IRC channel logs

2015-07-27.log

back to list of logs

<stis>it's also good to exersice the interface to find bugs!
<stis>hmm I found his github software, https://github.com/webyrd
<stis>ACTION is makeing a minikanren module from the kanren module
<nalaginrut>morning guilers~
<nalaginrut>I realized my design of Artanis server core is totally different from Guile inner server, so it means I can't take advantage of (web server). Besides, I need the activity of server use Artanis local config, which is another reason why I have to write a new one...
<nalaginrut>like, accecpt-body-size, charset, and a way to call sendfile
<nalaginrut>beyond, Artanis supports various protocols based on websocket, but Guile inner server is obviously a http server design
<nalaginrut>fortunately, (web server) is a good hint for me anyway
***jjmarin_ is now known as jjmarin
<paron_remote>stis: https://github.com/webyrd/polyconf-2015
<paron_remote>is the repo
<paron_remote>it contains some minikanren subdirectories, including one with "evalo"
***mario-go` is now known as mario-goulart
<davexunit>I think haunt just needs an auto-rebuild feature for 'haunt serve' and it's ready for 0.1
<dsmith-work>Morning Greetings, Guilers
<dsmith-work>Wow. Over 90 nicks in here.
<wleslie>and I am Gus.
<davexunit>every time wingo writes a new blog post, I get more excited for Guile 2.2
<davexunit> http://wingolog.org/archives/2015/07/27/cps-soup
<davexunit>"...we are able to do the complete set of optimizations from the "Optimizing closures in O(0) time" paper, including closure sharing, which I think makes Guile unique besides Chez Scheme."
<dsmith-work>sneek: later tell wingo master v2.1.0-1920-g4841239 is failing control.test
<sneek>Will do.
<paroneayea>davexunit: great post (though I don't totally understand it ;))
<sneek>paroneayea, you have 1 message.
<sneek>paroneayea, mark_weaver says: your pro-GPL article was mentioned in yesterday's LWN: https://lwn.net/Articles/651878/
<paroneayea>I read it earlier but thanks sneek ;)
<paroneayea>(as paron_remote that is)
<paroneayea>mark_weaver: but if I didn't say thanks for the pointer earlier, thanks! :)
<davexunit>paroneayea: yeah I definitely don't understand all of that stuff, but the stuff I can somewhat understand is very interesting
<amz3>O(0) is that even legal ?
<davexunit>rekado-: about haunt: I think you should add a 'post-file-name' keyword argument to 'blog'
<davexunit>that defaults to 'post-slug'
<rekado->yes
<davexunit>well..
<davexunit>hmmm
<davexunit>the arg should be 'post-slug'
<davexunit>and default to 'post-slug'
<rekado->There are also other things that might need customisation, e.g. "Recent Posts" as a hard-coded string is not pretty.
<davexunit>yeah of course
<davexunit>this was quick and dirty
<davexunit>it should also be able to paginate, but I'm not there yet.
<davexunit>rekado-: so, you could write your own 'post-slug' procedure that returns the post file name sans the ".html"
<davexunit>or whatever the extension is
<rekado->right
<rekado->I'm playing with this right now
<davexunit>an #:index-title arg could replace the hard-coded "Recent Posts" string
<davexunit>and a #:posts-filter procedure could replace the hardcoded 'posts/reverse-chronological' procedure call.
<rekado->hah, I was going to suggest that
<davexunit>#:posts-filter arg, that is.
<davexunit>;)
<davexunit>I wasn't really expecting anyone to actually try to use haunt, so you've caught me off guard a bit. :)
<amz3>davexunit: are you going to include a markdown parser in haunt 0.1?
<davexunit>amz3: nah
<amz3>ACTION is working on an article article about little markdown parser based on guile-log
<davexunit>I wouldn't bundle a parser anyway
<davexunit>it would be an extenal lib
<davexunit>the extension point is there to implement new post readers, so it can be added at any time, with or without me.
<amz3>yes
<rekado->davexunit: I just noticed that it will be too long till I get Hakyll packaged --- and haunt looked just right to get me going again and to let me play more with Scheme.
<davexunit>I'd rather just solve the issues that rekado- has run into and release, which will hopefully draw more interest.
<davexunit>skribilo/markdown support would be great for a future release
<rekado->It's great. Very easy to understand (unlike Hakyll, which took me a lot longer) and flexible.
<amz3>davexunit: when do you expect to release?
<amz3>indeed it's easy to understand
<davexunit>rekado-: that's very good to hear. I spent a good chunk of time thinking about how to make it as simple and extensible as possible.
<davexunit>amz3: no set date, whenever I have time. I just want to fix the blockers for rekado- and add a super naive auto-rebuild mechanism to 'haunt serve'.
<rekado->davexunit: it's a little ugly that 'blog' and the theme both need to know about how to generate a URL for a given post.
<rekado->'blog' is told how to do it via the POST-SLUG arg and PREFIX, while the theme defines its own post-uri procedure.
<rekado->not sure if this is worth avoiding as the theme is to be customised anyway.
<davexunit>rekado-: yeah, I noticed that when I wrote the code but wasn't sure what to do, so I did nothing. :)
<davexunit>the theme is meant to be customized so I guess it's up to the user to make sure it does the right thing
<davexunit>I expected most people to want the generated slugs.
<davexunit>rekado-: perhaps slug generation for posts should be a setting on <site>
<davexunit>since the whole site will likely need to know how to correctly form URIs to the rendered posts
<davexunit>your new slug scheme will also be incompatible with the atom feed generator.
<davexunit>so a site-global 'post-slug' procedure is looking pretty attractive.
<rekado->what's not so nice about this is that "site" now has to be threaded through more procedures, such as post->atom-entry.
<davexunit>I don't think it's so bad
<rekado->this could be made nicer with a reader monad, maybe.
<davexunit>we could use a parameter, but I'd like to stay away from them.
<rekado->davexunit: sent you a patch via email. Maybe not pretty but it works for me.
<davexunit>rekado-: thanks I'll look when I have a chance
<rekado->no rush.
<davexunit>rekado-: this patch is a good start. I have a few things I want to tweak and I think i noticed a bug. I'll clean it up and apply it later. thanks again!
<amz3>I've put online a small explanation of how guile-log parser combinators works http://www.hyperdev.fr/notes/2015/07/27/guile-log-parser-combinators/
<stis>amz3: nice!