[Haskell-cafe] Re: Thinking about what's missing in our library coverage

[Yitzchak Gale]

I agree with most of Alexander's many thoughtful comments
about Don's list of potential additions to HP. But I
disagree about pandoc.

Alexander Dunlap wrote:
> No. Pandoc is too actively developed to go into the HP.

It depends on the nature of the development. If the
API is currently very unstable and is expected to
stabilize soon, then wait a little bit. Otherwise, that is
no excuse to exclude something worthwhile.
Choose a well-tested numbered version and include
it.

Later, if we want to upgrade, just follow the usual
deprecation-upgrade process.

Umm - we do have a well-defined deprecation-upgrade
process, don't we?

> It's also much more of an end-user application than
> a "standard library"

pandoc provides an extensive library, and also a
command-line app. Is it a policy that in such a case,
we require the command-line app to be split off into
a separate package before we can include it?

I'm not sure that's so important, but if so, it should
not be hard to do that for pandoc.

> its applications are not general
> enough to be included in the standard
> distribution.

Text with markup is used in some way for almost
every application. This library provides tools to
convert between a wide variety of markup
formats. Sounds pretty general to me.

Other "batteries included" platforms contain
various tools for processing markup that are
far less general than pandoc. This is a place
where Haskell can shine.

So yes, pandoc should definitely be included
in the platform. All that said, though, I will
certainly agree that it is not currently in the top 5.

Regards,
Yitz