On 07/07/2010 14:31, Michel Dänzer wrote: > On Mit, 2010-07-07 at 13:54 +0200, Yves-Alexis Perez wrote: >> On 07/07/2010 12:21, Michel Dänzer wrote: >>> >>> Overall, I think building evolution-data-server with --enable-largefile >>> incurs >>> too much risk for Evolution users at this point. Unless there was a very >>> compelling reason for enabling it, please disable it again for now. >> >> I'm not especially attached to largefile, but in this case it's not >> exactly a good idea imho. The bugs weren't directly related to >> largefile, but enabling it make them appear on i386. > > I'm not sure why you say they're 'not directly related' to large file > support. Maybe I misinterpreted the developer comments on the upstream > bug report, but it seemed clear that large file support changes things > such that as-of-yet-unknown code in Evolution breaks in > as-of-yet-unknown ways.
What I meant is that the bug is not in largefile support but that enabling it triggers the bug, which is a different thing. In any case, bug should be fixed though. > >> That means some bugs may be present but just hidden. If there are >> crashes, hide them doesn't look like a good idea, it's better to fix >> them. > > I agree in general, but in this case I'm pessimistic about being able to > even identify, let alone fix all the problems in time for the release > given the history of the upstream bug report. Yeah but if a bug is just a lot more present when LFS is enabled, but still present without, we might have it triggered in some day, where it'll be harder to correct it. > > > What was the reason for adding --enable-largefile? Shouldn't we wait > until upstream enables it by default? The fact that 32 bit users might come to the 2/4 GB limit. I'm asking upstream advice on this one. As I said, I'm not against disabling it. I only maintain evolution because I do use it and nobody else seem to care about, and I don't use i386. So I don't really care about LFS support, but some other might. Cheers, -- Yves-Alexis -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
