Hi,

Reinier Lamers <[email protected]> writes:
> Op dinsdag 23 maart 2010 03:54 schreef je:
>> So I don't think this leaves much in favour of ignore-this, does it?
>> (Please bring up anything I might have omitted.)
>
> Just plain correctness. If someone once recorded in latin1 a patch
> with name "fix encoding handling so we never see garbage like 'é'",
> he will see "fix encoding handling so we never see garbage like 'é'"
> if he views the patch with newer darcs.
>
> Of course, we can accept this as a consequence of our own bug that we
> did not record the metadata encoding with the patch.
But that's not true -- this is a problem that is not solved by
ignore-this. You still have to guess an encoding to use for patches that
don't have the ignore-this utf8 tag.  There are always to be cases where
this fails, no matter how you choose the encoding, so it doesn't make
sense to talk about correctness. We can only think of more or less
successful heuristics. (In your case, if the user has an utf8 locale,
they still get é in that patch name.)

Moreover, I claim that utf8 falling back to current locale on failures
is usually no worse than going for current locale right away, and
sometimes is even better.

Since many distributions are defaulting to utf8 for a while now, we can
expect most repositories to come with utf8 patches.  Overall, I think
that the ignore-this tagging only benefits people with non-utf8
repositories on matching non-utf8 locales (*and* only in the case that
their patches happily decode as utf8 coincidentally), and such people
will only become rarer.

Yours,
   Petr.
_______________________________________________
darcs-users mailing list
[email protected]
http://lists.osuosl.org/mailman/listinfo/darcs-users

Reply via email to