Comments for Dan's blog http://www.dberlin.org/blog Where innovation meets insanity Fri, 29 Aug 2008 04:07:21 +0000 http://wordpress.org/?v=2.6 Comment on Points-to analysis and the real world by Christophe http://www.dberlin.org/blog/2007/03/18/points-to-analysis-and-the-real-world/#comment-11997 Christophe Tue, 02 Oct 2007 11:34:20 +0000 http://www.dberlin.org/blog/2007/03/18/points-to-analysis-and-the-real-world/#comment-11997 Hi Dan, It seems that you're really found of pointer analysis papers. I really liked your position on current research works. Since I'm starting research on this topic, it would be great if you could share your state-of-the-art bibliography on pointer analysis (for C, Java, etc.). Many thanks in advance, Cheers Hi Dan,

It seems that you’re really found of pointer analysis papers.
I really liked your position on current research works.
Since I’m starting research on this topic, it would be great if you could share your state-of-the-art bibliography on pointer analysis (for C, Java, etc.).
Many thanks in advance,

Cheers

]]> Comment on Attention horrible Westchester drivers by commercial hotspots http://www.dberlin.org/blog/2005/08/21/attention-horrible-westchester-drivers/#comment-11993 commercial hotspots Sat, 25 Aug 2007 07:12:34 +0000 http://www.dberlin.org/blog/?p=6#comment-11993 <strong>commercial hotspots...</strong> commercial hotspots... commercial hotspots…

commercial hotspots…

]]> Comment on How copyrights and patents work, for engineers by Luis http://www.dberlin.org/blog/2007/06/20/how-copyrights-and-patents-work-for-engineers/#comment-11987 Luis Mon, 25 Jun 2007 12:22:10 +0000 http://www.dberlin.org/blog/2007/06/20/how-copyrights-and-patents-work-for-engineers/#comment-11987 I've considered giving a similar talk at a couple conferences; would love to get my hands on these slides as a starting point if/when they are done. I’ve considered giving a similar talk at a couple conferences; would love to get my hands on these slides as a starting point if/when they are done.

]]> Comment on CVS2SVN and bitching about converted repository size by Daniel Jacobowitz http://www.dberlin.org/blog/2007/05/16/cvs2svn-and-bitching-about-converted-repository-size/#comment-11970 Daniel Jacobowitz Wed, 16 May 2007 17:59:16 +0000 http://www.dberlin.org/blog/2007/05/16/cvs2svn-and-bitching-about-converted-repository-size/#comment-11970 FYI: I fixed this problem last year and sent the cvs2svn developers a patch. It's taking them a while, but they're still in the process of integrating it. FYI: I fixed this problem last year and sent the cvs2svn developers a patch. It’s taking them a while, but they’re still in the process of integrating it.

]]> Comment on Points-to analysis and the real world by John "Z-Bo" Zabroski http://www.dberlin.org/blog/2007/03/18/points-to-analysis-and-the-real-world/#comment-11969 John "Z-Bo" Zabroski Mon, 14 May 2007 21:27:23 +0000 http://www.dberlin.org/blog/2007/03/18/points-to-analysis-and-the-real-world/#comment-11969 In a nutshell: ideally, <b>successful design anticipates all relevant and possible ways failure can occur</b>. Some people have a paranoidic approach to design, and they imagine that the <i>impossible could happen</i>. Failure in design is often non-quantitative, especially in software engineering, where accidental complexities are essentially non-numerical. Has anyone ever published historical case studies on these design failures in optimizing compilers? I wouldn't think so, because generally you implement an algorithm, it sucks, and you throw it away. Although these sorts of failures are not as dramatic as the events at IBM that contributed to the "Software Crisis", there are probably some lessons that can be learned. In a nutshell: ideally, successful design anticipates all relevant and possible ways failure can occur. Some people have a paranoidic approach to design, and they imagine that the impossible could happen. Failure in design is often non-quantitative, especially in software engineering, where accidental complexities are essentially non-numerical.

