Hello, I will release <#C038XMTQWBT|genomeassemble...
# pipeline-maintainers
n
Hello, I will release #genomeassembler today exceptionallyfastparrot the dev -> master merge is some 500 commits, should I do a squash merge or keep the commit history?
aaahhhh 1
exceptionallyfastparrot 3
j
Most people just merge, but I think it's up to you
👍 1
m
dev -> master, no squash merge
anything -> dev go squash merge
n
Ok
m
That's my opinion
We haven't came up with a proper rule on the nf-core side for this yet
👍 1
You are free to experiment
n
I am fine with doing a normal merge
m
but that seems sensible enough for me
n
Ok, another thing that I noticed on the release checklist is editing the description on github, I dont think I can do that on the nf-core repository
j
Now?
I've bumped you explcitily up to a maintainer of teh repo
with great power
n
Thank you!
I have create a release, whats the next step for zenodo?
j
We do that bit 🙂
n
Do you also update README.md?
👍 1
j
please hold
n
Amazing, thank you
j
Thank nf-test taking ages to run tests😎
Its uploading to Zenodo, will need to wait a bit
ai thumbsup 1
Here we are, for your reference: 10.5281/zenodo.14986999
n
Thanks!
j
I'll update your master branch now (README and nextflow.config) and then you can merge master into dev to have them in there
I'll let you know once udpated
Oops sorry, the canonical is 10.5281/zenodo.14986998
OK nextflow.config and README updated, I'll switch your default branch to master now
n
Thanks!
..6998 is for the pipeline, 6999 for release 1.0.0?
party this 1
j
Ok all done, you can merge master into dev
beer parrot 1
n
✅ 1
The nf-core website updates automatically at some point?
m
yes, I think it's twice daily
this 1
👍 1
n
One more thing: I am not receiving any notifications for issues opened on nf-core/genomeassembler and I do not know where / how I can enable those?
j
image.png
Have youcheked trhe 'watch' settingsw?
n
No, I havent, thanks!
m
> should I do a squash merge or keep the commit history When I familiarise myself with the code-base of any open-source project, I love using tools like GitHistory to see the edits sequentially and get an overview which parts are functionally related. That of course only works with intermediately sized commits. A single commit for a whole pipeline is useless and so are umpteenth
lint
,
lint
,
lint again
,
typo
,
aaaahg
,
revert
,
nooo
revert again
..
I quit my job
commits. Those, I would prefer being squashed selectively before being merged to
dev
. When I develop myself, I will often use
git commit --amend
on my feature branch (possibly with
push --force-with-lease
) until the edits amount to a functional and logical unit. Like this, I can make quick intermediate commits while building to have a fallback and still not push all of them to
main
.
n
That seems like a totally reasonable and good approach, but sadly I didnt do that and 80% of commits are tiny and felt like
aaaaahg
but i tried to give them somewhat more professional messages 😉
🤣 2