xpra icon
Bug tracker and wiki

Timeline


and

07/28/15: Today

13:17 Changeset in xpra [10102] by antoine
win32 whackamole (batch files are a bad joke): fix all the builds …
12:44 Changeset in xpra [10101] by antoine
try Werror with clang..
12:42 Versions/PendingFixes edited by antoine
(diff)
12:41 Changeset in xpra [10100] by antoine
found bug with clang: wrong type for hint
12:39 Changeset in xpra [10099] by antoine
enable strict warnings as errors with gcc, silence some clang warnings
12:38 Changeset in xpra [10098] by antoine
fix some clang warnings
12:12 CSC/Performance edited by antoine
(diff)
12:12 CSC/Performance edited by antoine
(diff)
12:03 CSC/Performance edited by antoine
(diff)
11:51 CSC/Performance edited by antoine
(diff)
11:28 Changeset in xpra [10097] by antoine
fix win32 build with py3k / without pycuda
10:49 Changeset in xpra [10096] by antoine
fix hmac warnings with newer versions of python
10:14 Changeset in xpra [10095] by antoine
thanks to the fixes in r10093 and r10094, we can now build with …
10:03 Changeset in xpra [10094] by antoine
more valid compiler warnings found via msvc
09:35 Versions/PendingFixes edited by antoine
(diff)
09:24 Changeset in xpra [10093] by antoine
fix compilation warning found with msvc
09:18 Changeset in xpra [10092] by antoine
fix errors with py3k build
08:44 Changeset in xpra [10091] by antoine
new upstream releases
07:53 Ticket #660 (stricter pre-release test plan) closed by antoine
fixed: Not heard back, closing - will follow up in #847.
07:47 CSC/Performance edited by antoine
(diff)
07:43 Changeset in xpra [10090] by antoine
r10089 for v0.15.x branch: also fix rgb to yuv
07:42 Changeset in xpra [10089] by antoine
same as r10087 but for rgb to yuv
07:38 CSC/Performance edited by antoine
(diff)
07:33 Ticket #927 (review our use of memoryview_to_bytes) created by antoine
As per #926, this can be expensive. We call it from the following …
07:27 Changeset in xpra [10088] by antoine
r10087 for v0.15.x branch: only use memoryview_to_bytes on memoryviews
07:26 Changeset in xpra [10087] by antoine
#926: only use memoryview_to_bytes on memoryviews
07:12 CSC/Performance edited by antoine
(diff)
07:11 CSC/Performance edited by antoine
(diff)
07:05 CSC/Performance edited by antoine
(diff)
06:54 CSC/Performance edited by antoine
(diff)
06:47 Changeset in xpra [10086] by antoine
r10084 for v0.14.x and v0.15.x branches: fix opencl platform and …
06:23 CSC/Performance edited by antoine
(diff)
06:01 Changeset in xpra [10085] by antoine
"macro" to try to butcher the english language a bit less!
05:56 Changeset in xpra [10084] by antoine
#926: opencl fixes: * fixup platform and device selection scoring * …
05:33 Changeset in xpra [10083] by antoine
new upstream release
04:59 CSC/Performance edited by antoine
(diff)
04:49 Ticket #926 (csc opencl performance has regressed) created by antoine
See wiki/CSC/Performance where I have added results for trunk. …
04:44 CSC/Performance edited by antoine
(diff)
04:40 CSC/Performance edited by antoine
(diff)
04:35 CSC/Performance edited by antoine
(diff)

07/27/15: Yesterday

16:47 Changeset in xpra [10082] by antoine
add latest fixes to the release notes
16:43 Changeset in xpra [10081] by antoine
use the same directory structure as for 0.15.x and trunk
16:23 Versions/PendingFixes edited by antoine
(diff)
16:22 Changeset in xpra [10080] by antoine
r10078 for v0.15.x branch: nvenc api correctness fix
16:20 Changeset in xpra [10079] by antoine
r10024 for v0.15.x branch: opencl fixes
15:59 Versions/PendingFixes edited by antoine
(diff)
15:54 Versions/PendingFixes edited by antoine
(diff)
15:53 Changeset in xpra [10078] by antoine
we should be setting this parameter, it exists for a reason
13:41 Changeset in xpra [10077] by joshiggins
html5 encryption * distinguish between cipher_in and cipher_out * …
12:38 Changeset in xpra [10076] by joshiggins
html5 protocol encryption * generate caps and pass through to the …
12:26 Changeset in xpra [10075] by joshiggins
html5 xpra_protocol make sure non-worker also has terminate function
12:10 Changeset in xpra [10074] by joshiggins
#850 html5 handle disconnection gracefully - only call the connection …
11:42 Changeset in xpra [10073] by joshiggins
html5 password authentication support * fix get_hex_uuid to remove …
10:41 Changeset in xpra [10072] by antoine
reduce debug logging (move more stuff to API logging)
10:20 Changeset in xpra [10071] by antoine
* require PyOpenCL 2015 or later (since we now use read-only buffers) …
10:02 Ticket #925 (opencl csc is broken) closed by antoine
invalid: Well, I have no idea what happened here, but it fixed itself after a …
09:55 Changeset in xpra [10070] by antoine
fix error logging of supported encoder csc modes
09:41 Changeset in xpra [10069] by antoine
always log to stderr, keep stdout for pipe comms
09:27 Changeset in xpra [10068] by antoine
* use integer division * log codec version in compress_image * remove …
06:24 Changeset in xpra [10067] by antoine
just try a few times and hope for the best!
06:05 Changeset in xpra [10066] by totaam
* log to stderr so it does not interfere when wrapping a subprocess …
05:42 Changeset in xpra [10065] by totaam
add HEVC GUID and stub definitions
05:11 Changeset in xpra [10064] by totaam
prettify error details, especially on win32
04:59 Changeset in xpra [10063] by antoine
only include VERSION_STATUS if there is something in it
04:54 Changeset in xpra [10062] by antoine
split cuda and pycuda info
04:37 Changeset in xpra [10061] by totaam
the error is a "GError" object, which cannot be serialized as it is …
04:30 Changeset in xpra [10060] by antoine
better debugging of pipeline setup failures
04:05 Changeset in xpra [10059] by antoine
* better log messages: let the prefix tell us if this is a source or a …

