care if any­one wants to go off and pro­duce their own da­ta in­ter­change for­mat, bi­na­ry or not, open or not, stan­dard­ized or not, mapped to XML or not; as long as they don’t call it XML. “Binary XML” is an oxy­moron. And I should point out that the peo­ple at Sun who are build­ing a bi­na­ry da­ta for­mat with a map­ping to XML are call­ing it some­thing else en­tire­ly. Th­ese Binary-XML peo­ple are charg­ing head­long on­to the top of a very long, very steep, very slip­pery slope. [Up­date: Fur­ther joy. I see that this poorly-labelled ta­ble as­serts that XML pre­vents both “processing efficiency” and “forward compatibility”. Glad to hear it.]
 
FSS: Chinese New Year · Fri­day Slide Scan #7 is from when­ev­er Chi­nese New Year was in 1984, in Vancouver’s Old Chi­na­town. It is egre­gious­ly Pho­toShopped ...
 
Upcoming Gig: NetBeans Day at JavaOne · This one looks like re­al fun: an all-day NetBeans bash in con­nec­tion with JavaOne. They want me to cov­er Coy­ote, but there’s lots of oth­er stuff to talk about too.
 
Not An April Fool’s Joke · Norm Walsh has a densely-technical post show­ing a nasty prob­lem that’s cropped up in the in­ter­ac­tion be­tween XIn­clude, xm­l:base, and XML val­i­da­tion. Un­less you’re a se­ri­ous XML geek you prob­a­bly don’t want to wade through the de­tail­s, but in his con­clu­sion, Norm rais­es a startling point: “I think what pains me most about this sit­u­a­tion is that XIn­clude was in de­vel­op­ment for just over five years. It went through eleven drafts in­clud­ing three Can­di­date Rec­om­men­da­tion­s. Why didn’t we no­tice this un­til sev­er­al months af­ter XIn­clude was a Rec­om­men­da­tion? I’ll grant that XIn­clude is a fair­ly odd spec­i­fi­ca­tion, in the sense that it’s pro­vid­ing func­tion­al­i­ty that you’d ex­pect to oc­cur down in the pars­er (like en­ti­ties), but it’s on­ly 8,563 words long. If we can’t get a 16 page spec right in three CRs, what hope do we have of get­ting the XSL/XML Query fam­i­ly of spec­i­fi­ca­tions right? By the same met­ric I used on XIn­clude, I get just over a half mil­lion words (505,779) in those doc­u­ments. ” Half a mil­lion word­s... pret­ty scary.
 
Text Encoding Progress · It’s good to see the IETF show­ing for­ward mo­tion on the vi­tal is­sues around how to store text ef­fi­cient­ly; check out the brand-new RFC4042 on UTF-9 and UTF-18. Good stuff.
 
author · Dad · software · colophon · rights
Random image, linked to its containing fragment

By .

I am an employee
of Amazon.com, but
the opinions expressed here
are my own, and no other party
necessarily agrees with them.

A full disclosure of my
professional interests is
on the author page.