IRC channel logs


back to list of logs

<pkill9>is there a repository of themes for haunt available?
<adhoc>pkill9: lists the no longer existing;
<pkill9>i'm on the homepage but there is nothing
<pkill9>i guess there isn't
<pkill9>i wish it was easy to import themes from other static site generator
<pkill9>hmm actually it might not be too difficult
<adhoc>is there not a default theme you can modify ?
<pkill9>well I want some fancy theme
<adhoc>find one, adapt the .css ?
<adhoc>trick seems to be to avoid .js as much as possible and do all the clever stuff directly in .css
<pkill9>yea that's what I'm gonna do, also you need to port the layout stuff
<pkill9>but porting it seems easier than i thought
<adhoc>well, yes, hopefullly there are enough of the required html5 tags in the template output ?
<adhoc>as long as the div tags have the right classes and id's you can do quite a lot with very little.
<pkill9>looking at this it seems fairly straightforward
<pkill9>infact it's mostly just the layout you need to port I think, as that will have all the relevant CSS tags
<pkill9>although jekyll has plugins that do things like pagineate, so i dunno
<pkill9>what's wrong with using .js stuff?
<adhoc>I'd like to figure out how to do something like the lowtech magazine solar site generator
<adhoc>damn, looks like it is down
<adhoc>ok, so if you are aiming to generate a static back end web site, you are not using .js to generate dynamic things ?
<pkill9>nah, i'd rather not use js anyways
<adhoc>is there a point to use .js for loading content and styling with the extra cpu overhead
<pkill9>everything can be done in css really i think
<adhoc>when .css does a fantastic job of layout and styling?
<pkill9>there was one really nice static site template example i saw
<pkill9>it was very nice
<adhoc>ther eis a place for .js, absolutely
<pkill9>looked like a magazine style
<pkill9>all in css i think
<adhoc>I took inspiration from many wordpress themes a while back, lots of good ideas there
<adhoc>especially for mobile first kinda layout =)
***jsoo_ is now known as jsoo
***janneke_ is now known as janneke
<tohoyn>sneek: botsnack
<pinoaffe>hi folks, I want to parse some date-time strings, is there some sort of module for this?
<pinoaffe>oh hey, strptime does the trick
<daviid>pinoaffe: srfi-19 as well, fwiw
<wingo>i landed a fun patch last night
<wingo>everything gets compiled through read-syntax, so symbols have source location info
***apteryx_1 is now known as apteryx
<lloda>i'm tempted
*spk121 is annoyed at all the extra gotchas in Gtk4's introspection
<leoprikler>It's more than just the new fundamental type?
<spk121>well for one, the new fundamental types need to hook up to GValue as well
<wingo>spk121: hey question, did you have a mingw dlopen branch somewhere?
<spk121>wingo: I've got some rough stuff, not pushed anywhere. I should get back to that
<spk121>there is a mingw branch with other patches that need to be considered first
<spk121>wingo: if you're thinking about releasing soon-ish, I think the mini-gmp stuff could be made ready. The MinGW stuff is no fun, so I keep getting distracted from it, but, with an actual deadline, maybe I could motivate
<wingo>well up to now there are three interesting changes that need testing -- the ltdl replacement, the gnulib update, and now the read-syntax change
<wingo>and we'll only get that testing if we release, so prolly should do that within the next few weeks
<wingo>i wish we had a web site to review patches
<spk121>Yeah. I'll try to get to MinGW (bleh). I think Civodul is still interested in mini-gmp, but, it is up to him.You should consider pulling in the vector-map and vector-last patches sitting in guile-devel@ They seem solid.
<tohoyn>spk121: is it ok to participate into Potluck with several projects?
*wingo should be a more responsible maintainer
<wingo>ah well, we do what we can
<wingo>will get to guile-devel before the release
<dsmith-work>Happy Friday, Guilers!!
<spk121>tohoyn: well since there is no judging and no real rules, I guess so
<spk121>tohoyn: but in future, maybe I should have thought it out better, because I do want to make sure I try out everything, and I may have bitten of more than I can chew, hahahaha
<tohoyn>spk121: I have made a patch to fix the cumulative time measurement in statprof. is this two tiny to be a Potluck project?
<spk121>tohoyn: the end product of all this should be a blog post where I talk about trying everything, and (if I can pull it off) some Guile stickers or postcards to mail out to participants. So I could imagine blogging about it, so I guess so.
<tohoyn>spk121: ok
<tohoyn>spk121: BTW I have posted the patch to guile-devel. What do you think about it?