07/26/15:

17:07 Versions/PendingFixes edited by antoine
(diff)
17:05 Changeset in xpra [10058] by antoine
use read-only buffers (we'll worry about opencl later - it is broken …
17:01 Changeset in xpra [10057] by totaam
wavpackenc needs an audioconvert to avoid errors (at least on win32)
16:49 Changeset in xpra [10056] by totaam
functions don't need self... cut&paste gone wrong
16:43 Versions/PendingFixes edited by antoine
(diff)
16:40 Changeset in xpra [10055] by antoine
r10026 + r10053 for v0.14.x and v0.15.x branches: fix shadow server …
16:23 Changeset in xpra [10054] by antoine
patch made a complete mess of it again, how on earth did that apply at …
16:08 Changeset in xpra [10053] by antoine
#637: refactor the code into utility functions and ensure the bug …
15:48 Changeset in xpra [10052] by antoine
r9921 for v0.14.x and v0.15.x branches: tidy up win32 build library paths
15:36 Changeset in xpra [10051] by antoine
shoot me now
15:35 Changeset in xpra [10050] by antoine
doh
15:31 Changeset in xpra [10049] by antoine
r10040 for v0.14.x and v0.15.x branches: fix log file error on win32 …
15:25 Changeset in xpra [10048] by antoine
allow us to have version-specific lib dirs
15:23 Changeset in xpra [10047] by antoine
r10046 for v0.14.x and v0.15.x branches: avoid error when no shadow …
15:17 Versions/PendingFixes edited by antoine
(diff)
15:16 Changeset in xpra [10046] by totaam
win32 and osx don't need a display name, and trying to guess one …
15:04 Changeset in xpra [10045] by totaam
tidy up the win32 build some more: move log files to win32 subdir and …
11:16 Changeset in xpra [10044] by totaam
* use integer division * log exact kernel arguments used
09:55 Versions/PendingFixes edited by antoine
(diff)
07:43 Changeset in xpra [10043] by antoine
fat warning for older versions of CUDA
07:07 Changeset in xpra [10042] by antoine
make lz4 version parsing and checking more robust and future proof
06:32 Changeset in xpra [10041] by antoine
this is the "standard" for pre-release package version naming I believe
06:22 Versions/PendingFixes edited by antoine
(diff)
06:21 Changeset in xpra [10040] by totaam
move the bug report launch code to a script so we can delay all …
06:02 Changeset in xpra [10039] by totaam
move imports to where they are used: importing this module is now more …
05:51 Versions/PendingFixes edited by antoine
(diff)
05:50 Changeset in xpra [10038] by totaam
we now assume the client has a "signal_disconnect_and_quit" method

07/24/15:

17:27 Versions/PendingFixes edited by antoine
(diff)
16:59 Changeset in xpra [10037] by antoine
important part of r10036 for v0.14.x and v0.15.x branches: only call …
16:44 Changeset in xpra [10036] by totaam
#918: exit fixes * don't assume closed flag is set in …
16:19 Changeset in xpra [10035] by totaam
update to new path api
16:16 Changeset in xpra [10034] by totaam
#918: add cleanup debug logging
16:16 Changeset in xpra [10033] by totaam
#918: add cleanup debug logging
14:42 Changeset in xpra [10032] by antoine
unlikely race, but I've managed to hit it!
14:40 Changeset in xpra [10031] by antoine
to restore the original behaviour, we just pass None!
11:28 Versions/PendingFixes edited by antoine
(diff)
11:27 Ticket #925 (opencl csc is broken) created by antoine
No idea when it broke, but it isn't doing anything anymore, just …
11:13 Versions/PendingFixes edited by antoine
(diff)
11:12 Changeset in xpra [10030] by antoine
r10029 for v0.14.x and v0.15.x branches: ensure the bug report tool …
11:07 Versions/PendingFixes edited by antoine
(diff)
11:05 Changeset in xpra [10029] by antoine
#637: re-create the bug report window whenever we need to (ie: after …
10:36 Changeset in xpra [10028] by totaam
#389: keep the same context until things change, log performance info
09:55 Changeset in xpra [10027] by antoine
make it possible to access the bug report tool from the launcher
09:37 Changeset in xpra [10026] by totaam
#637: capture screen using native pywin32 calls
09:33 Changeset in xpra [10025] by antoine
#888: forgot to update this caller
04:08 Versions/PendingFixes edited by antoine
(diff)
04:07 Changeset in xpra [10024] by antoine
* fix for memoryview: convert to bytes before upload * fix for …
03:47 Changeset in xpra [10023] by antoine
having only one encoding is usually a user mistake, turn it into a warning

07/23/15:

16:58 Changeset in xpra [10022] by antoine
* use stronger types * use the same logic as encoder for the number of …
16:41 Changeset in xpra [10021] by antoine
more of the same work on splitting api and regular debug
16:39 Changeset in xpra [10020] by antoine
fix invalid format string
14:54 Ticket #924 (OSX client - "Xpra quit unexpectedly") created by kverne
I'm observing the following about once per day. This may be right …
11:33 Changeset in xpra [10019] by totaam
make the shadow refresh delay configurable
08:55 Changeset in xpra [10018] by totaam
* don't bother with "which" on win32 * the "52" target is a no-go with …
08:43 Ticket #923 (AltGr broken after Ctrl-Escape / Escape) created by fladnag
--- environment --- server : debian 8, xpra 0.14.10 client : win 7, …
08:34 Changeset in xpra [10017] by antoine
make it easier to debug nvenc issues with better debug logging, use …
08:32 Changeset in xpra [10016] by antoine
add the ability to select the pulseaudio monitor device name using …
07:41 Versions/PendingFixes edited by antoine
(diff)
06:50 Changeset in xpra [10015] by antoine
* easy strong typing and optimizations found using Cython annotated …
06:07 Versions/PendingFixes edited by antoine
(diff)
06:01 Changeset in xpra [10014] by antoine
remove unused (and potentially dangerous?) line of code
05:27 Changeset in xpra [10013] by antoine
option to use the annotated html reports from cython, exclude the html …

