03:46 <smkelly> ddfreyne: cripes, you have nanoc generating PHP?? https://github.com/ddfreyne/stoneship-site/blob/master/content/404.erb

03:48 <bobthecow> smkelly: so do i.

03:49 <smkelly> why?

03:49 <bobthecow> it's hard to beat php for "mostly static"

03:49 <bobthecow> http://justinhileman.info/random/

03:49 <bobthecow> for example.

03:51 <smkelly> ok I get that use case

03:53 <bobthecow> so that php file is quite literally a giant array of urls, and i randomly pick one of 'em and redirect.

03:53 <bobthecow> wouldn't make sense to spin up a ruby or python or aything else app for that :)

03:57 <bobthecow> if i had "search" functionality on my site, i probably would implement that in PHP as well.

07:55 <ddfreyne> smkelly: Yeah, and I use PHP for mystonline.com's language redirect too

07:55 <ddfreyne> In fact... http://nanoc.ws/docs/guides/creating-multilingual-sites/#redirects :)

09:08 <darix> lighty mod magnet

09:09 <ddfreyne> darix: I used that in the past, but it is not a very portable solution

09:28 <ddfreyne> Idea: do not rely on implicit defaults in the nanoc config file. Everything always needs to be in nanoc.yaml. Feedback?

09:29 <ddfreyne> (it would not be as big in nanoc 4.0 as it is now)

19:03 <musicmatze> Yeah, I'm back blogging! I've setup a nanoc based minimalistic blogging site... until my static site compiler is ready...

19:08 <guardian> musicmatze: you know I'm serious when it's about coding in C but… what's the point of a static site generator in C? the key is letting the user customize/extend it. I don't imagine myself writing the kind of helpers I wrote for nanoc in C :/

19:12 <musicmatze> guardian: One of my goals is to write an interface to python or ruby or an other high-level language

19:13 <musicmatze> I think I can be faster for large sites. And when I write large sites, I mean really large sites: +10k items

19:13 <guardian> fair enough

19:14 <musicmatze> I'm currently only working on the kernel, which does nothing by itself. But I'm nearly at the state where I will write modules, which will then do stuff.

19:14 <guardian> yet, the helpers/plugins would be the bottleneck and the core DAG engine won't help when the slow helpers will run for thousands item :) anyways, I don't want to discourage you from doing it!

19:14 <musicmatze> 'does nothing' is not quite right... it provides core functionality

19:15 <musicmatze> I will write the critical stuff in C or an other fast language (I really would love to learn haskell, which is fast, too, I guess)

19:16 <musicmatze> I think the most expensive part will be the templating and the dependency thing...

19:18 <musicmatze> But I'm not at a state where I can tell you which part will be the most expensive. The kernel has to be finished, then I have to write at least four modules to be able to compile things... and then I can talk about templating and dependencies, as a templating mechanism is not included in the four modules...

19:20 <musicmatze> And I'm planning that the kernel can be released in v0.1.0 until end of this year and I will be able to build a site (with templating) at the end of next year. I know, that's a lot of time, but things are more complicated in C :-P

19:25 <musicmatze> What really annoying is: Writing the stuff alone. But I think thats common on open source projects...

19:29 <guardian> indeed