Frank Dana

@CtrlZmaster > Maybe I also misunderstood the boolean dependencies in rpm... The situation is that we also require rackup if there is rack >= 3. But if rack is older,...

(Note: I'm not saying the `if` form _wouldn't_ work, in fact I think it probably would because of the `Requires: rubygem-rack` that came before it, like with the second sstp-client...

~I rebased the PR branch onto `main`, hopefully that will solve the unsigned commits issue.~ And when that didn't work, I _reset_ the PR branch to `main` and re-applied my...

> So, given that the way they work makes no sense, and they're completely untested (0 coverage), it's likely nobody uses the `{get,set}_*_defaults` Graph methods at all. Well, it's possible...

@lkk7 > EDIT: Interestingly, there's also a result for the `get_node_defaults()`, and it's in NetworkX... Yikes! So it is, and they just blindly take the first element of the returned...

@aullidolunar > Any solution for this one? Here with meson 1.5.1 The solution is to just create empty `.po` files in the source dir, before running `meson setup` for the...

(You can quickly create all of the necessary files with this shell one-liner:) ```bash cat po/LINGUAS | while read _l; do touch po/${_l}.po; done ```

Looks like #9074 is where this got broken, the PR had good intentions but made the existence of `.po` files a new prerequisite for setting up the build system.

@eli-schwartz #13873 is the PR in question. (Which doesn't fix the issue, because I don't see how it can be fixed, but it does at least fix the documentation to...

If "real-world" examples are the goal, mightn't it be better to mine _actual real-world code_ that uses the Python API for examples? DNF comes to mind, as a blindingly-obvious source...