On Wednesday 14 June 2006 14:23, David Härdeman wrote: > I tested it until I could find no more bugs before I committed it to > SVN. I also posted to the list asking for review and comments before > committing. The changes for partman-lvm (essentially rewriting and > merging lvm-tools.sh from lvmcfg into partman-lvm) were fairly large, > so I'm not that surprised that bugs would be found. Yes, I saw that. No blame intended. I was just a bit surprised at how fragile the new code turned out to be. You probably know that I've basically been on VAC for the last 5 weeks and therefore was not able to respond to your request. I want to make this a priority now though. > I've been on vacation for a couple of days, I'll try to take a look at > the bugs later today Great. > Curious, I seem to recall testing this last week. Perhaps I used a > different version of the lvm tools since I use unstable as the d-i > suite for my testing images. Anyways, I'll take a look at it I used the daily images which means all udebs are from unstable! > > I would expect there are probably other, similar errors in the code. > > Not sure what your point is...? That we may need to check what exit codes the LVM tools give, especially if we use them in tests that are not hit very often and in "sanity checks" (which normally only give one result and thus the error case is never really tested). Thanks for your work on LVM and RAID. From what I've seen the functionality is now much more professional, but I expect it will need a lot more testing to get it as stable as the previous version was.
Attachment:
pgpnBi9Bw7DS8.pgp
Description: PGP signature