<spk121>I remember seeing and a thinking it looked okay at first glance, but, honestly I never evaluated it
<spk121>If you are 100% positive that it correct and fixes a real problem, and if you can write a test case, you should push it into the tree
<spk121>(says me, a non-maintainer)
***apteryx is now known as Guest82559
***apteryx_1 is now known as apteryx
<spk121>tohoyn: with regards to statprof... If you could write a script that demonstrates statprof, it would be easier to blog about
<tohoyn>spk121: done and sent to guile-user
<rlb>wingo: fwiw, debian's in soft-freeze so bar is higher and will be getting higher still, soon, wrt including any new version in bullseye (next stable). I was also wondering about mini-gmp. Is that actually going to break abi compat, and if so, does that require a 3.2?
<wingo>rlb: i haven't looked into it. right now there are problems linking guile and non-guile gmp users because of scm_gc_malloc -- so things are already broken in a particular way
<spk121>rlb: minigmp only affects the dependencies on which Guile relies, but, does not change the ABI that Guile presents to the world
<wingo>spk121: though, there is still scm_to_mpz
<rlb>Hmm, I thought civodul had said that it broke the abi (maybe indirectly?).
<wingo>which might change to need freeing or not
<rlb>It'll be unfortunate if we end up stuck with the broken 3.0 throughout bullseye, but might or might not be much we can do about it -- i.e. not sure I could get a transition like that into bullseye at this point even if we were ready (if one ends up being needed)...
<rlb>I can ping civodul later.
<spk121>rlb: at the moment, if we did incorporate it, there would be no need to enable that configure option in the debian build, unless debian is having the same problem w.r.t gnutls that mini-gmp is trying to fix
<rlb>It is.
<rlb>That's how it was discovered.
<rlb>I think, i.e. debian buildds wrt gnutls
<rlb>In any case, if we can figure out what the plan is, I can see what's plausible and/or talk to the release managers (depending).
<rlb>(See what's feasible.)
<civodul>rekado_: thumbs up for the Guile-Studio and Picture Language web sites!
<pkill9>i've started porting a jekyll theme to haunt, and now I want to create an importer for jekyll themes :)
<pkill9>it's easier than I thought, the only real thing that needs importing is the html layouts
<pkill9>which also have some jinja-style conditionals
<rlb>civodul: was wondering earlier if we knew for sure whether or not the mini-gmp changes were in fact going to break the abi, and if so, what the plan might be, i.e. are we going to have to have a 3.2, or...? Also wondering in the context of the debian release We'll be increasingly limited in what we might be able to get in there for the next 2 years, if it matters...
<mdevos>does someone know of a Common Lisp -> Scheme compatibility layer?
<mdevos> I'm porting some Common Lisp code to Scheme
***rekado_ is now known as rekado
<rekado>mdevos: it’s really far out there, but yale-haskell is written in Common Lisp and the Haskell implementation sits on top of a compatibility layer that looks close to Scheme.
<mdevos>I just asked #lisp, and they don't know a compatibility layer either
<mdevos>Also, I discovered Common Lisp allows docstrings in more places. E.g., variables and condition types can have docstrings. Perhaps something to consider
<rekado>this is about the CL Hurd bindings, right?
<mdevos>rekado: yes
<rekado>this had been on my “do this eventually” list for a very long time…
<mdevos>rekado: when there's something to see, I'll post a link to a repo
<mdevos>(and announce to the hurd mailing list)
<rekado>what always held me back was the difficulty of setting up a Hurd development environment, given the confusing state of the upstream repositories
<rekado>I’d be happy to check it out!
<mdevos>rekado: I actually have never developed anything on the Hurd, and never installed any development tools in the childhurd!
<mdevos>rekado: the initial porting is happening on Linux. Let's first make the code compilable by guile ...
<mdevos>A while ago, I was surprised that syntax splicing (#,@) works with lists of syntaxes, and not only with syntaxes that have a list inside. Can I rely on this?
<mdevos>(Actually I prefer to use only the latter, but sometimes the former is convenient)
<wingo>mdevos: you can rely on it. not documented, but if we changed, many things would break i think
<mdevos>wingo: ok. I probably won't rely on it myself, though.
<civodul>rlb: re mini-gmp, i'm not sure it'd be ABI-compatible, we need to check
<sneek>civodul, you have 1 message!
<sneek>civodul, mdevos says: substituter patch ->
<mdevos>civodul: /me is going afk, unless there are questions you would like to ask now?
<wingo>ah a civodul sighting
<wingo>trains passing in the night, methinks :)
<civodul>wingo: heheh :-)
<lloda>tried master and everything looks fine +1
<dsmith-work>Make check on master works for me. (Debina stable)
<pkill9>this is so freaking annoying, there is a'i' appearing int he top-left of this website generated by haunt