~ctwm/ctwm/trunk : changes

~ctwm/ctwm/trunk » Changes to add_window.c from revision 719

From Revision 719 to 564

Rev	Summary	Authors	Date
708	Bypass BorderedLayout for strut'd windows. Bypass BorderedLayout for strut'd windows. This is a hack, and does the wrong thing. However, due to the way we currently handle struts, the previous behavior would force a window to move outside its own strut, which is nonsense. A proper implementation would require evaluating struts individually, not cramming them into BorderedLayout (which already means something similar but distinct), and encapsulating those checks in all the places that do window positioning. That's a major change though; this at least avoids the extra-dumb behavior, and works no worse than 4.0.3 and pre-r644 code for those corner cases.	Matthew Fuller	2 years ago
707	Fixup window placement with DontMoveOff + 3d borders. Fixup window placement with DontMoveOff + 3d borders.	Maxime Soulé	2 years ago
703	Make session support built-time conditional. Remains enable... Make session support built-time conditional. Remains enabled by default for now. This allows disabling the X session management protocol support code. This is pretty pointless for the "restart-ctwm" case, since all it's storing is the info the windows and their properties would already have in the first place; it's really only meaningful if you actually use xsm or the like and save and restore full sessions with multiple running clients and their after a system/X restart. It's not clear this was ever a particularly widely used or supported feature across X clients in general, particularly in places where people are using WM's like ctwm as opposed to integrated DE's like Gnome and KDE. It's also a little grody, and for extra points is under slightly different license terms than the rest of the codebase. So, it's possible we may win from retiring it.	Matthew Fuller	2 years ago
702	Hide WindowBox support behind a build-time option, disabled ... Hide WindowBox support behind a build-time option, disabled by default, to be reaped. Some code nearby a few of these ifdefs can probably be slightly simplified once they're gone, but trying to do it conditionally would be too messy to bother with.	Matthew Fuller	2 years ago
697	Add a build option to disable VirtualScreens support. This ... Add a build option to disable VirtualScreens support. This is not documented anywhere, and is intended as a short-term tool for deorbiting the "feature". Unlike the similar handling of captive mode, there's no attempt to put the entire feature/removal behind the #ifdef, if only because it's _so_ deeply entwined in how _everything_ works. Much more will need to be done to actually remove all the tendrils of it. But this does remove 2/4 of the vscreen related fields from the ScreenInfo struct, and eliminates all handling of the config param. So, it's a start...	Matthew Fuller	2 years ago
693	Add a build option to en/disable captive mode. This is inte... Add a build option to en/disable captive mode. This is intentionally not documented, and is intended to be a short-term tool to cleanly remove the feature.	Matthew Fuller	2 years ago
689	Do various non-functional cleanups. Do various non-functional cleanups. Primarily this clears out all the remaining things clang-analyzer bitches about (apart from stuff in lex generated code, which ain't my problem). These basically consist of: - Dead store warnings. Some of these are obviously eliminatable; others are _correct_, but have instead been silenced because they're intentionally dead stores for defensive reasons. - Null pointer derefs. Some of these I'm pretty sure are impossible due to whole-program invariants, others are probably impossible due to how X works. I'm fairly certain none of them indicate any real problems; if they did, they'd be manifesting as segfaults, and we'd be hearing all about that. So this is mostly just about clearing out the list as far as possible to make it easier to find any meaningful ones that show up in the future. Also tweak cmake config a bit to get useful compile_commands.json back for dev use.	Matthew Fuller	3 years ago
683	Use SizeFont's height instead of the text's height for figur... Use SizeFont's height instead of the text's height for figuring the size of the window for RandomPlacement=off of windows. This seems to be a slightly less bad guess in practice, ignoring all the other b0rkedness in how this window is used in its various guises.	Michael van Elst	4 years ago
679	Some cleanups mostly about the window ring. Some cleanups mostly about the window ring.	Olaf 'Rhialto' Seibe...	4 years ago
