Mail Archives: djgpp/1997/10/05/14:17:45
On Sun, 05 Oct 1997 15:00:52 GMT T DOT Harte_ AT _btinternet DOT com (Thomas Harte)
wrote:
> >I just learned that SciTech just released their MGL graphics library free
> >of charge with full source code. It supports OpenGL and I do believe it
> >can compile under djgpp. Can someone please tell me the advantages of the
> >MGL library over say Shawn's excellent Allegro library? Please note I am
> >referring to the graphics library portion and nothing else.
>
> The advantages should in theory be obvious. It can do hardware
> accelerated 3d, using OpenGL and it can support DirectX surfaces, but
> these are not real advantages since the thing doesn't work! I've spent a
> total of five hours so far attempting to get it just to compile an
> example, and it won't. Has anyone actually got this to work? Yes, I am
> using the correct everything for DJGPP before you ask.
>
> -Thomas
As wrote Berret Kendall they broke DJGPP's support for MGL somewhere in
a middle of developing and for now support it again. MGL 4.01 has some
errors in a keyboard and mouse interrupts but pure graphic examples must
work.
I download MGL 4.01 sources but ScitechSoft use strange distribution
scheme with 'base' and 'full source' install - first thing is needed for
installation of the Sources ;) so I'll give solution&impressions later.
Here is a part of problems.txt from their ftp-site:
(ftp://ftp.scitechsoft.com/devel/problems.txt)
-------------------------------------------------------------------------
SciTech MGL Graphics Library
Version 4.01
Known Problems and Things To Fix
October 4 1997
Internal rendering functions
----------------------------
1. Ellipse rendering is done using a new floating point DDA algorithm
to avoid the overflow problems the old integer algorithm had. The
ellipse code needs to be re-engineered with a new integer algorithm
that wont sufffer from overflow.
All the ellipse assembler functions are currently stubbed out and
the emulated floating point routine is used. Although this routine
is fast, the assembler integer routines would be faster if they were
fixed.
[snipped - problems with other compilers]
DJGPP 2.01 32-bit DOS support
-----------------------------
1. Somewhere along the line the DJGPP support has gotten broken, and
the code appears to crash miserably on keyboard interrupts (or
mouse interrupts, I am not sure). It appears to be in one of the
DPMI real mode callbacks, so the problem may be simple to fix.
---- END OF PROBLEMS.TXT ----
Alexander Bokovoy, <bokovoy AT bspu DOT ac DOT by>
---== The Soft Age coming soon ==---
- Raw text -