07/22/15:

23:58 Changeset in xpra [10012] by nickc
#797 Updated the chart generator to handle multiple directories …
19:29 Changeset in xpra [10011] by joshiggins
html5 client: make generating the hello more like the python client in …
18:56 Changeset in xpra [10010] by joshiggins
html5 client include forge.js for encryption and hmac
17:48 Changeset in xpra [10009] by antoine
fix cuda version comparison: compare with version number (as a …
17:43 Changeset in xpra [10008] by antoine
* detect the CUDA version and only build targets the SDK is capable of …
16:32 Changeset in xpra [10007] by joshiggins
html5 encryption groundwork
13:35 Changeset in xpra [10006] by antoine
don't log packet type warnings when closing down
13:19 Changeset in xpra [10005] by antoine
r9997 for v0.14.x and v0.15.x branches: the list of output colorspaces …
13:19 Versions/PendingFixes edited by totaam
(diff)
13:17 Changeset in xpra [10004] by antoine
r9996 for v0.14.x and v0.15.x branches: API says we can't assume …
13:15 Changeset in xpra [10003] by antoine
r10000 for v0.15.x branch: support nvidia cards with more than 2GB of …
12:47 Versions/PendingFixes edited by totaam
(diff)
12:40 Changeset in xpra [10002] by totaam
#637 and #389: capture screen contents using Pillow, fallback to GTK …
11:59 Changeset in xpra [10001] by antoine
doh: wrong None check!
11:55 Versions/PendingFixes edited by totaam
(diff)
11:54 Changeset in xpra [10000] by totaam
fix overflow with 4GB+ nvidia cards: use 64-bit to hold memory in bytes
07:13 Changeset in xpra [9999] by totaam
fix nvenc on win32: * use WinDLL to avoid calling convention error …
05:41 Packaging edited by totaam
(diff)
05:40 Packaging edited by totaam
(diff)

07/21/15:

12:37 Changeset in xpra [9998] by antoine
since those asserts may well fire, ensure they have meaningful error …
11:17 Versions/PendingFixes edited by totaam
(diff)
11:16 Changeset in xpra [9997] by antoine
fix output colorspaces containing garbage! (lucky this didn't match …
09:28 Versions/PendingFixes edited by totaam
(diff)
09:28 Changeset in xpra [9996] by antoine
Using x264_picture_init instead of memset has been recommended since …
09:25 Changeset in xpra [9995] by antoine
log and send back the slice type (only IDR and P for now)
09:02 Changeset in xpra [9994] by antoine
make it possible to log individual nals using XPRA_X264_LOG_NALS=1 env var
08:15 Packaging edited by totaam
(diff)
08:05 Packaging edited by totaam
(diff)
08:05 Packaging edited by totaam
(diff)
06:26 Packaging edited by totaam
(diff)
06:18 Packaging edited by totaam
(diff)
06:17 Packaging edited by totaam
(diff)
06:16 Packaging edited by totaam
(diff)
05:57 Changeset in xpra [9993] by antoine
add latest fixes to release notes
05:40 Changeset in xpra [9992] by antoine
add optional message to client ack, so we can log the error details …
05:39 Changeset in xpra [9991] by antoine
catch invalid packet data early so the error message will be more …
05:19 Ticket #886 (OSX GUI won't connect (password box issue)) closed by totaam
fixed: Not heard back, I assume this is fixed - feel free to re-open if …
05:16 Versions/PendingFixes edited by totaam
(diff)
05:15 Changeset in xpra [9990] by antoine
r9978 + r9987 for v0.14.x and v0.15.x branches: refresh the window if …
05:14 Changeset in xpra [9989] by antoine
v0.14.x doesn't have the 'sound-source' option
05:08 Changeset in xpra [9988] by antoine
r9868 for v0.14.x and v0.15.x branches: look for Fedora 22 Xorg …
05:01 Changeset in xpra [9987] by antoine
don't schedule a refresh if the window is gone (it might be)
04:58 Versions/PendingFixes edited by totaam
(diff)
04:57 Changeset in xpra [9986] by antoine
r9983 for v0.14.x and v0.15.x branches: fix csc cython red and blue …
04:47 Versions/PendingFixes edited by totaam
(diff)

07/20/15:

18:29 Changeset in xpra [9985] by antoine
#465: support memoryview when running csc tests
18:07 Ticket #894 (0.15.1 client error: do_paint_rgb32) closed by totaam
needinfo: Unable to reproduce, so closing as needinfo.
18:06 Versions/PendingFixes edited by totaam
(diff)
18:05 Ticket #922 (csc_cython produces images with red and blue channels swapped) closed by totaam
fixed: trunk fix in r9983: I'm just not clever enough to figure out how to …
18:03 Changeset in xpra [9984] by antoine
this probably fixes big endian byte order
18:02 Changeset in xpra [9983] by antoine
#922: better to hardcode correct values than derive wrong ones
17:58 Changeset in xpra [9982] by antoine
use C "const" keyword where it matters
16:48 Changeset in xpra [9981] by antoine
* speedup clamp: make it inline, use const * log the byteorder (is it …
16:24 Changeset in xpra [9980] by antoine
support packed 24-bit colour output modes BGR and RGB
16:07 Changeset in xpra [9979] by antoine
use C "const" keyword where it matters

07/18/15:

