On 12-07-29 at 05:07pm, Didier 'OdyX' Raboud wrote: > Le dimanche, 29 juillet 2012 15.17:39, Jonas Smedegaard a écrit : > > On 12-07-29 at 02:20pm, Didier 'OdyX' Raboud wrote: > > > A nice solution to this bug is IMHO to replicate in > > > ghostscript-cups what has been done for most packages shipping > > > Cups drivers: transform the postinst code into a dpkg trigger and > > > let the Cups postinst do the job "as cups". > > > > Agreed, that's a nice solution. > > > > Didier, you listed yourself in that change. Please feel free to > > join us in maintaining ghostscript: Add yourself as uploader and do > > the change yourself. > > Thanks for the offer, but sorry, I already have too much on my plates > currently to add Ghostscript to the pile.
Fair enough :-) > > If you do, then please split into several commits: First actual code > > changes, and then (semi-)automated changes (e.g. using "git dch -a" > > to update changelog). That way it is much easier to later revert a > > change or cherry-pick across branches. > > Hrm. I usually commit without changelog entry, using pseudo-headers > and $(git dch --full --meta --release) and hand-edit the changelog, > and commit the release changelog. Oh, sorry: I did not mean to imply that your commit style was currently bad. I just described my preferred commit style in case you did choose to commit yourself *and* did not use that commit style already. But from your comment now it seems our commit styles are similar - or even that yours is even more elegant than mine, so thanks for sharing :-) > > Alternatively I can proof-read and apply the patch, but will then > > list it as done by me, crediting you with a "thanks" after the > > trailing bug closing hint. > > Whatever is fine for me, as long as it gets in. Btw, this machinery > has mostly been engineered by myself and Till (with him doing most of > the cleanup and polishing of my ugly code), see #637978, so there's > not much point in attributing this small change to me. You did bring this elegant approach to my attention, which I appreciate. I'll credit you both, then (unless you insist on not being mentioned). > > For Wheezy we should probably aim for the uglier but much simpler > > solution. the "master" branch is not intended for Wheezy, I will > > use a separate "master-wheezy" for that. > > I think I disagree. The patch I propose makes ghostscript-cups rely on > cups' postinst which is already proven working by more than 14 > packages, all of them already in Wheezy. I think this is a patch the > Release Team can accept and a patch that makes Wheezy a better release > by reducing useless code duplication in maintainer scripts. Oh, ok. Would you mind file the bugreport requesting freeze exception, when I have uploaded it to unstable? I am quite lousy at arguing such cases for the Release Managers :-/ - Jonas -- * Jonas Smedegaard - idealist & Internet-arkitekt * Tlf.: +45 40843136 Website: http://dr.jones.dk/ [x] quote me freely [ ] ask before reusing [ ] keep private
signature.asc
Description: Digital signature