Re: Unicode 7.0 released - some packages contain outdated embedded data copies
Adam Borowski <kilobyte <at> angband.pl> writes:
> On Wed, Jun 18, 2014 at 02:54:43PM +0200, Thorsten Glaser wrote:
> > > Unicode 7.0 was recently released. I discovered some source packages
> > > contain outdated copies of various Unicode data files. At minimum, the
> > I know that xterm’s wcwidth.c direly needs updating, and that mgk
> > doesn’t do that. I do have a drop-in replacement available (which
> > is what I use in mksh (16-bit only) and jupp (full 21-bit Unicode).
> > If the xterm maintainer is interested, or even upstream (Tom), I’d
> > be willing to provide patches.
> It would be better to use the copy provided by glibc. Of course, this one
> needs to be updated for Unicode 7.0 as well.
No, that’s an incompatible licence (LGPLv2.1+) and relies on internal
data structures and all that.
(Plus, some routines require different return values for certain
characters – for example, the one in jupp (and joe) wants the
string length of "<%X>" for control characters, as that’s the
display width in the editor.)
> Embedded copies of wcwidth.c are something that can be checked for
> automatically, but unlike your average embedded library it's often #ifdefed
> out when an autoconf check shows the existence of wcwidth() in the system
They are usually, yes, unless there is special need for separate
routines. That being said… most of the time, the copy in libc
is the outdated one… (in the general case, with upstream hat on).