17:54 Versions/PendingFixes edited by totaam
(diff)
17:53 Ticket #861 (xor error with differing pixel format) closed by totaam
fixed: r9978 ensures we refresh the whole window if we ever get client-side …
17:50 Changeset in xpra [9978] by antoine
#861: if we do get a decoding error, force a full refresh
17:17 Changeset in xpra [9977] by antoine
why paint with cairo for rgb32 when gdk.Drawable already has a …
17:10 Ticket #922 (csc_cython produces images with red and blue channels swapped) created by totaam
We don't use it much because we prefer opengl, and even opengl is …
16:36 Changeset in xpra [9976] by antoine
#861: verify that the pixel format is the same, makes the error more …
16:30 Changeset in xpra [9975] by antoine
* log reason for rgb reformatting * add docstring to explain delta …
16:26 Changeset in xpra [9974] by antoine
minor speedup: use more local variables
14:46 Changeset in xpra [9973] by antoine
update release notes with latest fixes
14:36 Versions/PendingFixes edited by totaam
(diff)
14:36 Versions/PendingFixes edited by totaam
(diff)
14:35 Changeset in xpra [9972] by antoine
Fedora 20 is no longer supported and Fedora 21 onwards have PyOpenGL …
14:08 Changeset in xpra [9971] by antoine
r9970 for v0.14.x and v0.15.x branches: fix delta compression errors
14:04 Changeset in xpra [9970] by antoine
#861: don't try to use delta if the pixel format is not supported by …
13:06 Changeset in xpra [9969] by antoine
* raise an exception if we fail to setup the shaders, as we cannot …
06:16 Changeset in xpra [9968] by antoine
don't try to set the WM_COMMAND x11 property before the window is realized!

07/17/15:

17:41 Usage/Docker edited by timthelion
(diff)
14:10 Versions/PendingFixes edited by totaam
(diff)
14:09 Changeset in xpra [9967] by antoine
* integrity hash does not work with mmap! * use stricter exception type
14:07 Changeset in xpra [9966] by antoine
* move more pillow code to the pillow module, * remove broken test …
11:27 Changeset in xpra [9965] by antoine
same as python2: use the Fedora package for opengl (remove accelerate …
10:05 Usage/OpenGL/Performance edited by totaam
(diff)
09:57 Changeset in xpra [9964] by antoine
watching for properties is already done in superclass!
09:51 Changeset in xpra [9963] by antoine
#794: allow the WM_COMMAND to be updated, moves the code to a …
09:33 Changeset in xpra [9962] by antoine
if gtk won't let us change the wmclass without spewing out warnings, …
09:31 Changeset in xpra [9961] by antoine
update tests and keyboard helper to new gtk2 x11 display source class …
09:11 Changeset in xpra [9960] by antoine
* move "class-instance" to a more generic place (but add warning …
08:51 Changeset in xpra [9959] by antoine
better error reporting when the "lpadmin" commands fails to execute
08:48 Ticket #921 (native win32 opengl client) created by totaam
This may help with testing and with things like #916, #917. It …
08:29 Ticket #920 (faster window video source logic) created by totaam
Even though this was much improved in 0.15, we still spend too much …
06:34 Changeset in xpra [9958] by antoine
also log device info
05:13 Changeset in xpra [9957] by antoine
#915: make it possible to parse a "pactl list" output file from the …
04:50 Versions/PendingFixes edited by totaam
(diff)
04:46 Changeset in xpra [9956] by antoine
#918: when we control-C the client, the cleanup can go wrong - catch …
04:32 Ticket #885 (honour window position exactly) closed by totaam
fixed
04:20 Changeset in xpra [9955] by antoine
#919: add "frame" debug logging
04:19 Ticket #919 (frame extents synchronization) created by totaam
Split from #885 where most of the work was done in time for 0.15. …
04:03 Changeset in xpra [9954] by antoine
add osx wrapper for the gui info
03:42 Changeset in xpra [9953] by antoine
#885: some properties should not be sent back to the client (usually …
03:34 Ticket #918 (windows 0.15.4 client hangs on control-c disconnect after changing ...) created by afarr
Testing with windows 8.1 client 0.15.4 r9951 against fedora 20 0.15.4 …
03:22 Changeset in xpra [9952] by antoine
move the srub list to the superclass where we handle the properties

07/16/15:

21:04 Ticket #917 (Use Msys2 for building Windows 32 and 64 bit builds) created by smo
I've been trying out MSYS2 after you mentioned it in another ticket so …
11:42 Changeset in xpra [9951] by antoine
these errors should be rare, so always include the stacktrace
09:43 Changeset in xpra [9950] by totaam
also ignore all the files generated by the win32 build system (.pyd)
09:31 Changeset in xpra [9949] by antoine
win32 error messages can be long, split lines
09:30 Changeset in xpra [9948] by antoine
oops: remove unused arg
09:25 Changeset in xpra [9947] by antoine
pillow minor fixes: * prettify error logging * remove encoding from …
08:38 Changeset in xpra [9946] by antoine
more helpful log warning with failed decoding
08:18 Changeset in xpra [9945] by antoine
we have to actually read / do something with the image data to ensure …
08:11 Changeset in xpra [9944] by antoine
oops: missed from r9935
08:03 Changeset in xpra [9943] by antoine
update release notes with latest fixes
07:53 Versions/PendingFixes edited by totaam
(diff)
07:53 Changeset in xpra [9942] by antoine
r9925 for v0.14.x and v0.15.x branches: safer cleanup / exit code on win32
07:52 Changeset in xpra [9941] by antoine
r9924 for v0.15.x branch: safer opengl platform checks
07:50 Changeset in xpra [9940] by antoine
r9918 for v0.14.x and v0.15.x branches: fix pulseaudio device count …
07:48 Changeset in xpra [9939] by antoine
r9898 for v0.14.x and v0.15.x branches: fix timer warnings and logging …
07:47 Changeset in xpra [9938] by antoine
r9915 for v0.14.x and v0.15.x branches: workaround invalid …
07:45 Changeset in xpra [9937] by antoine
r9888 for v0.15.x branch: if we get gibberish at either end, something …
07:45 Changeset in xpra [9936] by antoine
version bump
07:26 Changeset in xpra [9935] by antoine
* move pillow encoding and decoding to a codec module like most of the …
07:23 Changeset in xpra [9934] by antoine
* expose more details via get_info, include buffer api version * allow …
07:22 Changeset in xpra [9933] by antoine
ensure we discard encodings that have failed the selftests
06:36 Changeset in xpra [9932] by antoine
* new upstream release, required for running latest pycuda * merge …

07/15/15:

18:45 Changeset in xpra [9931] by totaam
tweaks needed for building nvenc5 on win32, we also have to force …
18:40 Changeset in xpra [9930] by totaam
we no longer need to support python2.4, remove the ugly code
17:48 Changeset in xpra [9929] by totaam
#916: use my directory structure if present, but keep the same default
15:42 Changeset in xpra [9928] by antoine
posix paths options should not have the command in them!
15:36 Changeset in xpra [9927] by totaam
allows us to build nvenc 4 and 5 on win32
14:57 Changeset in xpra [9926] by antoine
some compilers (like msvc) don't know about float_t
14:17 Versions/PendingFixes edited by totaam
(diff)
14:17 Changeset in xpra [9925] by antoine
I have seen an error here, logging a warning is a lot better than …
13:47 Versions/PendingFixes edited by totaam
(diff)
13:46 Changeset in xpra [9924] by antoine
the opengl platform checks should run before we try to import gtkgl, …
10:59 Changeset in xpra [9923] by totaam
#916: * minor quote fixing * exclude msys dlls for now
07:47 Changeset in xpra [9922] by totaam
#916 * fix gstreamer location on 64-bit builds * fix quoting problems …
06:08 Changeset in xpra [9921] by totaam
#916 allows us to move all the libs into a prefix (still "C:\" by default)
05:40 Ticket #916 (tidy up the win32 build) created by totaam
Related to #678 and #640. * building on 64-bit windows system is …
05:22 Changeset in xpra [9920] by totaam
make it easier to build on a 64-bit os: look for both program files …

07/14/15:

17:12 Changeset in xpra [9919] by antoine
fix typo in changelog
12:44 Versions/PendingFixes edited by totaam
(diff)
12:43 Changeset in xpra [9918] by antoine
fix device count
09:31 Changeset in xpra [9917] by antoine
allow platforms to supply alternative client implementations
06:37 Changeset in xpra [9916] by antoine
new upstream release
03:46 Logging edited by totaam
(diff)

07/13/15:

20:07 Ticket #915 (Sound Forwarding to client doesn't work (for me)) created by fervi
I tried to do with Xpra from version 0.14 to 0.16 (Beta of Winswitch) …
11:15 Versions/PendingFixes edited by totaam
(diff)
11:14 Changeset in xpra [9915] by antoine
ensure "help" options mistakenly stored in the config file won't make …
10:49 Changeset in xpra [9914] by antoine
document all the debug flags and show them using the "help" magic value
07:08 Changeset in xpra [9913] by antoine
more friendly error message: tell the user which socket failed
06:14 Changeset in xpra [9912] by antoine
#878: check the underlying liblz4 version if we have it
06:12 Changeset in xpra [9911] by antoine
#907: "focused" is now exported as part of wm-state, so we must define …
05:40 Changeset in xpra [9910] by antoine
#878: make a 0.8.0-rc1 build for testing

07/12/15:

05:38 Changeset in xpra [9909] by antoine
doh: should run lint before commiting: 'changelog not in descending …
05:25 Changeset in xpra [9908] by antoine
version bump

07/10/15:

11:13 Ticket #914 (screen is frozen until the window is resized/moved) created by azinuamid
We have a workstation (with Nvidia quadro + centos 7) with xpra. Of …

07/08/15:

14:53 Ticket #913 (html5 printing) created by joshiggins
Looking at #598 it seems as though we can register a printer to be …
11:33 Versions/PendingFixes edited by antoine
(diff)
11:25 Changeset in xpra [9907] by antoine
fix race and abort if we've lost the process already
11:20 Changeset in xpra [9906] by antoine
make it easier to debug fd leaks, add XPRA_DETECT_FDLEAKS env var
11:18 Ticket #912 ("Too many open files" caused by 0.14 clients with trunk (0.16) ...) closed by antoine
fixed: The leak was coming from the new palib (used in trunk only) and not …
11:16 Changeset in xpra [9905] by antoine
#912: the leak comes from palib - don't use it!
10:45 Changeset in xpra [9904] by antoine
refactor encoder and compression setup into a common utility function
10:28 Changeset in xpra [9903] by antoine
support any of the available packet encoders, without causing warnings
09:58 Changeset in xpra [9902] by antoine
more helpful message?
09:55 Changeset in xpra [9901] by antoine
don't log each accelerated handler at info level
09:35 Changeset in xpra [9900] by antoine
don't export the window "frame", we only set it
09:29 Changeset in xpra [9899] by antoine
add fault injection option for paints that we fail to report to the server
09:22 Versions/PendingFixes edited by antoine
(diff)
09:22 Changeset in xpra [9898] by antoine
* only remove timers that are still active (prevents source_remove …
06:54 Ticket #912 ("Too many open files" caused by 0.14 clients with trunk (0.16) ...) created by antoine
My OSX clients running in vbox get lots of restarts, which makes us …
06:40 Changeset in xpra [9897] by antoine
fix missing encoding from log output
05:41 Changeset in xpra [9896] by antoine
OR and trays are always shown
05:05 Changeset in xpra [9895] by antoine
don't bother trying to load the gtk2-x11 bindings on osx
02:48 Ticket #911 (GTK Menus not functioning) created by jasom
Description: All menus in GTK applications cannot be clicked on. …

07/07/15:

18:15 Changeset in xpra [9894] by antoine
#910: allows us to inject paint faults
18:00 Ticket #910 (fault injection for testing and triggering bugs) created by antoine
Hopefully this can help us with hard to trigger bugs and for catching …
17:45 Changeset in xpra [9893] by antoine
pretty logging
17:45 Changeset in xpra [9892] by antoine
"None" fault injection code for bencode + sound
17:26 Changeset in xpra [9891] by antoine
gibberish should trigger a warning (actually more than one - better …
16:56 Changeset in xpra [9890] by antoine
more logging tweaks: * better formatting for encodings (there are many …
16:55 Changeset in xpra [9889] by antoine
add fault injection hooks for testing using env var
16:41 Versions/PendingFixes edited by antoine
(diff)
16:40 Changeset in xpra [9888] by antoine
if we get gibberish at either end, something is wrong and we should stop!
16:04 Changeset in xpra [9887] by antoine
* more consistent log message format * avoid logging video failures …
15:57 Ticket #909 (10bpc support (40 bit color)) created by lapsio
I'm not sure if it's difficult or just few lines of code but in case …
15:43 Changeset in xpra [9886] by antoine
prettify warning message format
15:42 Changeset in xpra [9885] by antoine
avoid using gobject properties for simple calls
15:29 Ticket #908 (--with-Xdummy --with-Xdummy_wrapper options not honored across ...) created by jonathan.underwood
It used to be the case that you could specify {{{--with-Xdummy …
14:31 Changeset in xpra [9884] by antoine
add latest fixes to release notes
14:18 Versions/PendingFixes edited by antoine
(diff)
14:17 Changeset in xpra [9883] by antoine
parts of r9870 for v0.14.x and v0.15.x branches: * fix "print" with …
14:14 Versions/PendingFixes edited by antoine
(diff)
14:13 Changeset in xpra [9882] by antoine
r9820 + r9881 fixup for v0.15.x branch: log internal setup failures
14:11 Changeset in xpra [9881] by antoine
fixup r9820: compare the exception type!
14:10 Changeset in xpra [9880] by antoine
r9871 for v0.14.x and v0.15.x branches: ignore expose events before …
14:05 Changeset in xpra [9879] by antoine
r9851 for v0.14.x and v0.15.x branches: more reliable and clean …
14:03 Changeset in xpra [9878] by antoine
r9857 for v0.15.x branch: remove unsafe deprecated api
14:02 Changeset in xpra [9877] by antoine
cosmetic: consistent use of tabs vs spaces
12:10 Changeset in xpra [9876] by antoine
again, this shouldn't matter - but is cleaner
12:09 Changeset in xpra [9875] by antoine
this shouldn't make any difference since we are cleaning up at that …
12:03 Changeset in xpra [9874] by antoine
better warning message format and contents
12:03 Changeset in xpra [9873] by antoine
try to claim we support more window types - but not dnd yet, …
11:22 Changeset in xpra [9872] by antoine
new upstream release
11:17 Versions/PendingFixes edited by antoine
(diff)
11:15 Changeset in xpra [9871] by antoine
if we get an expose event before we're ready to paint the screen, just …
11:07 Changeset in xpra [9870] by antoine
control command refactoring: use a much nicer OO model and generic …
11:04 Changeset in xpra [9869] by antoine
#888: don't query the same real path twice
10:55 Changeset in xpra [9868] by antoine
we must also use this path to probe the Xorg version
08:01 Changeset in xpra [9867] by antoine
#888: ensure we don't pass None values to dotxpra
08:00 Changeset in xpra [9866] by antoine
support "--start-new-commands" without value specified, prevents weird …
04:33 Changeset in xpra [9865] by antoine
* cleaner and more generic automatic removal of encoding values …
04:13 Versions/PendingFixes edited by antoine
(diff)
02:53 Changeset in xpra [9864] by antoine
clarify what this is for

07/06/15:

18:07 Changeset in xpra [9863] by antoine
better connection failure messages (remove stacktrace)
17:57 Changeset in xpra [9862] by antoine
#888: when socker-dir is specified, we may still want to look for …
17:32 Changeset in xpra [9861] by antoine
don't bother sending the encoding_properties that are identical to the …
17:02 Changeset in xpra [9860] by antoine
use the dbus common code to get hold of the same systembus instance, …
16:55 Changeset in xpra [9859] by antoine
don't log a warning if there are no avahi publishers to start!
16:22 Versions/PendingFixes edited by antoine
(diff)
16:20 Changeset in xpra [9858] by antoine
bump version
16:19 Changeset in xpra [9857] by antoine
we've now removed "csc_modes" in favour of "full_csc_modes" which has …
15:50 Changeset in xpra [9856] by antoine
remove the old "csc_modes" option, we should always use …
15:35 Changeset in xpra [9855] by antoine
if we now know the socket dir, keep track of it
15:30 Changeset in xpra [9854] by antoine
"socket_dir" may not be set!
15:24 Changeset in xpra [9853] by antoine
oops: remove debug code..
15:23 Changeset in xpra [9852] by antoine
oops: other callers still need "cleanup_all_protocols"
15:19 Versions/PendingFixes edited by antoine
(diff)
15:18 Changeset in xpra [9851] by antoine
#896: more reliable exit (fixes server refusing to die with ssh …
15:01 Changeset in xpra [9850] by antoine
add missing ABI version
13:43 Changeset in xpra [9849] by antoine
prettify ssh exception handling
12:17 Changeset in xpra [9848] by antoine
#888: create the log_dir if we are going to need it
12:07 Changeset in xpra [9847] by antoine
improve opengl flush debug logging
11:50 Changeset in xpra [9846] by antoine
#792 when we process multiple regions at the same time (either a split …
10:37 Changeset in xpra [9845] by antoine
remove some outdated flags (assumed enabled since at least 0.12.x) and …
09:00 Changeset in xpra [9844] by antoine
* enable shape extension support for all X11 windows (including trays …
08:32 Changeset in xpra [9843] by antoine
more friendly error message format
06:42 Ticket #759 (xpra erroneously sends meta key from windows) reopened by antoine
As per comment:14, the bug is still there?
06:36 Changeset in xpra [9842] by antoine
#888 tmpfiles.d conf and packaging
06:31 Changeset in xpra [9841] by antoine
ugly (and probably wrong) py3k fix
05:42 Changeset in xpra [9840] by antoine
#888: use /var/run instead of /run by default, as this is more …

07/05/15:

