tusharheroFeb 22

Pretty tired and frustrated with things not working in Emacs (as well as I would want them to.)

fj.el is great but keeps blocking and sometimes just fails.

Emacs-Guix is great, when it works. (https://codeberg.org/guix/emacs-guix/issues/15) It has gone to the point I find it easier to just M-! guix command instead of using the interface. I wonder if others are also doing the same and hence tolerating it?

On the other hand, I have started learning more about gnus, seems to do a lot of things.

#guix #emacs #gnus

Guix-Geiser Repl doesn't start

so I was testing #9 and was met with this error while starting the guix geiser repl and it happens will all the commands ``` guix-geiser-eval: Error in evaluating guile expression: ice-9/boot-9.scm:1685:16: In procedure raise-exception: Throw to key `match-error' with args `("match" "no mat...

Codeberg.org
Show threadmousebotFeb 23

@tusharhero my gut feeling is that the performance issues with fj.el stem from the API itself. it's not that i think i'm a great programmer, fj.el is haphazard and full of problems, but it uses code v similar to lem.el, which is an order of magnitude faster on loading large threads or search requests. :/

but still, feel free to @ me with specific gripes.

lem.el has a design where a gnus frontend could be built atop its API layer. using the jag.el gen code for fj.el would mean achieving the same design. but still, you need someone who can write gnus frontends, which is uh... not me! :)

Edit: well, it's not just the API, it's also fj.el, making threads loading slow. but when i looked into it, i was astonished by how much faster lem.el is, which uses similarly amateur code, and i was unable to improve it any more on the fj.el side.

Show threadtusharheroFeb 23

@mousebot I have been reading various nn* backends. I found a nngitlab too (though it is readonly and non-free.) I think it would be possible to do it for forgejo.

Reading the old threads on [email protected] (gnus's dev mailing list), it seems the problem was that gitlab's api, it was making it all very slow.

Show threadtusharheroFeb 23
@mousebot as for the specific gripe, sometimes PRs don't load at all, and say "I am error!".
Show threadmousebotFeb 23
@tusharhero I am error! means the request failed. that's url.el or the API.. :/ I really only see that when codeberg.org is down....
Show threadtusharheroFeb 23
@mousebot it seems to happen with specific prs, let me see if i can reproduce it everytime.
Show threadmousebotFeb 23
@tusharhero ah ok, in that case do let me know if you can catch a particular PR that always breaks, sure.
Show threadtusharheroFeb 23
@mousebot I guess it doesn't really say "I am error!" in this case but I can't read the PR comments at all. https://codeberg.org/MonadicSheep/emacs-reader/pulls/126.
WIP: Implement features for text selection, search, highlight

This PR primary resolves #29 by adding basic features that are needed for text selection, search and manipulation. It will _try_ to implement some of the primitives and basic support for #37, but it is not the highest priority.

Codeberg.org
Show threadmousebotFeb 23
@tusharhero & please report such breaks, it's tne only way to fix stuff. i'm just a guy in a bedroom.
Show threadtusharheroFeb 23
@mousebot I understand, sometimes I just decide to endure it...
Show threadtusharheroFeb 23

@mousebot error in process filter: I am Error: 40cea318282c25586c7bd2f4df0d07a5a3fd45c4 Endpoint: repos/MonadicSheep/emacs-reader/git/commits/40cea318282c25586c7bd2f4df0d07a5a3fd45c4

pretty sure i did reload the package, let me try in a new emacs instance.

Show threadtusharheroFeb 23
@mousebot yeah it is reproducible.
Show threadmousebotFeb 23
@tusharhero using the new fix i can view the PR (your link above) with no issues in fj.el. is the error you just gave from trying to load it or something else?
Show threadtusharheroFeb 23
@mousebot from trying to load it.
Show threadmousebotFeb 23
@tusharhero the error you posted refers to the first "unreachable" commit, display of which was recently fixed and which works for me. fixed in fj.el 0.29/0.30....
Show threadtusharheroFeb 23
@mousebot maybe I am loading the wrong fj.el then
Show threadmousebotFeb 23
@tusharhero i set up a minimal emacs script that loads fj.el (handy to have), also working there...
Show threadtusharhero
@mousebot I evaluated fj.el and fj-transient.el from a file manually. The error still persists.
Feb 23 at 1:18pmWeb
Show threadmousebotFeb 23
@tusharhero ah, i guess you're only tracking tagged releases of fedi.el? i pushed a new tag release for it, 0.3. (the fj.el fix needs a fix in fedi.el too.) sorry about that
Show threadtusharheroFeb 23
@mousebot AH that must be it!
Show threadmousebotFeb 23
@tusharhero sorry for the runaround, i struggle to keep the releases/tags stuff in my head as i just run of my local dev branch...