Am 05.06.2010 23:24, schrieb Thomas Keller: > Am 05.06.10 18:26, schrieb Zack Weinberg: >> On Sat, Jun 5, 2010 at 4:59 AM, Thomas Keller <m...@thomaskeller.biz> wrote: >>>> I assume you meant "/dev/nul instead of /dev/null" there? >>> >>> No, I meant /dev/nul - try GNU patch and tamper the patch file to >>> "+++ /dev/nul" (which is of course wrong) - then you'll see that empty >>> files are kept and are only removed explicitely with -E. This is what I >>> mean - we'd tamper the test if we add -E, because then we just test GNU >>> patch to correctly interpret the -E option, but not to automatically >>> remove the file if it is empty _and_ has the target /dev/null. >> >> Oh, I see what you mean. Well, is the exact behavior of the system >> 'patch' utility the focus of the test? I'm kinda inclined to doubt >> it, but I have lost the message that says which test we're talking >> about here... > > The test is diff_patch_drop and IMHO it exactly tests the proper removal > of the files at the end. If not we could / should probably just remove > the last two checks.
I have changed the test that it now executes patch with -E on BSD's and that it only checks for a removed directory on non-BSD's (apparently GNU patch is much more aggressive). FreeBSD and openBSD now pass this test - the openBSD bot is still broken with three other tests though. Thomas. -- GPG-Key 0x160D1092 | tommyd3...@jabber.ccc.de | http://thomaskeller.biz Please note that according to the EU law on data retention, information on every electronic information exchange might be retained for a period of six months or longer: http://www.vorratsdatenspeicherung.de/?lang=en
signature.asc
Description: OpenPGP digital signature
_______________________________________________ Monotone-devel mailing list Monotone-devel@nongnu.org http://lists.nongnu.org/mailman/listinfo/monotone-devel
