Hi everybody,
I finally decided to get my hands dirty and figure this out. It looks
like the program (figinset.C) has a bug in it. It essentially keeps track
of the boolean variable gs_color which stores whether there is a local
color colormap (as opposed to a local gray colormap). The problem is that
when it decides how to render the EPS images, it decides based on
gs_color. Thus, if you don't need a local colormap (ie. you have
StaticColor or TrueColor), it incorrectly concludes that it should render
the EPS in gray.
The solution was to define a boolean variable, color_visual, which
stores whether the visual is color(based on the visual class obtained in
InitFigures()), and to use this to determine if it should be rendered in
gray or color.
The patch (for src/figinset.C in lyx-1.0.3) follows this email. I would
appreciate it if somebody could test this out on visual classes besides
TrueColor(I know this one works). For some reason, lyx dumps core on
anything but TrueColor for me, even with the unpatched RPM and when I'm
starting lyx without an argument/document). (Has anybody else had this
problem?) I'd like to make sure this works properly before merging it
into the main tree.
Dan
--------------------------------------------------------
111a112
> static bool color_visual; // is the visual color?
389a391,396
> if ( (vi->c_class == StaticColor) ||
> (vi->c_class == PseudoColor) ||
> (vi->c_class == TrueColor) ||
> (vi->c_class == DirectColor) ) color_visual = true;
> else
> color_visual = false;
621c628
< if (gs_color && !gs_gray)
---
> if (color_visual)
