summaryrefslogtreecommitdiff
path: root/system/snapscreenshot/NOTES.txt
blob: 3acfd2a9b994787f5361a4dd887771e83fe2a790 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Notes:

snapscreenshot is *fast*, when compared to fbdump and fbcat. However,
the reason it's so fast is that it doesn't read the graphic content of
the framebuffer. Instead, the text content of the screen is read (from
the /dev/vcs* devices) and rendered in the current console font... which
is the current font of the console you executed snapscreenshot from,
*not* the font in the console you're taking a snapshot of!

Also, custom fonts just don't work well with snapscreenshot. Depending
on the font, you may get an error (GIO_FONTX), a segmentation fault,
or else the resulting screenshot is garbled.

Because /dev/vcs* is used, snapscreenshot does NOT work for taking
screenshots of software that actually uses the framebuffer for graphics,
e.g. fbpdf, feh, "mplayer -vo fbdev", "links -g -driver fb". For these,
use fbcat or fbdump instead.

By default, the text on all the consoles ends up in the output image. To
take a shot of only one console, use a command like:

# snapscreenshot --firstwin 1 -c1 -x1 > shot.tga

...or use this:

# snapscreenshot.cur > shot.tgz

The snapscreenshot.cur script was written by the SlackBuild author,
and is not part of snapscreenshot itself.