Little quirk when using OVERLAY_CACHE_METHOD=sqlite*. When an overlay has been corrupted, layman -d <overlay> fails to delete it. To restore the overlay, you need to call eix-layman delete <overlay> before. Sorry no useful log error, except the message * Error was: Reproducible: Always Steps to Reproduce: 1.create a corrupted overlay 2.layman -d <overlay> 3. Expected Results: when deleting an overlay, layman should switch to eix-layman when OVERLAY_CACHE_METHOD='sqlite*' has been set. Or at least a useful warning. workaround, example: overlay x11 # layman -d x11 * Error was: # eix-layman delete x11 # eix-layman add x11 not sure if that still is necessary, # layman -a x11 version affected: app-portage/layman-1.3.0_rc1-r1, and probably other
I have difficulties understanding this bug report. What do you mean by "corrupted"? How do you know that "layman -d <overlay>" failed to delete it - any other indicators besides the error from layman?
(In reply to comment #1) > What do you mean by "corrupted"? you cannot sync the overlay and you cannot add/delete it either. * Errors: * ------ * * Failed to sync overlay "x11". * Error was: Syncing overlay "x11" returned status 256! > How do you know that "layman -d <overlay>" failed to delete it - Sorry, I made a lot of try and fail without taking notes. I cannot tell exactly the error returned, but one of them was a short and terse "* Error was:" Moreover, I deleted the x11 directory flushed all the references in make.conf, cache_*.xml and overlay.xml, but only the command 'eix-layman delete x11' was successful. > any other indicators besides the error from layman? The "layman -S" status was 256 but I cannot confirm it.
Jimmy, I have now commited code to Layman that may fix your problems: http://layman.git.sourceforge.net/git/gitweb.cgi?p=layman/layman;a=commitdiff;h=c654e79fd18adb0e9f1cae6b783f8c81b39dda61 http://layman.git.sourceforge.net/git/gitweb.cgi?p=layman/layman;a=commitdiff;h=7a4e280313f57f714c499aed2566c93744d60851 The easiest way to check it out would be to unmask and emerge =app-portage/layman-9999, the live ebuild of Layman. Please let me know how it works for you.
(In reply to comment #3) > Jimmy, I have now commited code to Layman that may fix your problems: > http://layman.git.sourceforge.net/git/gitweb.cgi?p=layman/layman;a=commitdiff;h=c654e79fd18adb0e9f1cae6b783f8c81b39dda61 > http://layman.git.sourceforge.net/git/gitweb.cgi?p=layman/layman;a=commitdiff;h=7a4e280313f57f714c499aed2566c93744d60851 > > The easiest way to check it out would be to unmask and emerge > =app-portage/layman-9999, the live ebuild of Layman. Please let me know how it > works for you. > I installed it this morning (1am) commit 7a4e280313f5. With about one upgrade per day we will soon see if the issue is over. Let me all this week to try it out. Thx
To be sure: are you aware that live ebuilds never update themselves because the version is not bumped as with usual ebuilds? So you have to run # emerge -av =app-portage/layman-9999 manually repeatedly. Ignore me if that's nothing new.
(In reply to comment #5) > To be sure: are you aware that live ebuilds never update themselves because the > version is not bumped as with usual ebuilds? So you have to run > > # emerge -av =app-portage/layman-9999 > > manually repeatedly. Ignore me if that's nothing new. > I did know that, thx anyway. I have used layman-9999 and eix-sync with sqlite back end lately and the issue has gone. Since eix is linked to layman, could you please read bug #303014? Unfortunately, I was not able to convince Martin of the benefit to use eix-sync without root rights. Using -e isn't convenient because it didn't flush logs. Thx
Thanks for testing! Layman 1.3.1 has just been released, these fixes included. This is a good occasion to get off the live ebuild again I would say. Just my 2 cents.
