[9fans] recent changes on sources

[9fans] recent changes on sources

[9trouble]

1. <u.h> now includes a definition for uintptr, which we define
to be an integer of the same width as a pointer.  Pointers and
uintptrs should be printed using %p.  

I believe that the C standard requires that uintptr_t is at least
as wide, but may be wider, than a pointer.  We've chosen to be
a little more restrictive in the definition in order to let code be
a little more lax - it can assume that pointers and uintptrs are
pushed on the stack the same, va_arg the same, and so on.

2. <libc.h> now defines Runeerror as 0xFFFD, the new official
Unicode error value.  We used to use 0x80, but the rest of the
world standardized on 0xFFFD.  

3. All fonts that have a glyph for 0x80 (a question mark in a box)
now use the same glyph for 0xFFFD.  For those of you who read
raw UTF8 bytes, get used to seeing "ef bf bd" as Runerror now
instead "c2 80".

4. We have a larger variety of sizes of Japanese bitmap subfonts,
prepared and contributed by the TIP9UG Japanese fonts project.
The subfonts have been added to the standard Unicode fonts as
appropriate.  TIP9UG started with BDFs from the efont project.
Thanks to both groups.

- 9trouble

Re: [9fans] recent changes on sources

[boyd, rounin]

> 2. <libc.h> now defines Runeerror as 0xFFFD, the new official
> Unicode error value.  We used to use 0x80, but the rest of the
> world standardized on 0xFFFD.  

i know i've posted this before, but it's just a simple filter
to debug rune encoding woe:

    http://www.insultant.net/code/plan9/vr.c

--
MGRS 31U DQ 52572 12604