HI Tomas,
This is good. Since we've had no complaints about your PS patch
I'll commit it tommorrow, if we can sort out some casting weirdnesses.
I first applied your patch on RedHat 6.2 which uses gcc2.92. Everything
went fine. On my RH 7.0 notebook, which uses gcc 2.96 I get errors
compiling
src/af/xap/unix/xap_UnixPSGraphics.cpp
because the array "buf" is declared
unsigned char buf[OUR_LINE_LIMIT*2]
and various strlen and strcpy functions want char * variables.
To get abi to compile in RH 7.0 I had to cast:
strlen( (char *) buf)
and
strcpy( (char *) pD, (const char *) glyph)
etc..
Now will that screw everything up for characters > 127? Is there a defined
standard of str* functions that will work on unsigned char * variables?
Cheers
Martin
On Sat, 10 Feb 2001, Tomas Frydrych wrote:
>
> This is how I envisage the future work on the BiDi functionality:
>
> Stage 1 -- geting it into the CVS, hopefully very soon.
>
> Stage 2 -- implementing the interface for non-Unix platforms; willing
> volunteers are sought to put in place around 6 checkboxes into
> already existing dialogues.
>
> Stage 3 -- modifying the importers/exporters, initially at least the
> Word imp/exp.
>
> Stage 4 -- a publicity campain. Being able to move this from
> experimental stage to working and realiable feature will depend on
> finding people who would be willing to use the BiDi enabled AW in
> normal conditions and report problems back, particularly people for
> whom BiDi is one of life's realities. Any suggestions where and how
> to reach such willing (and patient people) would be welcome. (They
> do not have to write code, although any hackers will, of course, be
> more than welcome.)
>
> Stage 5 -- making it a standard feature of AW. This should be a
> mere formality, if stage 4 gets well of the ground.
>
> In terms of a timetable, it is not realistic, I think, to expect to reach
> stage 5 by version 1.0, but we should be somewhere in stage 4 by
> then.
>
> Tomas
>
>
