sommerfeld: I've met another similar error, but for "network/bridge/build/i86/.built"

when did "gmake -n build"

head is oi/hipster , commit 00f7f4327ca7361dc28a29253f87530b26994778

vetal: that makefile hasn't been touched since 2018. likely requires fixes like the uwimap Makefile

wacki: I mean when release tag is set, in general it requires that whole things under that tag are rebuilt. And result artefacts placed to repository.

sommerfeld: Did you mean if I prepare a fix, you will apply it?

vetal_: I don't have the authority to apply it.

if you submit a well-formed pull request that fixes build breakage one of the maintainers would likely merge it.

@vetal_: We don't do that because of resource constraints. Whole OI is maintained by just a few people on a single server. Most people have only limited interests in a few packages they maintain.

Or simply put: I cannot do that because I am almost alone.

Of course you (and everybody else) can fix or enhance things and create a PR for it. If it looks good, builds and publishes for me, I will most likely merge it.

sommerfeld: Excellent!

wacki: Got it. That explains why I met errors during build. Probably I will have some time to help in that. Thanks!

Changing the default compiler to gcc-14 alone broke a lot of builds because many warnings of the past are errors now.

Then you can have all sorts of fun with new versions of a package that break others in subtle (and no so subtle) ways.

Because almost nobody but me checks PR's many problems will only be detected by users.

I have tried to wait for comments on PR's in the past but that mostly was pointless as very seldom someone looked at them.

gcc-15 (released earlier today) may break some more by changing the default to C23 - easy workaround is compiling with -std=c17 instead