I think the challenge with base is that it's a Tcl codebase, which is a language most developers these days are less familiar with. Folks write Portfiles, yes, but they don't have to write proper Tcl too much as much of Portfile development is very declarative.
I also understand the desire not to get too bought into Github, but like it or not, it's a massive default in the world of modern software development. I have a strong notion that if we started using Github Issues for tickets, after a while we'd see an uptick in tickets being opened by the community (granted that it would take a while for the understanding to propagate that Github Issues is available). That's because Github Issues is less friction for folks already using Github for all the other software they're working with, as compared to having to deal with an entirely separate ticketing system. Don't get me wrong: Trac is arguably low-friction, but a lot of folks would just not bother the minute they see it's something they have to go outside of Github to do. On Wed, Nov 8, 2023 at 7:33 PM Eric Gallager <eg...@gwmail.gwu.edu> wrote: > On Wed, Nov 8, 2023 at 3:47 PM Clemens Lang <c...@macports.org> wrote: > > > > Hi, > > > > On Sun, Nov 05, 2023 at 01:58:00PM -0500, Perry E. Metzger wrote: > > > On 11/5/23 13:37, Daniel J. Luke wrote: > > > > To clarify - this was in the context of commits to base/ > > > > > > > > (That may or may not change your perception - but like I said, I > > > > didn’t measure and this is totally measurable). > > > > > > I don't think the base tools are getting a lot of attention because > > > they mostly just work? > > > > Maybe that's the perception, but I'd argue it really isn't true. There > > are lots of improvements that could be had in base, if somebody had the > > time to implement them. > > > > Just to name a few ideas: > > > > - Hooks to run certain scriptlets (defined somewhere in the ports tree) > > when files in specific directories are installed > > - Switching selfupdate from rsync to http, which would avoid leaving > > users behind where the network blocks rsync > > - Switching to more modern signatures compared to the RSA signature over > > the RMD160 hash we're still using today > > - Improvements to make builds more reproducible (e.g., > > SOURCE_DATE_EPOCH, clamping mtimes of generated files, ...) > > - Finishing the 'port migrate' command to automate the migration process > > - Better post-build testing to detect problems in ports, e.g., scanning > > all compiled binaries to ensure that the files they link against are > > in their declared transitive dependencies. > > > > And these are just the ones I have worked on, or have on my todo list. > > > > Yeah, I've got a bunch written down, too; this Trac query gets the > ones that I've filed tickets for on Trac over the years: > > https://trac.macports.org/query?status=accepted&status=assigned&status=new&status=reopened&component=base&reporter=egall%40macports.org&reporter=egall%40gwmail.gwu.edu&reporter=cooljeanius&order=priority > ...and these are the ones I'm on cc for: > > https://trac.macports.org/query?status=accepted&status=assigned&status=new&status=reopened&cc=~egall%40macports.org&cc=~egall%40gwmail.gwu.edu&cc=~cooljeanius&component=base&order=id > > > -- > > Clemens > > >
