[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Freetype] Re: freetype behaviour below zero
From: |
Alexei Podtelezhnikov |
Subject: |
[Freetype] Re: freetype behaviour below zero |
Date: |
Tue, 2 Apr 2002 13:59:41 -0800 (PST) |
---------- Forwarded message ----------
Date: Tue, 2 Apr 2002 23:21:23 +0200
From: n++k <address@hidden>
To: Alexei Podtelezhnikov <address@hidden>
Subject: Re: freetype behaviour below zero
[Alexei Podtelezhnikov <address@hidden>]
| This was absolutely beautiful bug, if any.
| Your report, however, is missing important information:
|
| 1) version of freetype
2.0.9 and 2.0.4 tried both
| 2) compiler used to compile freetype and your program
gcc 2.95.3 for both
| 3) OS
linux [2.4.x]
| Mayby, it's just a stupid optimizations in your program like -O3.
program compiled with -g, freetype with -g -O2
it is perhaps a bug i introduced while converting my floats to FT_Vectors
suitable for the FT_Outline object: (see attached source code, outline.c)
static
outline_t* outline_conic_to(outline_t* self, vector2d_t* control, vector2d_t* t$
{
outline_add_point(self);
self->super.points[self->super.n_points-1].x = control->x * 64;
self->super.points[self->super.n_points-1].y = control->y * 64;
self->super.tags[self->super.n_points-1] = 0;
...
perhaps is my '* 64' incorrect ? (but then while would it fail with
negatives, and in such a weird way)
| Who knows? whould you attach the entire program to demonstrate the bug, not
| just a picture.
|
| Thanks,
| Alexei
provided tar.gz contains source and compiled 'tester'
program for linux|xwindow
-
n
++k
- [Freetype] Re: freetype behaviour below zero,
Alexei Podtelezhnikov <=