Hi,
On Thu, Aug 14, 2025 at 10:07:02AM -0700, Otto Kekäläinen wrote:
Please remember to check if your package has open MRs *at least once*
before the next upload!
This is good advice to put into an upload workflow, yes. For people who
have been around in Debian for longer than ten years this is a
significant change that needs to grow into finger memory. Please be a
bit more forgiving towards your fellow developers who might not be as
brilliant as you are.
I have now witnessed several cases where a maintainer blatantly
ignored the MRs their package received. It is demotivating to new
aspiring Debian contributors to put in significant effort to learn the
complexities of Debian packaging and submit an improvement, only to
see that the maintainer two months later didn't look at the MR at all,
and instead implemented the same change themselves, and the
submitter's work essentially got wasted.
I think it is well inside a package maintainer's prerogatives to not
accept a MR completely. Sometimes it is just faster to do one-liner
changes in an already checked out repository than going through the
sometimes arcane and complicated¹ Gitlab methods if there are issues
with the suggested changes.
When I am the person submitting an MR, I sometimes perceive comments
like "I don't like this, please change the wording to this, and do
better indentation there" like as if a teacher is talking down to their
students.
I would be grateful if the receiver of my MR would fix my MR in a way
that suits their needs. I am pretty sure that this workflow is saving
time on both the MR's submitter and receiver's side if the change is
trivial enough. Other people seem to take anything that doesn't end with
a click on the "Merge" button in Gitlab as a personal offense. I
disagree here. We should all assume good faith and accept actions of the
other side that move in to the right direction with appreciation and not
with contempt.
What I am trying to say is that not everybody ticks like you are. The
respective other side of the MR is acting in the same good faith like
you are, with very similiar results, just not in 100 % exactly the way
YOU find right. I would appreciate you trying to accept that other
people work differently then you are and trying to not lash out towards
them like you are doing in the message I am replying to.
I would not have written this if there hasn't been a pre-story between
the two of us.
A few days ago, you submitted a trivial but important MR to one of my
packages. I agreed with two of your three changes, but disagreed with
the third. I quoted the relevant part of policy in a review message (I
still think that I understand the Vcs-Git headers in debian/control the
correct way).
I politely thanked you for spotting my stupid omssions, tried to express
my appreciation and - in good faith - quickly pulled the two changes I
agreed with and created and pushed the two commits manually (of course
crediting you for your work in the commit message) instead of letting
them wait in the MR for us to sort out the third issue.
I promptly received a beating simliar to what you have written above,
saying that I demotivate you and that I am being disrespectful. I acted
in good faith, my only interest was to reduce work on both sides of the
MR² and to speed things up. I absolutely didn't want to be disrespectful
and I surely didn't want to demotivate you. It was me asking your for
advice in the first place and I appreciate your help (albeit still not
understanding what exactly you did to help and thus unable to reproduce
this on my other packages that will need the same procedure in the
future).
Receiving this lashing demotivated _me_ from working on Debian for the
rest of today's afternoon. I spent that time playing with and hugging my
adorable cats instead. Thank you for making this possible.
Greetings
Marc
¹ I still don't know how do do minimal changes to commit messages from
MRs like adding a "Closes:" line short of git rebase -i after doing the
merge. There must be an easier method to do that from the web interface
but alas, I didn't find it yet.
² I am interested to hear what additional work I caused on your side by
quickly pulling parts of your MR to the target branch manually. The
result is the same: Your changes are in debian/latest, the next package
version will be better because of that, you're credited and appreciated
in the commit message, it just didn't happen in your way. Nothing was
lost, just a bit of my self-confidence in interacting with my fellow
DDs.
--
-----------------------------------------------------------------------------
Marc Haber | "I don't trust Computers. They | Mailadresse im Header
Leimen, Germany | lose things." Winona Ryder | Fon: *49 6224 1600402
Nordisch by Nature | How to make an American Quilt | Fax: *49 6224 1600421
