Hi,
I sent emails to Thomas and Andreas about some patches I wrote for libGGI2D,
but have not received any reply from them. I'd like to get these changes put
into the CVS repository because currently the library does not work and I have
some code that uses it. (It used to work a few months ago but some changes got
made).
I am not sure who to email or ask about this because there is no traffic on any
of the mailing lists or the web site, and the bug tracker on source forge never
gets used either. Is the GGI project still alive at all or is it dead?
I have attached the email I sent previously to them ... if someone with CVS
access could make these changes (or fix it differently, or allow me to change
it or whatever) then that would be appreciated.
thanks,
Wayne
-------------------------------------------------------------------------------
Wayne Piekarski - PhD Student / Lecturer Phone: +61-8-8302-3669
Advanced Computing Research Centre Fax: +61-8-8302-3381
University of South Australia Mobile: 0407-395-889
Internet: [EMAIL PROTECTED]
Research & Development Manager [EMAIL PROTECTED]
SE Network Access Pty Ltd http://www.cs.unisa.edu.au/~ciswp
From: Wayne Piekarski <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>; <[EMAIL PROTECTED]>
Subject: libGGI2D Bug Fix Patch For CVS
Date: Friday, 26 January 2001 11:06 PM
Hi,
I sent you a report a few days ago with some bugs in the libGGI2D library. I've
been using it for quite some time and then some point in the last 6 months
things were changed and the library broke.
I have produced a patch for the library which fixes all the problems I've been
experiencing. The patch isn't the best code (I'm not an expert with the code)
however it does fix everything up to the point where it works again. It fixes
everything mentioned in my email forwarded below. The patch libggi2d-first.path
is a complete one for the whole ggi2d directory, and the two others hline-diff
and hline-complete are for another file I changed afterwards when I realised
I'd made a mistake the first time.
I had two directories, degas and degas-original, with the source trees, and ran
diff -u to produce the patch. I can produce it in any other format if you like.
If these could be integrated into the CVS (either as is or preferably with
other changes of your own) I would be grateful and I'm sure others would be as
well.
thanks for your time,
Wayne
----- Original Message -----
From: Wayne Piekarski <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Tuesday, January 23, 2001 3:36 PM
Subject: Problems Compiling Recent libGGI2D from CVS
> Hi,
>
> I have been using the GGI, GII, and GGI2D libraries for a year or so, and I
> have noticed a few bugs. Some have been present for a long time, others only
> have appeared recently.
>
> Lately there were some major changes made to the rest of the system, and
GGI2D
> no longer compiles. The error message during make is:
>
> Making all in ggi2d
> make[2]: Entering directory `/root/degas/lib/libggi2d/ggi2d'
> /bin/sh ../libtool --mode=compile
> -DHAVE_CONFIG_H -I. -I. -I.. -I../include -g -O2 -D_REENTRANT -D_THREAD_S
AF
> E -g -Wall -c main.c
>
gcc -DHAVE_CONFIG_H -I. -I. -I.. -I../include -g -O2 -D_REENTRANT -D_THREAD_SAF
> E -g -Wall -Wp,-MD,.deps/main.pp -c main.c -fPIC -DPIC -o main.lo
> main.c: In function `changed':
> main.c:73: too few arguments to function `ggiExtensionLoadDL'
> make[2]: *** [main.lo] Error 1
> make[2]: Leaving directory `/root/degas/lib/libggi2d/ggi2d'
> make[1]: *** [all-recursive] Error 1
> make[1]: Leaving directory `/root/degas/lib/libggi2d'
> make: *** [all-recursive-am] Error 2
>
> If you jump to the line of code with the error, the ggiExtensionLoadDL
function
> has been extended with new arguments and does not match the current version.
I
> tried to look at other extension code and fix it up myself, but had problems
> and when the library was called by the demos it would seg fault internally.
>
> Also, in the directory degas/lib/libggi2d/default, there are three targets,
> linear_16, linear_8, and stubs. By default, the makefile only compiles the
> stubs, and does not touch the rest of them. However, later on when you run
your
> program, it dies because it cannot find linear_8 or linear_16.
>
> Another bug which I discovered was in the code that draws filled objects in
> 16-bpp mode. When filling in the scan lines between the edges of the object,
it
> assumes you are on a 8-bit display, so only half the scan line is filled, and
> it is the wrong color. It uses a memset or something to set the bytes. To
make
> it work I hacked in a call which calls some function to fill in a scan line,
> but it was a real hack job and needed a proper fix. I don't know enough about
> how everything works to do this properly as doing 16-bit pixels can be a
pain.
>
> I have noticed that the GGI2D libraries have not been updated for a very long
> time, and thought I would report these problems before the CVS tree is put
out
> as a stable final release after the beta's finish getting tested.
>
> I really like the GGI libraries, and it would be really great if these subtle
> little problems could be removed because it is tedious to continuously patch
> each new version I download.
>
> Thanks for your effort!
>
> thanks,
> Wayne
libggi2d-hline-diff.patch
libggi2d-hline-complete.patch
libggi2d-first.patch
