[darcs-users] darcs patch: Resolve issue1292: re-encode line input from the Haskeline backend.

[Petr Rockai]

Hi!

Judah Jacobson <judah.jacobson at gmail.com> writes:
> Wed Jan 21 09:24:22 PST 2009  Judah Jacobson <judah.jacobson at gmail.com>
>   * Resolve issue1292: re-encode line input from the Haskeline backend.
>   This patch bumps dependencies to haskeline-0.6.* (which provides the required
>   functionality) and terminfo-0.3.* (which is required by that version of
>   Haskeline).  Haskeline is also enabled by default now that non-ASCII
> line input
>   works correctly.
>
>     M ./darcs.cabal -7 +5
>     M ./src/Darcs/Utils.hs -2 +7

Kowey: The patch looks OK to me -- I have done a cursory review.

Judah: Thanks for the patch. I have one question though, and that is, how hard
it is to get working haskeline on Win32? If it amounts to "cabal install
haskeline" without needing cygwin and the like, I'd like to propose dropping
the non-haskeline backend for 2.3.

It would also constitute a first step to stop putting the literal bytes we get
from terminal into persistent metadata. Even if not a complete solution, it
would possibly make 99% of newly recorded patches utf8-clean. Step like this
would of course need to be mentioned in release notes, although I highly doubt
there are any projects that keep their metadata in pure latin1 or anything. On
the contrary, I have seen enough repositories with mixed-encoding content,
which is a nuisance.

Yours,
   Petr.

-- 
Peter Rockai | me()mornfall!net | prockai()redhat!com
 http://blog.mornfall.net | http://web.mornfall.net

"In My Egotistical Opinion, most people's C programs should be
 indented six feet downward and covered with dirt."
     -- Blair P. Houghton on the subject of C program indentation