[October 25] #2225
Byron
announced in
Progress Update
[October 25]
#2225
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
"By this time I think it's fair to say that I am mostly busy with GitButler, and
gitoxidedoesn't receive more than basic maintenance and the occasional fix."That was what I started out as in the previous month, and it's still very true today.
Important fix to blob merges
Previously, when merging the bytes of blobs, there were two conditions which effectively auto-merged what should have been a conflict. This in turn made
gitoxidereport a merge-success where it should have conflicted, just like Git.There was one existing test that covered it, but the expectation was wrong, and I didn't notice the signal in quite an indirect test for another purpose.
Now there is specific tests for this case, and the fix was just to remove this special case. If memory serves, these were a remnant of an intermediate state of the implementation and I never dared removing them.
Thanks to Caleb Owens for reporting this issue!
The MSRV is now Rust 1.82
In the drive to get closer to removing
once_cellfrom the dependency tree, we have upgraded the minimum supported Rust version to 1.82, which is also used by Helix, entirely without coincidence.Now I am looking forward to eventually hitting Rust 1.90, which should allow the crates to switch to Rust Edition 2024 🎉.
Community
Meta Open Source donation on OpenCollective
By the end of August, well within the reporting period of the previous month, Meta Open Source endowed Gitoxide with 20.000USD. This generous donation is now used as maintenance fund, allowing me to get paid for maintenance which doesn't have to come from my already non-existing free time anymore. ("free time" here is defined as time were I can do just what I want 😁).
Thank you, Meta Open Source, for making such a meaningful donation!
SHA256 preparations
A small but meaningful contribution by Lorenz Leutgeb helped pave the way for SHA256 support, by making the current SHA1 implementation optional, and the respective
gix-hash::Kindenum non-exhaustive.This also allows to one day create builds that only support SHA256 as safer (and much faster) alternative.
gix commit signLast but not least, and kindly contributed by Christoph Rüßler, we now have the
gix commit signsubcommand, which does exactly like it says: it signs a commit using GPG, like Git would, without implementing all the configuration variables and special cases that Git would.As such, it's a prototype, but also a first step towards implementing signing just like Git would, so it can eventually move into the
gixcrate 🎉.Gix in Cargo
There is nothing new here, but let's keep the horizon active:
Cheers
Sebastian
PS: The latest timesheets can be found here (2025).
Beta Was this translation helpful? Give feedback.
All reactions