On Fri, Apr 23, 2021 at 09:33:07PM +0200, Léo Le Bouter wrote:
> I knew about this but I didnt feel like telling Raghav to do yet
> another rebase. I felt like Raghav was taking on with so much already.
> The rebase was specially complicated because Raghav's commit changed
> indentation, git has bad quite bad UX for cases like these. At the time
> I had lots of things to handle also and couldnt spend lots of time on
> it myself. I didnt feel like blocking the merge of these patches for
> commit history was worth it at all. Such blocking could have hindered
> the GNOME upgrade effort even more. Thankfully now there's lots of
> energy being put to it, at the time there wasnt anyone else than Raghav
> and me.

I'm sympathetic. There is an imbalance between the work that we want to
complete, and the time and energy that we can give to it.

And in the case of GNOME, we have already fallen short of our goals
several times, having missed multiple upgrades.

I too have felt the temptation to cut corners with Git when I know that
the final result will be "okay". But Guix is not just about the final
product (a release, or a merge). We also have the --commit option to
Guix commands, and `guix time-machine`. So the Git history is important
too.

And I have also spent several hours at a time, focused on completing
(after several restarts) a complicated rebase involving dozens of
commits. And I've done that many times.

I do think that Mark is being hyperbolic about the wip-gnome branch. The
name says "work in progress" and we don't hold those branches to a high
standard. But what happened on core-updates *must not happen again*.

For a task as large as "updating GNOME in Guix", history tells me that
it has to be a group effort. In many cases, the hardest part of a
project is coordination and leadership, not coding. I hope that this
current effort continues, and that more people decide to join.