6:19 p.m.
kellin opened a new pull-request against the project: `releng` that you are following:
``
DNF & Python3 Updates for Critpath.py
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
12:48 a.m.
ignatenkobrain commented on the pull-request: `DNF & Python3 Updates for Critpath.py`
that you are following:
``
why? dnf is supported on py2
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
9:48 a.m.
kellin commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that you
are following:
``
@ignatenkobrain this is a request from @mohanboddu to support both YUM and DNF on
python2.
The rationale is that older releases were calculated with YUM and, given there are
differences between the packages YUM and DNF return as dependencies, we should maintain
YUM compatibility to use against older releases.
We will use this under Python2 with the --dnf flag on newer releases.
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
10:29 a.m.
ngompa commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that you
are following:
``
@kellin: I would change this to default to DNF for both Python 2 and Python 3, and have a
`--yum` flag for older releases. If the script is slightly release aware, then you could
also add logic for "if fedora <= 26 or rhel/epel <= 7, use yum". The goal
should be to be forward-looking by default.
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
11:22 a.m.
kellin commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that you
are following:
``
I have updated per @ngompa 's request.
FWIW: as far as I can tell this script is not used against EPEL. If it should be; then
that's outside the scope of fix for this PR and we can discuss that as a future
feature.
@mohanboddu - please take a look as well and let me know if this works for you.
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
12:03 p.m.
ngompa commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that you
are following:
``
`list(range(12,28,1))` can generate your list in a simpler chunk of code.
c.f.: https://docs.python.org/3/library/functions.html#func-range
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
12:19 p.m.
puiterwijk commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that
you are following:
``
You mean you tested this with python1? Awesome!
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
12:20 p.m.
puiterwijk commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that
you are following:
``
Even better, since you're looping over it and not returning anything, you can leave
out the list() part and not have the actual full list object in memory.
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
12:23 p.m.
puiterwijk commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that
you are following:
``
Just move this all out into a expand_critpath_dnf please?
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
12:23 p.m.
puiterwijk commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that
you are following:
``
Let's rename this to expand_critpath_yum?
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
12:24 p.m.
puiterwijk commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that
you are following:
``
I'm pretty sure that this line is not pep8 valid. Just move the `(count,` to the next
line.
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
12:26 p.m.
puiterwijk commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that
you are following:
``
Why this renaming? This makes it harder to read...
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
12:27 p.m.
puiterwijk commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that
you are following:
``
Why.... Why not just use `import dnf.rpm` and then use `dnf.rpm.foobar()`?
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
12:27 p.m.
puiterwijk commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that
you are following:
``
This line is missing a space
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
12:29 p.m.
puiterwijk commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that
you are following:
``
This should probably be `if not using_dnf`.
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
12:30 p.m.
puiterwijk commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that
you are following:
``
Why this craziness? Why not just `if yum.__version_info__ <= (3, 2, 24)` ?
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
12:31 p.m.
puiterwijk commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that
you are following:
``
Maybe actually error out?
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
12:32 p.m.
puiterwijk commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that
you are following:
``
(In which case you should also move this down to where that's defined probably)
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
12:39 p.m.
puiterwijk commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that
you are following:
``
Also, why do you have this import? You don't seem to use it.
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
12:40 p.m.
puiterwijk commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that
you are following:
``
Also, random idea, but why not just get a directory list and use that, so that we
don't have to update this in the future?
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
12:44 p.m.
ausil commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that you
are following:
``
I would not call this a hack
fakearch = {'i386':'i686', 'x86_64':'x86_64',
'ppc64':'ppc64', 'ppc':'ppc64',
'armhfp':'armv7hnl', 'aarch64':'aarch64',
'ppc64le':'ppc64le', 's390x':'s390x'}
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
12:51 p.m.
puiterwijk commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that
you are following:
``
Also, general comment: it would be great if you added `from __future__ import
print_statement` at the top of the file: that forces all print's to be function
calls.
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
12:53 p.m.
ngompa commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that you
are following:
``
I suggest using the real arches, and using the conversion functions in yum and dnf to get
the basearch.
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
10:07 p.m.
kellin commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that you
are following:
``
Updates have been made.
@ngompa & @puiterwijk - I changed the static release list to a range generator.
@puiterwijk - renamed functions to expand_X_critpath per request and moved dnf code into
its own spot. I didn't do this because originally this had caused major breakages but
it looks like somewhere along the way I cleaned up what had caused those in the first
place.
@puiterwijk - I had renamed the import to tempdir because originally had been using
TemporaryDirectory from Python3 and was renaming on import to use the same thing in two
places. Eventually this gave way to the form it was in, so the renaming was superfluous.
Changed it back to a straight import of mkdtemp
@puiterwijk - removed the oddly renamed unused import - it was a relic borne out of late
night frustration sometime last week that I didn't remember to remove.
@puiterwijk - the yum version comparison now uses tuples, thanks for the suggestion though
it wasn't something I'd changed in this PR, just moved the code to another spot.
@ausil - I removed the verbage for calling it a hack, however, it's not a good thing
to use a hard coded list of releases.
@puiterwijk - added the __future_ import per request.
There were several other comments above: with regard to attempting to use arch/basearch
conversion functions and using a directory listing those are items that will be resolved
in a fix that is coming after bodhi freeze.
The main point of updating this script is just to get DNF dependency solving in place
before Bodhi freeze so that the functionality is there. The longer term fix required more
work/testing after @mohanboddu 's request and I didn't feel comfortable trying to
shoehorn it all in while attending FLOCK and travelling with the time constraints.
Thanks for all the reviews - if there's anything I didn't cover please let me
know.
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
10:09 p.m.
ngompa commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that you
are following:
``
Erk... you'll want to change `28` to `27`. We have no release tree for Fedora 27
yet!!
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
10:14 p.m.
kellin commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that you
are following:
``
@ngompa - changed, though in usage that would just check to see if the user's input to
the script was sane and, if it wasn't, it'd still be caught in the error handler
for DNF and pop a message to the user.
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
10:15 p.m.
ngompa commented on the pull-request: `DNF & Python3 Updates for Critpath.py` that you
are following:
``
Looks good to me :wine_glass:
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
1:35 a.m.
ignatenkobrain commented on the pull-request: `DNF & Python3 Updates for Critpath.py`
that you are following:
``
you shouldn't really install all groups at the same time because in theory some of
packages might conflict. I would rather do group_install(), resolve(), reset() and then
for next group
``
To reply, visit the link below or just reply to this email
https://pagure.io/releng/pull-request/7007
2424
days inactive
2426
days old
rel-eng@lists.fedoraproject.org
27 comments
5 participants
participants (5)
-
Dennis Gilmore -
Igor Gnatenko
-
Kellin
-
Neal Gompa
-
Patrick "マルタインアンドレアス" Uiterwijk