Hrvoje Niksic <hniksic(a)srce.hr> writes:
"Kirill M. Katsnelson" <kkm(a)kis.ru> writes:
> | True, but that variable doesn't exists for strings. For instance:
> | (md5 "some Japanese text")
> | This is what the CODING-SYSTEM argument and the guessing code is there
> | for.
> This is what undecided streams are for, but the comment in md5 about
> hit and miss is unfortunately rather correct...
OK, should I default the coding system to `undecided'? BTW, Jareth,
could you please enter this discussion? I still haven't heard back
from you regarding md5.c. I'd like to know that my changes to it are
correct before attempting to work more on that file.
Your changes work just fine on my test cases. I don't have a problem
with it at all. As for changing the coding system to undecided, what
does that buy us? I think we've got enough decision logic in there at
the moment to cover the bases. What we really need to do is rethink
Mule and coding system functionality completly.
> coding-system-for-write is absolutely no friend of yours,
> is constantly nil, and can be nothing else.
Well, that variable doesn't look very useful then.
At the moment, it is next to useless. None of the environments
properly set it. Again, serious rethinking is needed.
Jareth Hein | jareth(a)camelot-soft.com | ハイン ジェラス
Toolsmith & Program lead | http://www.camelot-soft.com
Camelot Software, Ltd. | |（株）キャメロット
"Contrary to popular belief, penguins are not the salvation of modern
technology. Neither do they throw parties for the urban proletariat."