Hi John (and others following this issue),I dare respond through our BTS rather than privately, as I see nothing in need of discretion in your mail and we should generally work transparently and in the open whenever possible[1].
I have commented below the quoted text... On Tue, Dec 15, 2009 at 10:39:22AM -0800, John MacFarlane wrote:
This is unfortunate. Pandoc uses template Haskell to splice the contents of some data files into source files, so they can be built into the executable. But apparently not all architectures have a version of GHC that is capable of handling template Haskell. This will keep pandoc 1.3 from entering testing, I believe, because 0.46 did compile on these architectures. (Correct me if I'm wrong about the criteria.) 0.46 used a different strategy -- it created DefaultHeaders.hs and other data files from templates before compiling, and didn't use template Haskell. My new plan is to wean pandoc off of template Haskell, which seems to be more trouble than it's worth. http://code.google.com/p/pandoc/issues/detail?id=186 John
I do not know these Haskell-specific details of coding style. But if the new way is "the proper way", then we do have the alternative of packaging only for those architectures supporting this better approach.
Do you perhaps have some references to discussions by others about template Haskell not working on some archs? Perhaps the Debian Haskell team have discussed similar issues with other Haskell apps/libs?
- Jonas [1] We won't hide problems: http://www.debian.org/social_contract ยง3 -- * Jonas Smedegaard - idealist & Internet-arkitekt * Tlf.: +45 40843136 Website: http://dr.jones.dk/ [x] quote me freely [ ] ask before reusing [ ] keep private
signature.asc
Description: Digital signature
