Let’s start with Python. First of all, Guido’s lengthy Python 3000 Status Update (Long!). A little bird tells me they’re going to use UTF-16 internally. I’m horrified, that seems egregiously wrong, but Guido’s no fool. Also from Pythonia: Django status update: June 26. Just because the world is breathing all heavy over Ruby these days doesn’t mean Python’s less interesting. If Ruby vanished tomorrow I could transfer my world to Python with fairly little pain. And my code would run faster too.
Moving on to Atom, Trey Drake has more on his project: Filling the Atompub AAA void with an OpenDS backed Atom server. To be brutally honest, most of the Atompub crowd is clueless about these large-scale AAA issues.
Still in REST-land, there’s Announcing Jersey. I gather they’re considering changing the logo from a maillot to a cow.
On another subject entirely, Eric Armstrong feels the pain of a six-year DOM hangover.
And finally... to hell with iPhone, I’m gonna get me a FIC Neo 1973, the world’s first Free Phone, via OpenMoko. It’ll be interesting to see how a pretty-decent and radically-open platform, freely-programmable by any geek with the chops, makes out against Apple’s tightly-locked objet d’art.
Comment feed for ongoing:
From: Sylvain Hellegouarch (Jun 28 2007, at 12:11)
> To be brutally honest, most of the Atompub crowd is clueless about these large-scale AAA issues.
Tim, I appreciate your PR regarding AtomPub and your work around it but I wonder where you could make such claim from?
AtomPub does not make Authentication nor Authorization any different from plain HTTP and well, is this something new? The Accounting bit is even doable already.
So please explain us where you got that feeling that the AtomPub crowd was so knowledgeableness about it?
[link]
From: Andrew Dalke (Jun 28 2007, at 12:25)
Could you ask your little birdy for clarification on Python 3000 using UTF-16 internally? Everything I've read says it will be the same as Python 2.x; compile-time option for UCS-2 or UCS-4.
[link]
From: Matthias Ernst (Jun 28 2007, at 12:28)
> A little bird tells me they’re going to use UTF-16 internally. I’m horrified, that seems egregiously wrong, but Guido’s no fool.
What would you propose? UCS-2 + Javalike codepoint APIs? UTF-32? Thinking about this, I wonder how important random character access to a string really is - so how bad is a variable length encoding actually?
[link]