665	Merge support for EWMH _NET_FRAME_EXTENTS from Maxime Soulé,... Merge support for EWMH _NET_FRAME_EXTENTS from Maxime Soulé, with some cleanups. This provides clients with information about how much space our decorations around the window are using up, giving them an additional source of information about where on the screen a given point inside the window winds up.	Matthew Fuller	5 years ago
644	Merge RANDR and Layout support. Merge RANDR and Layout support. This adjusts internal data structures defining our display layout to be more flexible. When build with RANDR support (default) and the X server supports the extension, the structures are built with knowledge of the invididual monitors composing the desktop and how they're laid out. The various zoom'ing functions now zoom just on the monitor they're on; .x*zoom functions will zoom across the entire Screen. This makes multi-monitor setups work much more fluidly, especially when they're of differing resolutions or otherwise don't add up to an even rectangular output.	Matthew Fuller	6 years ago
642	Comment, refactor, and lightly rework the startup process, p... Comment, refactor, and lightly rework the startup process, particularly the per-Screen handling loop. More initialization is moved into the Init function which now also handles allocation etc. Other initialization steps are better documented, including their cross-relationships and ordering requirements. Variables are moved to smaller and in some cases artificial scopes for readability. No visible changes expected.	Matthew Fuller	6 years ago
626	Merge fixes for handling of WM_HINTS changing after a window... Merge fixes for handling of WM_HINTS changing after a window was created. We now properly apply our various overrides then, as well as during the initial AddWindow().	Matthew Fuller	6 years ago
616	Merge support for EWMH _NET_WM{_ICON,}_NAME, and our own ove... Merge support for EWMH _NET_WM{_ICON,}_NAME, and our own overriding CTWM_WM{_ICON,}_NAME.	Matthew Fuller	6 years ago
615	Doxygen-ify struct TwmWindow. Many existing comments were m... Doxygen-ify struct TwmWindow. Many existing comments were minorly tweaked into doxygen format, some were significantly reworked, and others summoned from scratch. This also includes pulling it out into its own #include (still pulled in via the main ctwm.h, but split out for dev simplicity), and minor wording tweak to some comments that weren't worth breaking out on their own.	Matthew Fuller	6 years ago
603	Add DontNameDecorations config var to disable the XStoreName... Add DontNameDecorations config var to disable the XStoreName()'s on our wrapper/decoration windows.	Matthew Fuller	7 years ago
579	Land some fixups of OTP and EWMH interaction. Land some fixups of OTP and EWMH interaction. This involves restructuring the OTP calculations to take some EWMH information as a default if the user doesn't have anything (e.g., from the config file) to override it, and allowing that and the various OTP f.whatever functions to set a base priority. Then various EWMH related actions can also set affect flags which add offsets from that base priority to determine the effective priority. Those calculations can then be run at any time to determine where the window is actually placed.	Matthew Fuller	8 years ago
569	Stop explicitly overriding focus so we always give windows f... Stop explicitly overriding focus so we always give windows focus, and add ForceFocus config param to enable it if/when somebody needs it. This seems to fix behavior in some cases where windows explicitly tell us not to give them focus because they handle their input differently, and our doing it anyway breaks them.	Matthew Fuller	8 years ago
566	Unwind the mess of occasional pointer-sharing between the wi... Unwind the mess of occasional pointer-sharing between the window name and icon_name; just let them always be separate. With the removal of full_name, this lets us remove the somewhat twisty function that tried to free some subset of the group, and just free each piece when it makes sense. This will probably lead to dozens of bytes more memory usage for the situation where you have multiple windows that don't set an icon name, but do set a window name. For more straightforward code, that's a ridiculously easy tradeoff.	Matthew Fuller	8 years ago
564	Remove TwmWindow.full_name usage completely; it was nowhere ... Remove TwmWindow.full_name usage completely; it was nowhere anything but a duplicate pointer of TwmWindow.name, so it doesn't add anything.	Matthew Fuller	8 years ago

Older »