Has anyone ever published historical case studies on these design failures in optimizing compilers? I wouldn’t think so, because generally you implement an algorithm, it sucks, and you throw it away. Although these sorts of failures are not as dramatic as the events at IBM that contributed to the “Software Crisis”, there are probably some lessons that can be learned.

]]> Comment on Supreme court smacks down Federal Circuit, twice by Laurent GUERBY http://www.dberlin.org/blog/2007/04/30/supreme-court-smacks-down-federal-circuit-twice/#comment-11966 Laurent GUERBY Tue, 01 May 2007 09:18:10 +0000 http://www.dberlin.org/blog/2007/04/30/supreme-court-smacks-down-federal-circuit-twice/#comment-11966 Hi Daniel, here is a compiled link lists about the two IP decisions: http://guerby.org/blog/index.php/2007/04/30/159-la-cour-supreme-des-usa-secoue-le-monde-du-brevet For me obviousness comes from the "improve progress", there's no way you improve progress if every single thing unser the sun is patented :). Hi Daniel, here is a compiled link lists about the two IP decisions:

http://guerby.org/blog/index.php/2007/04/30/159-la-cour-supreme-des-usa-secoue-le-monde-du-brevet

For me obviousness comes from the “improve progress”, there’s no way you improve progress if every single thing unser the sun is patented :).

]]> Comment on Why prior art doesn’t matter - combining references by Steve Bragg http://www.dberlin.org/blog/2005/08/10/why-prior-art-doesnt-matter-combining-references/#comment-11964 Steve Bragg Mon, 30 Apr 2007 20:22:52 +0000 http://www.dberlin.org/blog/?p=5#comment-11964 The Supreme Court <a href="http://www.bloomberg.com/apps/news?pid=20601087&sid=aHigP6BkC8pE&refer=home" rel="nofollow">just challenged</a> the "teaching-suggestion-motivation" test. The Supreme Court just challenged the “teaching-suggestion-motivation” test.

]]> Comment on Warmer, healthy, productive by Steven http://www.dberlin.org/blog/2007/02/21/warmer-healthy-productive/#comment-4343 Steven Fri, 23 Feb 2007 16:22:29 +0000 http://www.dberlin.org/blog/2007/02/21/warmer-healthy-productive/#comment-4343 Fitter, healthier, more productive... http://lyriki.com/index.php?title=Radiohead:Fitter_Happier :-) Fitter, healthier, more productive…
http://lyriki.com/index.php?title=Radiohead:Fitter_Happier

:-)

]]> Comment on Optimizing the fsfs backend by Don Lapre James http://www.dberlin.org/blog/2006/06/26/optimizing-the-fsfs-backend/#comment-2648 Don Lapre James Thu, 04 Jan 2007 23:57:12 +0000 http://www.dberlin.org/blog/2006/06/26/optimizing-the-fsfs-backend/#comment-2648 I agree EdG. Sounds like it would work better that way. Jim Don Lapre James webmaster@linuxrocks.org www.linuxrocks.org   I agree EdG. Sounds like it would work better that way.

Jim
Don Lapre James
webmaster@linuxrocks.org
http://www.linuxrocks.org  

]]> Comment on Optimizing the fsfs backend by DannyB http://www.dberlin.org/blog/2006/06/26/optimizing-the-fsfs-backend/#comment-2429 DannyB Fri, 29 Dec 2006 01:33:59 +0000 http://www.dberlin.org/blog/2006/06/26/optimizing-the-fsfs-backend/#comment-2429 These discussions occur mainly on the dev list. The reason no method exists for doing it yet is because nobody who wants to expend the development effort has had a need. GCC's repository has > 100k revisions, all in a single dir, with absolutely no noticeable performance effect (it is backed by ext3, which these days supports hashed directories). Like most projects, the reality is there are a lot of people who simply sit there and believe that if they wish hard enough, the work will get done :) These discussions occur mainly on the dev list.
The reason no method exists for doing it yet is because nobody who wants to expend the development effort has had a need. GCC’s repository has > 100k revisions, all in a single dir, with absolutely no noticeable performance effect (it is backed by ext3, which these days supports hashed directories).

Like most projects, the reality is there are a lot of people who simply sit there and believe that if they wish hard enough, the work will get done :)

]]>