As you all can see there is not an awful lot of activity here, basically what has happened since I took over maintainership is that I have commited one or two contributed patches.
Since I am currently busy working on GIMP I expect project management to remain in status quo for the foreseeable feature.
This message is addressed at recordMyDesktop's users.
I believe it is time for me to step down completely
from the project and let someone else take it to new
Looking at some commit stats, it became clear that I
haven't made any real contributions to the project since
December, 2007. After that, all I did was mostly push
upstream any translations or patches that I got.
But I haven't had a vision for the application for a very
long time.... read more
As reported by users, v0.3.8 had a bug (some headers were missing), that prevented
recordMyDesktop to compile in some setups (RHEL4, Slackware 12.1 and possibly others).
Thanks to the thorough feedback received, this problem should now be resolved, in v0.3.8.1.
recordMyDesktop 0.3.8 is mostly a bugfixing release, driven by the contributions of Martin
Nordholts, who is now a cotributing member with an SVN account . Yay!
Here are the ChangeLogs:
* Extensive refactoring and cleanups of the code from Martin Nordholts.
* Cursor-flickering bug has been fixed, by Martin.
* Improved consistency on the format of the commandline options
and updated the manpage to reflect the changes (also by Martin ;) ).
* Applied patch by Luca Bonavita, that corrects the rgb
to yuv algorithm and provides near perfect colorspace conversion
(fixes bug #1961505)
* Compositing Window managers are now recognised through the _NET_WM_CM_Sn
hint instead of a simplistic namelist (which means that dual-mode wm's like
kwin or yet uknown ones will be handled correctly).
* Changed the way Jack support is offered, from dlopening at runtime to normal
linking at compilation.
* other minor changes and bugfixes.... read more
Fixed bug #1886713 that manifested as a
crash with the following message:
'recordmydesktop: xcb_xlib.c:82: xcb_xlib_unlock:
Assertion 'c->xlib.lock' failed.' ,
when recordMyDesktop was used with libxcb-xlib.
0.3.7.2 is (also)a quick release with the
1) Updated src/skeleton.c from libboggz trunk
(fix for big endian hosts), in recordMyDesktop.
2) Updated the Polish translation by Paweł
Smoliński, in the two frontends.
3) gtk-recordMyDesktop has now the
button/dialog for workdir selection,
that in the previous version was added
only in qt-recordMyDesktop.
This is a quick release, in order improve
compliance with the latest Xiph.org
1) Files contain an Ogg skeleton bitstream now
2) The extension is kept .ogv (serializing
happens before it and it is appended if missing)
bugfixes, new features and translations.
recordMyDesktop has now configurable, global shortcuts
and the ability to rescue(encode later) crashed sessions.
Also, it now draws a frame that surrounds the capture area
(this functionality existed on the frontends, but it was a
lot less accurate).
gtk/qt-recordMyDesktop have also gained shortcuts for the most
usefull actions of their interface.... read more
0.3.6 is mostly a bugfixing release dealing with some problems
introduced in 0.3.5.
Along with the bugfixes, there's a new feature, called follow-mouse,
where the capture area moves around following the mouse cursor.
The frames in qt/gtk-recordMyDesktop have been made compatible with
this behavior and also move around, but it might be better to just turn
them off, when using this feature. Continuously moving a window around
can make the Xserver use too much CPU. And there's also some latency by the
fact that the frame is drawn and moved by qt/gtk-recordMyDesktop and not
recordMyDesktop.... read more
This is a bugfixing release addressing some critical bugs:
* Corrected segmentation fault when recording non-fullscreen areas
All recording of areas with x and/or y not zero is broken in 0.3.5.
Many thanks to Sander Jansen for providing the bug-report and the
solution to this problem.
* Corrected incorrect allocation of queued sound buffers
(Caused segmentation faults in FreeBSD,
possibly on non x86 linux platforms, too)
* Corrected dlopen lookup on *bsd (libc instead of libdl),
which blocked recordMyDesktop from being compiled with
Jack support on this platform.
Thanks to Kevin Lo for reporting the above two problems.... read more
qt-recordMyDesktop-0.2 and gtk-recordMyDesktop-0.3.5
should be used recordMyDesktop == 0.3.5. Other versions
might create problems. The dependency goes both ways.
Corrected cursor top/left behavior, fixed xfixes cursor placing.
Double buffered capturing on full-shots, to convert only changed pixels to yuv.
Cache block information is now is more efficiently extracted from the damage list
(when using xdamage) and during the rgb2yuv conversion(when using full shots).
Jack client uniquely named, to allow simultaneous recordings when using it.
Updated man page.
Removed uneccesary calculation of the alpha component when
calculating average pixel value for the u and v planes.
Ub,Vr replaced with UvBr.... read more
qt-recordMyDesktop is a port of gtk-recordMyDesktop in PyQt4.
Minimum requirements to run it are Qt4.2, PyQt4.1 and
The interface and functionality provided is more or less identical
to that of gtk-recordMyDesktop.
