All Projects

No need for description

Initially fixes (believed to be) only required for older devices.
Now also seen on newer devices

See also  FS#261 

This is a meta-item depending on bugs/features required to complete before releasing nettype 1.4.4

This is a meta-item depending on bugs/features required to complete before releasing nettype 1.5.2.

This is a bugfix-only release.

Previous:  FS#353  (1.5.1),  FS#215  (1.5),  FS#252  (1.4),  FS#203  (1.3)
Next: FS#329 (1.6)

This is a meta-item depending on bugs/features required to complete before releasing nettype 1.5

Previous:  FS#203  (1.3),  FS#252  (1.4)
Next: FS#329 (1.6)

Version 1.12.3 breaks on FreeBSD due to the usage of GNU-specific extensions of expr, sed and an incompatible Bash RE.

As reported by Dirk Reiners (paraphrased)

When steps get too small they might round to 0 ⇒ infinite loop in compute_timecodes.

Currently the extended set always contains the standard captures. It might be nice to add some time-shift to them so that they’re different.

See also  FS#122 

Complete manpage.

Safe length measuring, when probing the video length, prints only the number of seconds and milliseconds, it would be better to display a more readable format.

As found in mtn.

Try to avoid capturing mostly-blank frames.

At least on some systems, the console/terminal needs a manual reset (or tset) to be usable after running vcs.

When repeated captures are avoided, captures end up being filtered more than once

Suggested by Adjel and Vel

Add a possibility to change icons, possibly customise them with themes

The function that vcs.conf in current dir accomplished is covered by profiles. If wanted it can still be manually loaded with -C :pwd or -C vcs.conf

The filmstrip mode has had a placeholder implementation since the start, it should look closer to an actual film strip.

E.g.: http://sites.google.com/site/elsamuko/gimp/sprocket

mplayer: FPS1

i.e.:
ID_VIDEO_CODEC=fffraps
ID_VIDEO_FORMAT=FPS1

ffmpeg: fraps

i.e.:
Stream #0.0: Video: fraps, yuvj420p, 1680x1050, 30 fps, 30 tbr, 30 tbn, 30 tbc

Internal: Stale code from prior versions, numsecs inside process() uses mplayer’s value instead of the best value to determine video length (and keep or discard timestamps).

As a side-effect, newer code being more strict about array vs string handling can fail easily.

When a timestamp is present in more than one set (highlights/std/extended) there’s no need to re-capture it.

See also  FS#123 

No need for description

Bash 2 syntax has been the norm for most versions (although broken in some).

So far none of the major distributions is using bash 2, bash 4 is already gold but not widely-used enough.

Only a few special purpose/minimal distros like DSL retain bash 2.05b, but vcs requires recent versions of ImageMagick, Mplayer and FFmpeg anyway. I can’t see a reason to retain such backwards compatibility.

Parsing of the “-U” option (optional argument) is incorrect, it always receives an argument (empty if none provided) which must be shifted manually. Currently it’s only shifted if provided explicitly.

“–fullname” is not affected since it doesn’t take any arguments.