16:20 Changeset in xpra [9839] by antoine
#907: move away from ugly do_get_property_XYZ overrides and update the …
12:55 Changeset in xpra [9838] by antoine
#907 use managed_connect wherever possible
12:53 Changeset in xpra [9837] by antoine
wm no longer manages the xkb bell
12:53 Changeset in xpra [9836] by antoine
#907: * use managed_connect wherever possible * route bell events …
12:18 Changeset in xpra [9835] by antoine
* even with multiple clients, the server doesn't need 1 second to …
11:59 Versions/PendingFixes edited by antoine
(diff)
11:59 Changeset in xpra [9834] by antoine
update release notes with latest fixes
11:55 Changeset in xpra [9833] by antoine
remove cuda_common now that nvenc is gone
11:51 Changeset in xpra [9832] by antoine
r9799 for v0.15.x branch: force enable can bypass platform checks
11:45 Usage/Docker edited by antoine
(diff)
11:41 Changeset in xpra [9831] by antoine
r9796 for v0.15.x branch: support the new socket: syntax for …
11:36 Changeset in xpra [9830] by antoine
r9770 for v0.14.x and v0.15.x branches: more helpful dpi warning
11:34 Changeset in xpra [9829] by antoine
r9803 for v0.15.x and nvenc3 codec
11:33 Changeset in xpra [9828] by antoine
fix error handling of missing cuda bin files
11:31 Changeset in xpra [9827] by antoine
r9808 for v0.15.x branch: default socket permissions should be 0600
11:30 Changeset in xpra [9826] by antoine
r9784 for v0.14.x and v0.15.x branches: fix unhandled failure code …
11:26 Changeset in xpra [9825] by antoine
don't internalize an invalid window type atom
11:24 Changeset in xpra [9824] by antoine
make it easier to simulate hello errors
11:12 Changeset in xpra [9823] by antoine
oops, missed from r9822
11:12 Changeset in xpra [9822] by antoine
#907: * split the huge window class into the models submodule (but …
11:05 Changeset in xpra [9821] by antoine
* always use full frames for tray * don't assume tray is OR * don't …
10:11 Changeset in xpra [9820] by antoine
log at warn level all failures except "Unamanageable" ones
07:56 Changeset in xpra [9819] by antoine
trays cannot ever be fullscreen!
05:08 Changeset in xpra [9818] by antoine
log xshm flag
04:52 Changeset in xpra [9817] by antoine
#907 add support functions for XGetWMProtocols, getParent wrapper for …
04:51 Changeset in xpra [9816] by antoine
use stricter parsing method

07/03/15:

16:27 Ticket #616 (HEVC single frame encoding support) closed by antoine
wontfix: #784 is more likely to succeed.
16:25 Versions/PendingFixes edited by antoine
(diff)
16:25 Ticket #28 (large windows on android will cause out-of-memory problems) closed by antoine
wontfix: See ticket:29#comment:6
16:23 Ticket #29 (android client does not pass key events at all!) closed by antoine
wontfix: I think the way forward is to abandon this code and use …
14:33 Ticket #907 (refactor the window attributes and signal code) created by antoine
As it's a bit of a mess... In particular: * keep in mind #41.. * try …
08:47 Changeset in xpra [9815] by antoine
minor tweaks to showconfig: no xvfb or displayfd with py3k, dbus-proxy …
08:30 Changeset in xpra [9814] by antoine
* add pulseaudio command to options which are generated during build …
06:28 Changeset in xpra [9813] by antoine
#888: split the fixups from the part that applies the settings, which …
06:21 ReportingBugs edited by antoine
(diff)
05:43 Changeset in xpra [9812] by antoine
oops: missed from r9802
05:40 Changeset in xpra [9811] by antoine
refactoring more platform bits for "showconfig": * move os bits from …
03:29 Changeset in xpra [9810] by antoine
try a bit harder to bunch up the CUDA log output in one logical chunk

07/02/15:

18:19 Versions/PendingFixes edited by antoine
(diff)
16:29 Ticket #906 (Disable or change local keyboard shortcuts on OS X when ...) created by jasom
The Xpra client registers a few keyboard shortcuts (e.g. ⌘Q ⌘H). …
16:05 Changeset in xpra [9809] by antoine
#888 showconfig subcommand constify DEFAULT_SSH_COMMAND properly …
15:42 Changeset in xpra [9808] by antoine
default socket permissions should be 0600! (that's what we have in the …
15:26 Changeset in xpra [9807] by antoine
only add the magic value to the environment if we are going to need it …
15:24 Changeset in xpra [9806] by antoine
#888: only instantiate DotXpra? if we actually need it
15:17 Changeset in xpra [9805] by antoine
despite the nested try+except, this preserves the exception
15:05 Changeset in xpra [9804] by antoine
use a more consistent message format, log level and indentation - and …
14:55 Changeset in xpra [9803] by antoine
fail slightly more gracefully when the cuda bin file is missing
13:35 Changeset in xpra [9802] by antoine
#888 * add "socket-dirs" so we can look for sockets in multiple …
12:33 Changeset in xpra [9801] by antoine
also missed from r9720
10:23 Changeset in xpra [9800] by antoine
add newline to stdout before logging the signal info (prevents C from …
09:49 Changeset in xpra [9799] by antoine
allow us to force enable opengl even platform checks would have it disabled
09:38 Changeset in xpra [9798] by antoine
fix the new libav module build on win32: * add the ugly hard-coded …
08:53 Changeset in xpra [9797] by antoine
"raise" preserves the stack, "raise e" does not!
03:01 Versions/PendingFixes edited by antoine
(diff)
03:01 Changeset in xpra [9796] by antoine
#888 support finding the socket by explicit path

07/01/15:

