Re: IPA coding and UTF8
Aidan Kehoe wrote:
Ar an deichiú lá de mí na Nollaig, scríobh Ben Wing:
> [...] BTW part of unicode-internal is a redoing of the internal mule
> coding so that any of the 96 possible charsets can be either dim-1 or
> dim-2. (and a slightly more advanced version doubles the number of
> possible charsets to 96x2 = 192.) i also added unicode charsets that
> cover all characters unrepresentable in some other way. we might consider
> backporting this to 21.4, so that we can at least open, if not display,
> all possible unicode chars.
21.4’s Unicode support is entirely through Mule-UCS, the buggy CCL
implementation that increases the process’s resident size by 13 MB or so.
Backporting your Unicode implementation for 21.5, which I agree would be
nicer to our users, would be a big change to the stable branch, and I’m not
sure Vin would be okay with it.
no, didn't mean backporting the whole shebang; just adding coding
systems that can be used to handle unknown unicode characters rather
than having them get corrupted.