Reputation: 2631
I work with a software vendor who generally distributes code as git am
-like patches in .patch
files, which works reasonably well.
Occasionally, instead of providing a commit as a patch file, which patches a source file (i.e. .c
), they tend to patch up one of the older patch files.
In case it's not clear what I'm describing, the diff often looks like this:
second.patch:
diff --git a/first.patch b/first.patch
index 4a2f020..2ea53fe 100644
--- a/first.patch #notice this isn't a source file, but rather a older patch provided
+++ b/second.patch
<whatever>
I'd assume there should be a very good reason, since they do it very rarely, and it's very annoying.
To me, I see a huge downside of having to maintain older .patch
files to apply the new one, which should be completely unnecessary.
Is there a reason why one would do this?
Upvotes: 2
Views: 57
Reputation: 2631
Turns out it's their way of doing git commit --amend
so that they don't actually modify the older patch files
Upvotes: 1