Skip to content
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

LANG: python 3.10 update (amplicon) #367

Merged
merged 9 commits into from
Sep 18, 2024

Conversation

lizgehret
Copy link
Member

local install of all external dependencies was successful, so now it's time to see what breaks!

@lizgehret lizgehret self-assigned this Sep 6, 2024
@lizgehret
Copy link
Member Author

lizgehret commented Sep 6, 2024

Okay so current issue here is from q2-fragment-insertion's pin of dendropy. This pin is because sepp depends on an old version of dendropy (because it imports a private function that no longer exists in the current versions) which is (of course) not compatible with python 3.10. We'll need to figure out what to do about this because this blocks q2-fragment-insertion from being included in the 3.10 update.

@lizgehret
Copy link
Member Author

lizgehret commented Sep 10, 2024

Okay so it looks like the latest version of sepp resolved this issue (less than two weeks ago!) and builds are now up on bioconda. Paired PR in fragment-insertion that removes this pin - we'll see if this works!

@lizgehret
Copy link
Member Author

Another paired PR in alignment - this punts the mafft pin back to alignment's recipe, using the min pin previously present

@ebolyen
Copy link
Member

ebolyen commented Sep 11, 2024

Wow, it looks like it practically worked!

Now that you have a solve, I would recommend putting mafft back into the seed env with the new pin (whatever that ended up being), and also maybe we should do the same for dendropy? That way we can control all of these in one place.

(I realize that you needed the >= operator to try and solve for mafft, which wouldn't have worked in the seed env... probably?)

@lizgehret
Copy link
Member Author

lizgehret commented Sep 11, 2024

PR to address test failures in vsearch (very minor): qiime2/q2-vsearch#99

I'll try adding mafft as a min pin in the seed env @ebolyen - conda's docs do state that pin ranges are supported, and that might be better for this situation. I'm debating between pinning sepp and dendropy, since sepp is the root offender. Now that I've written that out, I think I'm leaning towards a min pin for sepp - that way we're just ensuring we have at least the build that's fixed that old dendropy dependency.

EDIT: so even though conda supports pin ranges, it doesn't seem to be easily compatible with the jinja templating for package versions (within each individual package) - so i've just set the pins for mafft and sepp, and we'll just fuss with them if/when they stop working.

@lizgehret
Copy link
Member Author

We've got a working build! 🥳

@lizgehret lizgehret marked this pull request as ready for review September 17, 2024 22:55
@lizgehret lizgehret merged commit f78e465 into dev Sep 18, 2024
@lizgehret lizgehret deleted the Language-python310-update/amplicon branch September 18, 2024 03:22
@lizgehret lizgehret removed their assignment Sep 18, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
No open projects
Status: Completed
Development

Successfully merging this pull request may close these issues.

DEP: q2-fragment-insertion can't be upgraded to Python 3.10, due to pinned SEPP dependency on Python <=3.9
3 participants