03:21 <Rennex> i'm trying to get nanoc to copy a certain subdirectory of content/ to the output without any modifications, but there are text files in there encoded in iso-8859-1, and nanoc blows up with "RuntimeError: Could not read content/public/blah.htm because the file is not valid UTF-8."

03:23 <Rennex> so... yes, i know it's not valid utf-8. i don't want to touch it, and the files i want processed are utf-8. When it's simply copying it untouched, it shouldn't care about the contents

03:23 <Rennex> so is there a way to fix that? all i'm finding on the web is just "convert your input file" or "specify the right encoding"

03:24 <Rennex> my rule for the dir is: compile '/public/**/*' do; write item.identifier.to_s.sub(%r{^/public}, ""); end

03:36 <Rennex> ok, i removed the default text_extensions from the config and left only haml, txt, and css. now it ignores those .htm[l] files :)

05:20 <Deledrius> :)

05:33 <Rennex> is there a way to force re-generating given pages? nanoc sees that my index.haml depends on a .txt file (my code parses it and removes past events from it), but updating its timestamp with "touch" doesn't force re-generation

05:36 <Rennex> hm, or maybe it does. either it skips compiling it because the .txt content hasn't changed, or it doesn't list it as updated because the output didn't change. need to examine more...

05:41 <Rennex> ok, touch makes it recompile, and if the output actually changes, it does show that it updates it :)

07:59 <travis-ci> zoosky/nanoc-git/master d5ddd7e zoosky: The build has errored.

08:45 <travis-ci> zoosky/nanoc-git/master aa22d85 zoosky: The build passed.