Dienstag, 2. März 2010 link
"Compcert is a [verified] compiler that generates PPC and ARM assembly code from a large subset of the C programming language. The particularity of this compiler is that it is written mostly within the specification language of the Coq proof assistant, and its correctness was entirely proved."
Extremely impressive work by [create Xavier Leroy] et al.
Jon Stokes: "Perhaps [the] most likely reason behind Apple's silence, is that the A4 just isn't anything to write home about [..]. If Apple were to tell you what's in the A4, most of the focus would be on what the chip is not, rather than on what the iPad is."
I've got to confess to having indulged in speculations about the A4 in the past. May the above help to finally dispel quad-core-featuring myths. Even if reality will catch up with those in a month or so anyway.
0 comments
Samstag, 20. Februar 2010 link
LLVM is coming to Haskell, in the form of David Terei's LLVM backend which is being merged into mainline GHC. Based on that, GHC seems to be also dropping it's C backend, in favour of their own native code generator and LLVM as fall-back, wherever the former is not yet available/feasible.
And while we are in Haskell-land: the
uvector package is now deprecated in favour of vector, which now provides the same performance but not only makes all those U suffices go away, but is also more versatile. Why should you care at all about vector (or uvector, for that matter)? Two words: Loop fusion (based on DPH work)! To spread his excitement, Don Stewart has started "Numeric Haskell: A Vector Tutorial" which, based on the NumPy tutorial, nicely introduces using vector in Haskell.Addendum 2010-02-22: Don Stewart did some micro-benchmarking of GHC's native, C and LLVM backends using the
vector package.1 comments (by headcore)
Mittwoch, 17. Februar 2010 link
A rose by any other name ... Helma NG is now RingoJS.
0 comments
"Moblin and [create Maemo] are merging! We are taking the best pieces from these two open source projects and are creating the MeeGo software platform."
0 comments
Mittwoch, 10. Februar 2010 link
Doug Gregor: [On 2010-02-04], Clang completed its first complete self-host! We built all of LLVM and Clang with Clang (over 550k lines of C++ code). The resulting binaries passed all of Clang and LLVM's regression test suites [...]"
Which is fantastic news!
0 comments
Peter Kirn: "The iPad, in a single device, embodies the exact opposite of all the reasons I’ve invested so much time in computing for the last 25 years."
Well said. Looks like a seducingly beautiful device, though.
For a different (optimistic) angle, read Stephen Fry's iPad About. The key point for me: let's hope that the iPad also delivers a "100,000 volt taser shot up the jacksie" to the market.
Take, for example, Lenovo's IdeaPad U1. On the one hand, it's a beautifully engineered device, hardware-wise, that easily matches (if not beats) the iPad in specs. Looking at the IdeaPad's tablet user interface, on the other hand, gives me the creeps: could it really be more awful and sluggish?
Apple, having total control, end-to-end, from hardware to software is obviously at a distinct advantage here. But it is also this very control and general locked-down-ness that makes me wholeheartedly agree to Peter Kirn as quoted above. And I don't think that Apple's advantage is too big to overcome, in general. Let the smartphone market be my witness for this: no, an Android phone will probably never completely match an iPhone in sleekness. But 90% of the sleekness coming along with a more open ecosystem makes the whole package vastly more compelling to me.
So I, for one, am hoping for the "taser shot up the jacksie" to bootstrap an exciting market.
0 comments
Sonntag, 24. Januar 2010 link
"The Internet 2009 in numbers" -- Executive summary: Internet 2009 = Facebook + YouTube + Email Spam (+ BitTorrent).
0 comments
Montag, 11. Januar 2010 link
"The trouble with the world is that the stupid are cocksure and the intelligent are full of doubt."
-- Bertrand Russel
Chromium Issue #31482: "What is the expected result? I expected Google Chrome to teleport maybe three, maximum five goats!"
0 comments
Donnerstag, 31. Dezember 2009 link
Onward!
0 comments
Roughly a year after the last release (2008-12-12), beanstalkc 0.2.0 is out! This release comes with a few minor incompatible API changes, which are detailed in the release announcement.
Here's the unabridged list of changes:
$ git log --reversed --pretty=format:"[%h] %s (%an)" v0.1.1..v0.2.0
[0049ffb] Minor nit: make Job inherit from object (Eric Florenzano)
[4160b3c] Remove unused time import. (Eric Florenzano)
[ce041e5] Fix typo in -p command in Getting Started section of tutorial (Chris Scott)
[7880eaa] Fix for Python 2.4 (Andreas Bolka)
[943d572] Fixed typo (Andreas Bolka)
[a4e7d1a] Prefix private methods with an underscore (Andreas Bolka)
[d04cfb1] Fix delete to expect a NOT_FOUND response (Andreas Bolka)
[ac7c037] Fix beanstalkd command in tutorial (Andreas Bolka)
[6e282a0] Change YAML-related names for clarity (Andreas Bolka)
[92ed027] Document our PyYAML dependency and alternatives (Andreas Bolka)
[0053c5b] Document job priorities (Andreas Bolka)
[f7a32d4] Update beanstalkd URL (Andreas Bolka)
[2f5f61c] Mention PyYAML dependency in readme (Andreas Bolka)
[89583b1] Correct typo (Andreas Bolka)
[65a279f] Job priorities are documented now (Andreas Bolka)
[37a67c1] Use socket#sendall to fail faster (Andreas Bolka)
[23eaf7a] Make silent socket errors explicit (Andreas Bolka)
[a49928f] Wrap all socket exceptions in SocketError (Andreas Bolka)
[2f0a663] Update copyright notice (Andreas Bolka)
[b1c5f40] Wrap long line (Andreas Bolka)
[a494e1d] Gracefully handle missing PyYAML (Andreas Bolka)
[38122ab] Relax doctest to cope with more stats info (Andreas Bolka)
[c076e4c] Prefix parse_yaml, socket, socket_file with _ (Andreas Bolka)
[56b0e00] Implement pause-tube support (Andreas Bolka)
[d970f04] Release version 0.2.0 (Andreas Bolka)
0 comments
Sometimes, a simple
flatten zip a b goes tragically wrong:mixin: funco [a [series!] b [series!] /local v][
parse a: copy a [
some [skip if (v: first+ b) insert v] a:]
]
head clear a
]
WTH?
0 comments
Donnerstag, 17. Dezember 2009 link
{$\x#,(1 1f;1 0f)}3 comments (by headcore, earl)
search
recent stores (more)
recent comments