This is also the very first release so it's expected to be rough around
the edges (at least more than gtk-recordMyDesktop).
Version 0.3.4 of the programs is incompatible to
any previous ones so if updating, please do so for
Fixed behavior of the program when using both the MIT-Shm and Xdamage extensions.
As a result, less memory is now required and all options related to conditional usage of
shared memory are gone. Shared memory is now on by default and the --with-shared
option is substituted by --no-shared.
Fixed configure script to exit when zlib is not found.
The --drop-frames option has been removed.
The dummy cursor has been updated to look better.... read more
recordMyDesktop is now available to Fedora users through
the Fedora Extras repository. So anyone running
Fedora Core 6 or Fedora devel (the upcoming Fedora Core 7),
can simply install it with
yum install recordmydesktop or
yum install gtk-recordmydesktop, for the gtk frontend.
(At the time of writting, gtk-recordMyDesktop might still be
at the stage of building, but it should become available soon)... read more
It appears that the previous release had some
issues when recording with JACK was performed.
Also the interface on configuring the ports was
not well thought and could end up too big to be
These issues are now resolved in 0.3.3.1 thanks
to extensive feedback and help in debugging
provided by Carlos Pino.
Also gtk-recordMyDesktop 0.3.3.1 has an updated Italian
translation by Giovanni Scafora.... read more
Changes in this release for recordMyDesktop:
i) Improved cache format. Blocks are now cached at 16x16 for y plane
and 8x8 for u,v planes. Thus, unneeded caching is minimized,
which can result in significant savings of disk space
ii) Added support for compilation with OSS instead of ALSA and made
minor other changes, to make compilation on FreeBSD possible
(recordMyDesktop 0.3.3 has now been tested to compile and run under
FreeBSD 6.2).... read more
This release has been focused on code refactoring.
The only notable change from a user perspective is that
resolutions with 16bpp color depth are now supported for
Also a bug appearing on the amd64 arch(possibly others, too),
that made the cursor appear distorted, has been fixed.
*This release is compatible with current latest gtk-recordMyDesktop(0.3.1.1).
And unless specifically noted otherwise, all 0.3.* version of the programs,
are to remain compatible with each other.
This is a bugfixing release for recordMyDesktop.
The most important changes are:
i) Fixed BadWindow/segfault bugs at program startup.
ii) Corrected sound buffer size/number of frames to be read in
every loop, which could cause sound to skip.
iii) Interleaving of audio and video streams is now handled properly.
iv) Fixed missing end-of-stream in generated files.
On the other hand, gtk-recordMyDesktop v0.3.1 is
a release focused in translations.... read more
The sid repository has been removed
from the downloads section, as recordMyDesktop
is now in the official Debian Sid repositories
Many thanks go to Jose L. Redrejo Rodriguez
who maintained the program in its previous repository,
is now the official Debian maintainer
and has helped in many ways this project.
A correction on this news item:
The gtk interface was designed by Daniel Campos
to whom I owe many thanks, too.... read more
gtk-recordMyDesktop is a graphical interface for recordMyDesktop,
a screencasting utility for linux.
The frontend currently needs to be translated to as many languages as possible,
in order to make it accessible to everyone.
The latest pot file can be obtained here:
It only consists of 77 messages, so it should be a very easy task.... read more
recordMyDesktop-0.3.0r2 bears only a correction in the project files
in order to avoid including unrelated files on the dist-tarball.
(a CVS directory has been included in all previous releases)
gtk-recordMyDesktop-0.3.0r2 is also out, including now a
french translation by David Valot
(who also was the one that reported the tarball bug in recordMyDesktop).
So, thanks David!
p.s. gtk-recordMyDesktop now also produces a log in case of a crash
during recording, to be included in bug reports.
Pausing is completely broken in 0.3.0, when
recording with sound and not encoding on the fly.
(the program keeps counting frames and it acts as
if all the pause-time has been lost frames)
This is fixed now in 0.3.0r1.
Italian translation has been updated and is
available in the new release of gtk-recordMyDesktop (0.3.0r1).
Thanks go to Giovanni Scafora for his work and for being so fast!
Version 0.3.0 of both the program and the frontend is out.
This release has some important changes in the way the program
functions, as now encoding happens on a step of its own, after
recording finishes(the old mode, where encoding was performed
on the fly is still available as an option).
Also, the program now checks if it is run under a known (by me)
3d compositing window manager and sets its behavior accordingly,
to avoid the problem, where only the cursor was recorded.... read more
The rMD-exp module is no longer in use and it won't be used again(it was a mistake anyway).
All changes have been commited in the recordmydesktop module and the latest version will
be available for checkout there. Last snapshot prior to the commit has been tagged and branched
Changes so far focus mainly around transforming the program in a two-stage recorder. This means
that encoding into theora takes place after the recording. To do this, a simple interframe format
is used for cache, with an optional light gz compression (at the time of writting, there is
some work needed on this yet, but basic functionality is there, if anyone wants to check it out).... read more