On Fri, Mar 10, 2006 at 05:37:38PM +0900, Stephen J. Turnbull wrote:
> Ie, as long as the user supplies the right encodings.  Thus condemning
> the Chinese and Japanese to (at best) inadvertant binary treatment of
> many of their files, and (at worst) corruption of data.

Actually, the same logic should also work for at least Big5,
Shift-JIS and EUC. UTF-16 is more tricky, though I imagine Darcs
could check for a BOM if anyone felt the need.

What's the alternative? Make Darcs encoding-aware? I suspect that
would cause more problems than it solves. Insist on UTF-8? How's that
an improvement?

> As for backward compatibility, the time to break with backward
> compatibility is almost always "as early as possible", because all too
> soon you end up with "ten minutes ago or never".

I think think that depends on the reason. 'Because maintaining
compatibility is causing problems', or 'because we're crusading'. I
see few problems with the status quo (at least w.r.t. Darcs), and
nothing to gain by being more strict.

-- Jamie Webb

_______________________________________________
darcs-users mailing list
[email protected]
http://www.abridgegame.org/mailman/listinfo/darcs-users

Reply via email to