[darcs-devel] status of libdarcs?

[Alexander Staubo]

On 3/29/08, David Roundy <droundy at darcs.net> wrote:
> On Sat, Mar 29, 2008 at 09:56:52PM +0100, Alexander Staubo wrote:
>  > Let me rephrase the above proposition, then: Move all of the Darcs
>  > engine code into libdarcs, rewrite the current Darcs to use it.
>
> When you're interested in doing this, let me know.

I am interested in helping to design, document and test the C API, and
develop Ruby and Objective-C bindings for such a library.

I am not a Haskell developer, however, and the time required to become
one, and become acquainted enough with Darcs' largely undocumented
source code to contribute, seems an insurmountable obstacle. The
situation would be different if Darcs were written in a mainstream,
non-functional language. That's just the way my brain is wired.

I am writing a version-control application for Darcs [1] in my spare
time, and my development time and productive energy is best spent on
writing the code -- a daunting and time-consuming enough project as it
is -- and in delineating what kinds of bindings and support I need for
this tool.

If I were to invest all my time in learning Haskell, I would never get
done. As it stands, the reason I am *not* done is because Darcs is
very hard to automate -- or at least, very hard to automate the way I
want in order to create an intuitive user interface -- and for some
things, like the recording problem, I simply don't have satisfactory
solutions.

[1] It is looking increasingly like I have to change focus in the
direction of Mercurial and/or Git.

Alexander.