
Yes I was thinking that too. Bringing in Pandoc brings in too many
other dependencies.
Perhaps it could be left in an uncompiled example or readme?
On Mon, Oct 18, 2010 at 6:20 PM, Michael Snoyman
Just for the record, while I like the markdown-based interface, I can't include it in the mime-mail package: I'm not going to make Pandoc a dependency of mime-mail, both for weight and license reasons. It would be awesome to see a markdown-mail or similar package, however.
Michael
On Mon, Oct 18, 2010 at 4:33 PM, Robert Wills
wrote: I've updated the gist to now include automatic conversion through pandoc of markdown to html.
On Mon, Oct 18, 2010 at 2:32 PM, Robert Wills
wrote: On Mon, Oct 18, 2010 at 1:57 PM, Michael Snoyman
wrote: This is *exactly* the kind of high-level interface I was hoping someone would provide ;). Anyone have any objections to this being the de-facto "simple" interface for mime-mail?
Cool. It might be good to have something a little more composable so eg you could add ccs without having to drop down to the lower-level.
I can help out there. A lazy ByteString is nothing more than a lazy list of strict ByteStrings. This can be more efficient since we don't need a gigantic single block of memory, and can also allow us to generate data lazily, one chunk at a time. Converting a lazy ByteString to a strict one can be done with:
Great that worked. I'll have a look at HaskellNet and see how easy it'll be to change.
I've put the file I was playing with up here: http://gist.github.com/632222
-Rob