IRC channel logs

<rekado>civodul: I did another reproducibility talk at the MDC, yes. How did you hear?

<rekado>I'm a little concerned about the unintended effects that serializing our work on branches has had on timely updates.

<rekado>the r-team branch has been ready for weeks. It was ready even before R 4.4.0 was released. Now it also include the upgrade of R 4.4.0.

<rekado>this branch is now a lot longer-lived than it used to be, because it's waiting in the queue

<rekado>the python-team branch on the other hand is never truly ready (it's just decided to be good enough until the next iteration), and it's currently holding up all the other branches.

<rekado>not great.

<civodul>rekado: hi! i saw your slides in the repo :-)

<civodul>i share your concerns

<civodul>not sure how to best handle that

<civodul>you could still ask whoever is first to merge before them

<rekado>ah, right, this time I did actually push the slides to the repo!

<rekado>it's currently the python-team branch that's waiting to be merged (and it seems that I'm sorta driving that one), but r-team is not the next after that, so jumping the queue would certainly make other people unhappy.

<rekado>it's annoying because all the R stuff is pretty self contained

<rekado>this time it's exceptional because of the openblas update

<rekado>but this procedure has seriously slowed down my ability to provide timely updates to R things.

<rekado>by the time I finally get to merge the branch it will be outdated again

<rekado>there has been a new Bioconductor release in the meantime, for example.

<civodul>rekado: yeah, i think we need to discuss it and adjust it somehow

<civodul>not sure exactly how to change it, but i agree that it’s not efficient

IRC channel logs

2025-04-01.log