On Wed, 2017-02-15 at 14:32 +0000, James Hogarth wrote: > On 14 Feb 2017 11:24, "James Hogarth" <james.hogarth@xxxxxxxxx> > wrote: > > > > On 14 Feb 2017 09:28, "Tomas Tomecek" <ttomecek@xxxxxxxxxx> wrote: > > Quoting James Hogarth (2017-02-13 18:02:23) > > On 13 February 2017 at 16:40, James Hogarth <james.hogarth@xxxxxxxx > > m> > > wrote: > > > On 13 February 2017 at 15:36, Tomas Tomecek <ttomecek@xxxxxxxxxx> > > > wrote: > > > > Hello, > > > > > > > > am planning to update package python-docker-py from 1.x series > > > > to 2.x > > series in > > > > rawhide. The update should happen rather soon, so we're sure it > > > > gets > > to F26 (and > > > > hence we can have docker-compose 1.11 in F26). The reason this > > > > is > > important is > > > > that 2.x is not backwards compatible with 1.x. Here is a list > > > > of > > breaking > > > > changes: > > > > > > > > https://docker-py.readthedocs.io/en/stable/change-log.html#b > > reaking-changes > > > > > > > > > > > > Here is a list of affected packages: > > > > > > > > $ dnf repoquery --whatrequires \*docker-py > > > > atomic-0:1.15.2-2.fc26.x86_64 > > > > docker-compose-0:1.9.0-3.fc26.noarch > > > > flr-0:0.0.1-1.fc26.noarch > > > > python-atomic-reactor-0:1.6.19-5.fc26.noarch > > > > python-dockerpty-0:0.4.1-4.fc26.noarch > > > > python2-docker-squash-0:1.0.5-3.fc26.noarch > > > > python3-atomic-reactor-0:1.6.19-5.fc26.noarch > > > > python3-docker-squash-0:1.0.5-3.fc26.noarch > > > > python3-dockerpty-0:0.4.1-4.fc26.noarch > > > > python3-sen-0:0.5.0-1.fc26.noarch > > > > > > > > > > > > Maintainers of these packages, shall I help you with porting to > > docker-py-2? > > > > > > > > > > > > (This is the third time I'm trying to send this e-mail, this > > > > time > > without CCs; > > > > sorry if spamming) > > > > > > > > Since I'm resending, I already managed to rebase the package > > > > and all > > the changes > > > > are in dist-git. Will submit a build once we are sure the > > > > update > > doesn't break > > > > anything. > > > > > > > > > > > > > > Since this is a breaking change in the module and upstream have > > > formally renamed from python-docker-py to python-docker might I > > > suggest that it is more appropriate to issue a fresh package > > > review > > > for python-docker (which can then in due course perhaps obsolete > > > python-docker-py or at least just retire it without obsoleting) > > > would > > > be more appropriate? > > > > > > This will break the scripts of anyone using the docker python > > > module > > > after all ... > > > > > > Related to this are you aware of any plans to rebase from 1.9.0 > > > in RHEL > > extras? > > > > Further to this looks like the F26 mass rebuild picked up this > > change > > by accident ... > > > > https://koji.fedoraproject.org/koji/buildinfo?buildID=854128 > > > > At the very least you need a self-contained F26 FESCO change ticket > > for > > this ... > > > > The correct procedure, given upstream has renamed, is to epoch bump > > this back to 1.10 and then to do a package review for python-docker > > though ... > > _______________________________________________ > > devel mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxx > > To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxx > > Oh! > > Thanks for letting me know. > > The rename is actually a really good point. What bothers me is that > upstream git > repo is still named docker-py, same for documentation. PyPI package > is named > docker, as you pointed out. They even check if the 1.x series is > installed: > > https://github.com/docker/docker-py/blob/master/setup.py#L12 > > So let's rename then. I will submit new review. But first let's fix > rawhide. > > Regarding RHEL extras: not sure; first I want to make it happen in > Fedora, > then > wait for everyone to get used to it (especially atomic command). The > thing > is > that plenty of tools and scripts in RHEL ecosystem use this library. > I don't > want to break those. If this bothers you, feel free to open bugzilla > and we > can > discuss there. > > > Thank you, James. > > > Tomas > > > No problem and great stuff > > Feel free to cc me when you submit the review and I'll be happy to > pick it > up. > > I have scripts that use this in my workplace so useful to know when I > might > need to update them :) > > > > I see you've submitted the review request > > I've grabbed it and will do the review this evening. Can you stop using HTML please? It makes all messages just to loose quoting which makes messages unreadable... > _______________________________________________ > devel mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxx > To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxx -- -Igor Gnatenko
Attachment:
signature.asc
Description: This is a digitally signed message part
_______________________________________________ devel mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxx
