-
-
Notifications
You must be signed in to change notification settings - Fork 15.1k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
python3Packages.cassandra-driver: fix build by packaging new dependencies #100989
python3Packages.cassandra-driver: fix build by packaging new dependencies #100989
Conversation
@risicle Thanks a lot for picking this up! |
8ef3d69
to
0a21340
Compare
@ofborg build python37Packages.cassandra-driver python38Packages.cassandra-driver |
getting a failure from geomet, there is a 0.2.1.post on pypi
|
Hmmyes. There's no tagged release for the 3.8 fix, but I might just pull it in as a patch... |
0a21340
to
96a82da
Compare
@ofborg build python38Packages.cassandra-driver |
@risicle please solve the merge conflict. |
Ping @risicle. Do you want to continue on this? |
Sure, I just have a few things on my TODO list - currently working on a second attempt of #93083 |
b3d7e66
to
51147a1
Compare
Updated, retested macos 10.14, non-nixos linux. @ofborg build python37Packages.cassandra-driver python38Packages.cassandra-driver python39Packages.cassandra-driver |
I don't get it - why didn't you request these changes when you made your first pass? More to the point, how many times do we have to iterate this process until I've satisfied every last one of your personal style preferences? From the fact that I've had this on my TODO list since October, you should see I don't have a lot of time for this stuff, but you'd prefer to keep it bouncing. |
I think we should be good to go after this. I am human and I can always miss something and notice it later. I think that's better than not noticing it at all. Also in the last two months I learned a lot of things and nixpkgs evolved. For example stdenv.lib got deprecated and is no longer allowed to get re-introduced.
I can't know if you don't have enough time ,you forgot it or just missed the notification. I let it bounce once to you which I think is reasonable to do especially after two months. |
This is a semi-automatic executed nixpkgs-review with nixpkgs-review-checks extension. It is checked by a human on a best effort basis and does not build all packages (e.g. lumo, tensorflow or pytorch). Result of 9 packages built:
|
This is a semi-automatic executed nixpkgs-review with nixpkgs-review-checks extension. It is checked by a human on a best effort basis and does not build all packages (e.g. lumo, tensorflow or pytorch). Result of 9 packages built:
|
51147a1
to
04a96e3
Compare
My point is that being overly picky about stylistic things really just puts off contributors. If something straightforward (like perhaps deciding to upstream one of your team's fixes after work for the good of the community) is going to become a multiple week saga bouncing back and forth talking about newlines and semi-colons, people just aren't going to bother because they have other things to do that feel a lot more productive. |
I agree with that. Getting a PR into nixpkgs is so much work already, even for smaller contributions. Formatting issues should be addressed by an automatic formatter. They shouldn't cost our precious human's time. |
I am not overly picky. Most of the requests are to have a somewhat uniform style to make treewide changes in the future way easier by not having to many variants of essentially the same thing. I did not order your inputs or sort your phases. Run nixpkgs-hammer on your package and you should see what overly picky looks like.
It isn't if you follow the contributing guide. Take a look at fabaff PRs since December and a lot of them got merged without any review comments or only very small changes.
There is no formatter that is one stop and does everything. nixpkgs-hammering and nix-fmt should be a good option but they do not cover everything. I am not familiar with programming formatters so I can't improve the situation in a meaningful amount of time. |
@SuperSandro2000 What exactly is the contributing guide? https://github.com/NixOS/nixpkgs/blob/master/.github/CONTRIBUTING.md is the only thing I know in that direction, and it doesn't give me enough information to submit PRs that can just be merged as-is. I'd like to learn more. |
There is also https://github.com/NixOS/nixpkgs/tree/master/doc/contributing and the language specific doc. |
This is a semi-automatic executed nixpkgs-review with nixpkgs-review-checks extension. It is checked by a human on a best effort basis and does not build all packages (e.g. lumo, tensorflow or pytorch). Result of 9 packages built:
|
Motivation for this change
ZHF: #97479
This is heavily based on @turion 's #91662, but I ended up modifying it so much I thought I should just post it separately rather than give him a long list of things to fix. Though I've still left @turion as the maintainer of these - I hope that's ok @turion?
To get
NIX_REDIRECTS
working I had to addlibredirect.so
toLD_PRELOAD
, and also add a dummy/etc/resolv.conf
for some later tests. Also converted topytestCheckHook
and disabled some tests causing trouble for various reasons.All of the packages had to be switched to the github source because the tests weren't included in any of the pypi tarballs.
gremlinpython
's tests do actually run if we override some version constraints and disable some sets of tests which want to connect to a running tinkerpop server.Things done
sandbox
innix.conf
on non-NixOS linux)nix-shell -p nixpkgs-review --run "nixpkgs-review wip"
./result/bin/
)nix path-info -S
before and after)