Closed maresb closed 2 weeks ago
As far as I can tell, there's currently no direct way to add a new package while keeping the rest fixed. Right? So I would just remove that section.
I think it's possible by adding the new package to environment.yml
and then run conda-lock --update newpackage
, though I haven't tested it.
Unfortunately --update
is pretty broken at the moment due to https://github.com/conda/conda-lock/issues/639. :see_no_evil:
I think it's possible by adding the new package to
environment.yml
and then runconda-lock --update newpackage
, though I haven't tested it.
That works, but also updates all the existing packages. Same without --update
Is it actually updating the existing packages or is it merely clobbering the extensions?
I'm not sure I understand what you mean by clobbering extensions. But I think it's updating. For example, I have numpy=1.25 in the lockfile, then add pandas to the environment file and run conda-lock again. After that I have numpy=1.26 in the lockfile.
Ah, thanks for the clarification, I guess it is updating the existing packages. (By "clobbering extensions" I meant replacing all .conda
builds with .tar.bz2
.)
It seems that part of #632 doesn't accurately reflect the current behavior of how existing lockfiles are used when relocking.
https://github.com/conda/conda-lock/pull/632#issuecomment-2156733823