Comments for Fernando J. Pereda's blag

Comment on How I migrated Paludis to Git by Linus’ Git Talk Index Git Script to Show Largest Pack Objects and Trim Your Waist Line! Spring Modules Fork Eclipse and IPhone+Google Fail spring-modules-ehcache and ehcache issues you should be aware of Fighting Scala – Scala to Java List Conve

Thu, 27 Sep 2012 09:31:38 +0000

[...] For other migration tips (svn) – see here: http://fpereda.wordpress.com/2008/06/11/how-i-migrated-paludis-to-git/ [...]

Comment on The Burning Ship fractal by Cialis

Cialis — Thu, 04 Mar 2010 07:25:51 +0000

PeMuC6 Thank you for the material. Do you mind if I posted it in her blog, of course, with reference to your site?

Comment on MSDNAA, clever? by Dexter

Dexter — Fri, 28 Aug 2009 20:58:23 +0000

yeah .. but that downloader gets them from somewhere .. a simple tcpdump would show you the download site of the .sdc :) ( mine is software.msdnaa.net … )

Comment on How I migrated Paludis to Git by Git Script to Show Largest Pack Objects and Trim Your Waist Line! « Stubbisms – Tony’s Weblog

Fri, 10 Jul 2009 01:07:42 +0000

[...] For other migration tips (svn) – see here: http://fpereda.wordpress.com/2008/06/11/how-i-migrated-paludis-to-git/ [...]

Comment on MSDNAA, clever? by zimbatm

zimbatm — Thu, 11 Jun 2009 10:37:38 +0000

Had the same problem and got my copy by running the program under wine. It needed multiple attempts but finally I got it.

Comment on MSDNAA, clever? by Fernando J. Pereda

Fernando J. Pereda — Fri, 29 May 2009 12:09:39 +0000

Yeah… but thing is, I don’t have a WIndows copy handy :)

— fpereda

Comment on MSDNAA, clever? by Fran

Fran — Fri, 29 May 2009 11:25:24 +0000

Wine is our friend, but yes, it’s hilarious.

I only use MSDNAA to get legal windows’ serials.

Comment on Small C trivia by Fernando J. Pereda

Fernando J. Pereda — Mon, 27 Apr 2009 09:28:40 +0000

I think you are just half-right. The problem here is that floating point (as specified by IEEE754, that is) addition isn’t associative. But you don’t need to involve big and small numbers, the following sequence will give you different results if summed left to right and right to left: 0.1 0.2 0.3 0.4 0.5

Also, for the record, neither does the distributive property hold.

— fpereda

Comment on Small C trivia by Fernando J. Pereda

Fernando J. Pereda — Mon, 27 Apr 2009 09:20:25 +0000

Yeah, that’s a silly typo. Let me correct that.

Comment on Small C trivia by Snaury

Snaury — Mon, 27 Apr 2009 09:00:16 +0000

As for the right answer, if the line there reads f[i-1] = atof(argv[i]), then no, they can be different. The problem is how numbers are stored as floating point, and when a very large number (i.e. that has no bits left for fractional part) is summed with a very small number, then result doesn’t change. So if the first argument is a very large number, and all the rest are comparatively small, but summed give you some comparatively large number (in relation to the first number), then from left to right result won’t change, and will stay equal to the first number. But from right to left you could get a big number, and two big numbers summed would be different from the first number.

…sorry for my poor explanation. ;)