18:23 Changeset in xpra [9795] by antoine
the batch delay should always be an int (we don't care about sub …
18:21 Changeset in xpra [9794] by antoine
calling cleanup here may help in ensuring that the wrapped object …
18:00 Changeset in xpra [9793] by antoine
avoid this warning: "gbrp output is not supported with half chroma …
17:35 Changeset in xpra [9792] by antoine
better libvpx ABI error messages, log warning if we can't do YUV444P: …
17:28 Changeset in xpra [9791] by antoine
work on a copy of the list so we display the correct values in the …
17:10 Changeset in xpra [9790] by antoine
new upstream release
17:07 Ticket #905 (newer libav versions fail to decode vp9) closed by antoine
fixed: Re-enabled VP9+YUV444P decoding in r9789 and added some warnings …
17:04 Changeset in xpra [9789] by antoine
#905 / #832: * don't even bother trying VP9 with versions older than …
16:16 Changeset in xpra [9788] by antoine
ensure the rgb threshold is never bigger than half the video region
15:50 Changeset in xpra [9787] by antoine
* always keep the batch delay as an int, we don't care about …
15:24 Versions/PendingFixes edited by antoine
(diff)
15:21 Changeset in xpra [9786] by antoine
#905: don't fail the whole decoder module if one of the encodings has …
15:09 Ticket #905 (newer libav versions fail to decode vp9) created by antoine
Which is a big problem since that makes us fail the whole …
14:48 Changeset in xpra [9785] by antoine
#902: make it possible to see dec_avcodec2 log messages using our own …
10:00 Versions/PendingFixes edited by antoine
(diff)
09:59 Changeset in xpra [9784] by antoine
#902: a zero return value also indicates a failure, just an unspecified one
09:29 Ticket #904 (dbus interface for control channel) created by antoine
Some of the server's features could be exposed via dbus, this would …
08:20 Changeset in xpra [9783] by antoine
r9782 for v0.14.x and v0.15.x branches
08:06 Ticket #903 (Please update to GStreamer 1.x) closed by antoine
duplicate: As of 0.16, it will be possible to run gstreamer 1.x with the GTK2 …
08:02 Ticket #903 (Please update to GStreamer 1.x) created by onlyjob
As reported in …
07:33 Changeset in xpra [9782] by antoine
Fedora's choice of packaging breaks our repo, fun
07:24 Changeset in xpra [9781] by antoine
very minor upstream build fix - just bumping the version here, …
02:50 Ticket #902 (getting h264 Missing reference picture errors client-side, while page ...) created by afarr
Using 0.14.x clients (anywhere from 0.14.20 to 0.14.25, win32, osx, …

06/30/15:

13:15 Changeset in xpra [9780] by antoine
minor debug logging tweaks
13:12 Usage/OpenGL edited by antoine
(diff)
12:27 Changeset in xpra [9779] by antoine
add ability to track where the server thinks the mouse is (which could …
12:25 Changeset in xpra [9778] by antoine
ensures exceptions get propagated (void does not), and add logging …
12:24 Changeset in xpra [9777] by antoine
missed from r9720
11:30 Changeset in xpra [9776] by antoine
cleanup old junk code
09:19 Changeset in xpra [9775] by antoine
#896 now that we don't use daemon threads in the protocol class, we …
08:27 Changeset in xpra [9774] by antoine
add ability to build with or without dbus bits
08:26 Changeset in xpra [9773] by antoine
move dbus bits to their own module, not in x11
08:23 Changeset in xpra [9772] by antoine
let's turn on this warning, so we can find the types that don't work …
08:07 Changeset in xpra [9771] by antoine
welcome to 2015, and show it as text
06:10 Changeset in xpra [9770] by antoine
add more helpful dpi warning
04:53 WikiStart edited by antoine
link to the plain html version instead (diff)
04:52 WikiStart edited by antoine
(diff)
00:40 Ticket #901 (Windows Client crash after locking screen) created by maxmylyn
Versions: * Windows 8.1 trunk r9554 client http://xpra.org/beta * …

06/29/15:

21:07 Ticket #900 (google chrome fail to start) closed by jiang.qian
fixed: I can verify that this is fixed in 0.15.2-2. Thank you for being so …
14:42 Changeset in xpra [9769] by antoine
force package rebuild with decorations fixup
14:38 Changeset in xpra [9768] by antoine
fixup r9753
11:59 Changeset in xpra [9767] by antoine
uuid is not really used for much, don't log it
07:58 Ticket #900 (google chrome fail to start) created by jiang.qian
After upgrading to xpra 0.15.2, from the apt repository, google chrome …
05:52 Building edited by antoine
(diff)

06/28/15:

17:00 Changeset in xpra [9766] by antoine
had problems with pypi, use the upstream source instead
16:51 Changeset in xpra [9765] by antoine
fix md5
16:50 Changeset in xpra [9764] by antoine
bump version of ffmpeg to 2.7.1
16:05 Changeset in xpra [9763] by antoine
use same level of logging: with 0.14.x, this may actually fire
15:47 Changeset in xpra [9762] by antoine
r9761 for v0.14.x and v0.15.x branches: avoid warning with shadow servers
14:49 Changeset in xpra [9761] by antoine
we may also call disconnect: same as connect, just log a warning here …
14:41 Changeset in xpra [9760] by antoine
fixup r9751: the new code uses sys
14:26 Changeset in xpra [9759] by antoine
fixup the fixup: field is global to server
14:22 Changeset in xpra [9758] by antoine
once again, patch made a mess - fixup for r9677
14:14 Changeset in xpra [9757] by antoine
add latest changes to release notes
12:36 Versions/PendingFixes edited by antoine
(diff)
12:33 Changeset in xpra [9756] by antoine
r9682 for v0.15.x branch: use slow and safe approach with non-opengl …
12:32 Changeset in xpra [9755] by antoine
r9694 for v0.15.x branch: fix config file defaults and location on win32
12:28 Changeset in xpra [9754] by antoine
r9711 for v0.15.x branch: we don't need the mainloop to still be …
12:25 Changeset in xpra [9753] by antoine
r9715 for v0.14.x and v0.15.x branches: honour the decorations bit flag
12:22 Changeset in xpra [9752] by antoine
r9727 for v0.15.x branch: avoid packet encoder errors if pipeline …
12:21 Changeset in xpra [9751] by antoine
r9737 + r9738 for v0.15.x branch: skip max-size hook with GTK3, run …
12:17 Changeset in xpra [9750] by antoine
r9743 for v0.14.x and v0.15.x branches: avoid invalid warning
12:16 Changeset in xpra [9749] by antoine
r9728 for v0.15.x branch: fix size-hints
12:14 Changeset in xpra [9748] by antoine
r9725 for v0.15.x branch: acquire GIL before running python code
10:51 Ticket #899 (generic shadow improvements) created by antoine
Tracker ticket. See also: * #389: win32 * #558: win32 nvenc * #390: …
10:03 Versions/PendingFixes edited by antoine
(diff)
10:01 WikiStart edited by antoine
(diff)
10:00 Versions/PendingFixes edited by antoine
(diff)
09:56 Versions edited by antoine
(diff)
09:55 Versions edited by antoine
(diff)
09:54 Versions edited by antoine
(diff)
09:54 Versions/PendingFixes created by antoine
Note: See TracTimeline for information about the timeline view.