Recently I made two #guile contributions:
- A rewrite of `(ice-9 peg string-peg)` that supports PEG properly (before it didn't) and also supports [^...] syntax, which is very present in modern PEG.
- Finished a port of the JIT compilation library for RISC-V I started 3 years ago and tested it successfully on Guile, in real hardware.
I don't feel any of those are ever going to be included in the project, and that's very discouraging.
Guile needs more people.
Maybe more #Guix people?
It's funny because the reason why the first was broken and needed to be fixed and the second was abandoned 3 years ago is the same that makes me feel this is never going to be merged.
There have been efforts to before improve the PEG module, but no answer was given to them.
3 years ago I passed all the lightening tests but I didn't finish the work, because I didn't know how. With a little of help, we could've done it. I got no help and I abandoned until I had the knowledge to finish.
In the meantime the work was almost done, but we couldn't enjoy a JIT-ed version of Guile in RISC-V for 3 years. Same with PEG.
It's a shame.
Maybe it's just me, but it would be easier for me to accept it if the work wasn't done. But often it is done, it's in the mailing list, waiting to be applied in the project. Which never happens.
@ekaitz_zarraga Ahem, I think I did answer to the PEG patches, didn’t I?
I haven’t checked the next round yet, and it’s already been way too long, and really the problem is that nobody else is looking at the backlog currently, but… we’ll get there.
Now, I agree that there’s a real issue to be solved in Guile.
@civodul Yes, but as an special case because you know me and follow me in here and you saw I did it, and I bothered you in IRC about it.
We cannot rely on you being nice to people. It's great, don't get me wrong. But in order to keep you being nice for long time, we cannot squeeze the niceness too hard out of you.
Am I explaining myself properly here?
@civodul I think a good improvement to Guile's issues could be to motivate more Guix developers to take part on Guile (I'll put the ML on fire again with some essay about why we should pay Guile more attention).
I think many people, and now I'm talking more in general, are afraid to take part in the development and the maintenance of a programming language, probably because the covers of the books that explain them have a very scary dragon.
But this we need to improve.
@ekaitz_zarraga @civodul I would contribute to guile but it seems like none of the 7 bugs this month have received a reply, not very welcoming :/
Furthermore, contributing to a new project involves many questions and asking them on e-mail or IRC feels like sending a message bottle down a dark pit, and this also goes for #guix.
I've seen projects be so much easier to get into: projects with associated discord servers are a huge success in this way. Providing a way for newcomers to actually chat with other users and maintainers in a friendly and casual manner.
In my experience, e-mail and IRC are extremely unwelcoming technologies to use.
@ekaitz_zarraga @civodul The simple fact that you have to keep your IRC client open to receive messages makes it the only messaging app where you don't receive the messages. And e-mail is the only one where messages might go anywhere if you don't re-read your headers a bajillion times.
https://xmpp.org/software/
- replies
- 1
- announces
- 0
- likes
- 0