aboutsummaryrefslogtreecommitdiff
path: root/nx-X11/programs/Xserver
Commit message (Collapse)AuthorAgeFilesLines
* Client.c: save some linesUlrich Sibiller2020-01-051-43/+28
|
* GC.c: condense codeUlrich Sibiller2020-01-051-34/+2
| | | | remove some empty lines to make the code more compact
* GC.c: reformat commentsUlrich Sibiller2020-01-051-53/+39
| | | | save some lines
* GC.c: fix format specifiers and parametersUlrich Sibiller2020-01-051-2/+2
|
* GCOps.c: fix format length specifiersUlrich Sibiller2020-01-051-2/+2
|
* GCOps.c: reformat commentsUlrich Sibiller2020-01-051-114/+84
| | | | save some lines
* Display.c: shorten codeUlrich Sibiller2020-01-051-58/+32
| | | | scope improvements, designated initializers
* Binder.c: scope improvementsUlrich Sibiller2020-01-051-8/+3
|
* Image.c: scope improvementsUlrich Sibiller2020-01-051-87/+37
|
* Error.c: scope improvementsUlrich Sibiller2020-01-051-22/+11
|
* Cursor.c: scope improvementsUlrich Sibiller2020-01-051-69/+60
| | | | + use of designated initializers
* GC.c: scope improvementsUlrich Sibiller2020-01-051-52/+24
|
* GCOps.c: scope improvementsUlrich Sibiller2020-01-051-64/+43
|
* NXrender.c: call upstream versions of (S)ProcRenderDispatchUlrich Sibiller2020-01-052-35/+41
|
* NXrender.c: modify code to match upstream as close as possibleUlrich Sibiller2020-01-051-1/+3
|
* NXrender.c: mark NX changesUlrich Sibiller2020-01-051-25/+96
|
* NXrender.c, render.c: fix memory handling bugsUlrich Sibiller2020-01-052-8/+27
|
* nxagent: remove additional keyboard handling from Reconnect.cUlrich Sibiller2020-01-052-20/+32
| | | | | | | | At reconnect we check if a keyboard reset is required. If so, we are calling the corresponding code from Keyboard.c. No explicit keyboard code in Reconnect.c required. fixes ArcticaProject/nx-libs#886
* nxagent: central check for keyboard clone modeUlrich Sibiller2020-01-053-18/+14
|
* Args.c: more debug outputUlrich Sibiller2020-01-051-0/+3
|
* Args.c: do not parse nx/nx marker as optionUlrich Sibiller2020-01-051-9/+32
| | | | slight optimization
* Args.c: fix: do not modify options stringUlrich Sibiller2020-01-051-2/+9
| | | | | This resulted in a garbled option string on reconnect where everything after the first '=' was lost!
* Keyboard.c: scope cleanupsUlrich Sibiller2020-01-051-76/+54
|
* Keyboard.c: Use __func__ at fprintf callsUlrich Sibiller2020-01-051-51/+50
|
* Keyboard.c: use "\0" instead of "0"Ulrich Sibiller2020-01-052-5/+5
|
* Keyboard.c: fix: handle empty stringUlrich Sibiller2020-01-051-0/+6
|
* Display.c: drop icon file supportUlrich Sibiller2020-01-052-153/+17
| | | | | | | | | | | | | | | | | | | The normal case was to use the builtin icons. There were two cases where the icon file was read from disk: 1) /usr/NX/share/images/nxagent.xpm was existing which normally is not there 2) case 2 was not true and nxagent.xpm was existing somewhere in the PATH (!) (replace nxagent.xpm by x2goagent.xpm if in x2gp mode) Scanning the path from the PATH variable for xpm files is kind of unexpected and dangerous, too (think of automounter triggers or invalid xpm files). Also remove the xpm files from the distribution.
* nxagent.xpm: reduce colours from 242 to 8Ulrich Sibiller2020-01-051-371/+137
| | | | | | | | | This leads to a tremendous startup speedup (several seconds, depending on latency) because libXpm had to do a lot of X requests (have not further examined which). With this colour-reduced version things are quick again for those connections without losing a lot of quality. Fixes ArcticaProject/nx-libs#824
* x2goagent.xpm: move copyright to the end to make it a valid xpm fileUlrich Sibiller2020-01-051-16/+15
| | | | gimp refused to load it, nxagent crashed if this was loaded as icon file!
* nxagent manpage: improve descriptions and typesUlrich Sibiller2019-11-241-25/+42
| | | | | Most <int> options are <bool>. Change that and add missing defaults. Also add list of descriptions for the defer option.
* nxagent manpage: extend clipboard documentationUlrich Sibiller2019-11-241-3/+15
|
* Clipboard.c: report more supported targetsUlrich Sibiller2019-11-151-3/+2
| | | | | fix one FIXME. This is a rewritten and extended version of a quick patch by Danil Pleshakov and Dimbor.
* Clipboard: fix debugging outputUlrich Sibiller2019-11-151-4/+10
| | | | Correctly determine the Atom names from the real X server.
* Clipboard.c: automatically calc the array sizeUlrich Sibiller2019-11-151-2/+2
|
* Clipboard.c: improve XChangeProperty codeUlrich Sibiller2019-11-151-2/+2
| | | | | | | | | | The previously used "Atom" type was confusing. Clarify this by writing the code as requested in the manpage of XChangeProperty: "If the specified format is 8, the property data must be a char array. If the specified format is 16, the property data must be a short array. If the specified format is 32, the property data must be a long array."
* Rootless.c: rewrite loop to allow for scope improvementsUlrich Sibiller2019-11-151-17/+10
| | | | | make i a local loop variable everywhere by changing the while loop to a for loop.
* Rootless.c: fix linefeeds when printing listsUlrich Sibiller2019-11-151-5/+5
|
* Rootless.c: silence compiler warning about wrong parmeter sizesUlrich Sibiller2019-11-151-5/+3
|
* Rootless.c: disable unused functionUlrich Sibiller2019-11-151-0/+2
|
* Rootless.c: silence compiler warnings about wrong fromat specifiersUlrich Sibiller2019-11-151-15/+15
|
* Rootless.c: Fatal error if malloc failsUlrich Sibiller2019-11-151-1/+5
|
* Rootless.c: cleanup function prototypesUlrich Sibiller2019-11-151-10/+8
|
* Rootless.c: reformat commentsUlrich Sibiller2019-11-151-34/+26
|
* Rootless.c: scope improvementsUlrich Sibiller2019-11-151-58/+38
|
* Rootless.c: use __func__ instead of hardcoded function nameUlrich Sibiller2019-11-151-78/+74
|
* Rootless.c: use designated initializersUlrich Sibiller2019-11-151-15/+9
| | | | This fixes a possible unitialized struct variable.
* Pixmap.c: reformat commentsUlrich Sibiller2019-11-151-56/+44
| | | | plus some minor formatting fixes
* Pixmap.c: scope improvementsUlrich Sibiller2019-11-151-125/+86
|
* Pixmap.c: drop code that does not make senseUlrich Sibiller2019-11-151-10/+2
| | | | | | This check has already been done before we reach here. Also add some comments about a condition that will never occur...
* Pixmap.c: call miModifyPixmapHeader twice instead of duplicating codeUlrich Sibiller2019-11-151-83/+11
|