=== schmidtm_ is now known as schmidtm
=== mamarley is now known as Guest22518
=== mamarley_ is now known as mamarley
=== yofel_ is now known as yofel
[13:51] <tseliot> mamarley, ricotz: change of plans, we need to revert the change for snaps in the nvidia packages
[13:52] <mamarley> tseliot: Already done. :)
[13:52] <tseliot> mamarley: great :)
[13:52] <mamarley> (For the PPA packages, anyway.)  I reverted it as soon as the reports of X not starting came in.
[13:55] <tseliot> yes, they're going with a better solution that shouldn't cause LP: #1589006
[13:55] <ubottu> Launchpad bug 1589006 in snapd (Ubuntu) "Failed unmounting Mount unit for nvidia support in snappy" [Critical,In progress] https://launchpad.net/bugs/1589006
[13:58] <mamarley> tseliot: What is the new solution?  I don't see anything about it in that bug.
[13:59] <tseliot> mamarley: "mvo: we just discussed the bind mount of libgl in our snappy standup and we would like to revert the bind mount unit again, sorry for the trouble, but the issues that croped up are too servere and we will use a different strategy (bind mount in the launcher in a private mount namespace to isolate all of this from the real system)"
[13:59] <mamarley> OK, so something that doesn't involve the NVIDIA package at all.
[14:12] <tseliot> yes, and that's the way I like it :P
[14:12]  * mamarley too :)
[14:13] <mamarley> To be honest, I had a bad feeling about that patch anyway.
[16:51] <ricotz> mamarley, note, next time when introducing pre/post/inst/rm changes let them sit for a while before porting to other series
[16:51] <ricotz> ;)
[16:52] <mamarley> I had already decided that :)
[16:52] <ricotz> this really created mayhem for some users
[16:53] <mamarley> I have given a detailed explanation of how to correct the problem for everyone who has asked.
[16:53] <ricotz> even Mark shouldn't demand it to be fixed without real testing ;P
[16:53] <ricotz> mamarley, thanks for that, just noticed another message on the list
[16:54] <mamarley> I also think maybe there should be an argument for dpkg that causes it to ignore errors in prerm/postrm scripts.
[16:54] <ricotz> absolutely not, those script should have been tested more careful
[16:55] <ricotz> or have it swallow errors with || true
[17:01] <mamarley> Anyway, I will definitely wait for the -proposed testing to be done before applying non-critical patches like that in the future.
[17:04] <ricotz> +1