Allocation of 810 FState objects (32400 bytes in total) could be performed in MSVC x64 build but worked in 32-bit version
This fixes https://mantis.zdoom.org/view.php?id=602
- added arrays to the config to hold entries for the softsynths' config files. This is not active yet, but will later be used to give the user a list of config options instead of having to type it by hand.
The needed headers are now included in the repo, which for these libraries is possible thanks to a stable ABI (at least on Windows, the other platforms still need to be checked but the headers only add, never remove or change existing content.)
The big advantage of this setup is that it allows building the project on Windows without any necessary setup - all that needs to be provided is the DLLs from the binary package.
This still requires some fixes for macOS and Linux. On MacOS the proper library names are missing and the ones for Linux are not verified. Both platforms should work, though, if the dynamic loading is disabled.
For some files that had the Doom Source license attached but saw heavy external contributions over the years I added a special note to license all original ZDoom code under BSD.
src/sound/oplsynth/musicblock.cpp:3:10: fatal error: 'muslib.h' file not found
src/sound/oplsynth/oplio.cpp:410:12: error: use of undeclared identifier 'cos'
src/sound/oplsynth/oplio.cpp:410:41: error: use of undeclared identifier 'sin'
This was very poorly done without ever addressing the issues a composite render style can bring, it merely dealt with the known legacy render styles.
The same, identical code was also present in two different places.
The oversight that AlterWeaponSprite overrode even forced styles was also fixed.
OpenGL is not implemented yet but with the problems eliminated should be doable now.
This is to ensure that the Class pointer can be set right on creation. ZDoom had always depended on handling this lazily which poses some problems for the VM.
So now there is a variadic Create<classtype> function taking care of that, but to ensure that it gets used, direct access to the new operator has been blocked.
This also neccessitated making DArgs a regular object because they get created before the type system is up. Since the few uses of DArgs are easily controllable this wasn't a big issue.
- did a bit of optimization on the bots' decision making whether to pick up a health item or not.
The global variable holding a pointer to this thinker should be a weak reference to the instance in the thinker chain, there is no need to mark this global variable, as the thinker's lifetime is only determined by the thinker chain.
- committed a few Posix related file the last commit missed.
I have no idea why they were even in there, as they intentionally circumvented all GC related features - they declared themselves fixed if prone to getting collected, they all used OF_YesReallyDelete when destroying themselves and they never used any of the object creation or RTTI features, aside from a single assert in V_Init2.
Essentially they were a drag on the system and OF_YesReallyDelete was effectively added just to deal with the canvases which were DObjects but not supposed to behave like them in the first place.
Let's use inline checkers in PType instead of constantly having to do clumsy IsKindOf checks etc. Once complete this also means that the types can be taken out of the class hierarchy, freeing up some common names.
src/p_pspr.cpp:363:37: warning: more '%' conversions than data arguments [-Wformat]
src/gl/textures/gl_texture.cpp:845:21: warning: comparison of integers of different signs: 'int' and 'unsigned int' [-Wsign-compare]
Fixed bunch of compilation errors:
cannot pass non-trivial object of type 'FString' to variadic method; expected type from format string was 'char *' [-Wnon-pod-varargs]
Fixed linker erorr:
g_doomedmap.cpp.o: In function `InitActorNumsFromMapinfo()':
src/g_doomedmap.cpp: undefined reference to `PClass::FindActor(FName)'
Combining these two groups of data has been the cause of many hard to detect errors because it allowed liberal casting between types that are used for completely different things.
Having these two types related can cause problems with detection in the compiler because for some parts they need quite different handling.
Common handling for the fields has been moved into PSymbolTable but overall redundancy was quite minor as both types share surprisingly little functionality.
A pointer to an object has quite different semantics so let's do this with proper virtual inheritance. This should allow to simplify a lot of pointer related checks in the compiler.
- fixed: IsVisibleToPlayer and optimized it a bit more.
- moved SourceLumpName to PClass, so that it can also be used for non-actors (there's a lot of non-Actor classes already.)
- separated the serializer for PClassPointer from PPointer. Even though not doable, a pointer to a class type is something entirely different than a class pointer with a restriction so each should handle its own case.
- added a few access functions for FActorInfo variables.
With PClassActor now empty the class descriptors can finally be converted back to static data outside the class hierarchy, like they were before the scripting merge, and untangle the game data from VM internals.
For these fields maps have no advantage. Linearly searching a small array with up to 10 entries is nearly always faster than generating a hash for finding the entry in the map.
- some optimization of access to OwnedStates in old DECORATE.
- consolidate all places that print a state name into a subfunction.
- allocate states from the ClassDataAllocator memory arena. States do not need to be freed separately from the rest of the static class data.
This reinstates the old FActorInfo as part of the meta data a class can have so that the class descriptor itself can be freed from any data not directly relevant for managing the class's type information.
This is not needed anymore because classes do not need to be replaced. The only reason this was implemented was the original design with the class descriptors taking on all the metadata themselves.
This is an incredibly costly way to do a debug check as it infests the entire VM design from top to bottom. These tags are basically useless for anything else but validating object pointers being passed to native functions (i.e. mismatches between definition and declaration) and that simply does not justify a feature that costs execution time in non-debug builds and added memory overhead everywhere.
Note that this commit does not remove the tags, it only discontinues their use.
See https://forum.drdteam.org/viewtopic.php?t=7588
Processing order is now the same as in Chocolate Doom
prBoom+ loads separate files after all WAD lumps though
This makes sense but would change loading sequence existed in ZDoom for years
This is somewhat brute-force thanks to the surprising lack of good documentation for the Ogg headers. The only other option would have been some rather bloated library for a function that should be 25-30 lines at most.
The idea is to have more control on the game side instead of dealing with these formats in the backend, which was done for FMod because it already had the decoders implemented.
However, with OpenAL this setup makes no sense and only complicates future extensions that can be better handled at a higher level.
No more error when running with +map command line parameter with classic HUD:
> VM execution aborted: Attempt to draw to screen outside a draw function
> Called from BaseStatusBar.DrawImage [Native]
> Called from DoomStatusBar.DrawFullScreenStuff at gzdoom.pk3:zscript/statusbar/doom_sbar.txt, line 140
> Called from DoomStatusBar.Draw at gzdoom.pk3:zscript/statusbar/doom_sbar.txt, line 41
- all 5 settings affected by uiscale have been changed to have the exact same semantics: -1, if supported means special scaling, this is available for HUD and status bar, 0 means to use uiscale, any larger value is a direct scaling factor.
- scaling is cut off when the factor is larger than screenwidth/320 or screenheight/200 because anything larger will definitely not fit.
- a lot of code has been cleaned up and consolidated. Especially the message code had an incredible amount of redundancy.
- all scaling options have been moved into a submenu. This menu is not complete, though - it still requires a special menu widget to convey the intended information without confusing the user.
- activated the RenderOverlay event, now that it can be called from the correct spot, i.e. right after the top level HUD messages are drawn. The system's status output will still be drawn on top of them.
- changed the effect spawn prevention of the Hexen flame strike weapon and reverted the attempt to fix this in FastProjectile.
This cannot be fixed in the base class, which was doing everything right. It's the flame missile that was doing undefined things by stopping its movement without clearing its missile flag. This cannot work because missiles are given some minimal forced velocity to ensure collision detection and any attempt to address this without clearing the missile flag is doomed to fail.
This allows using the UI scale or its own value, like all other scaling values.
In addition there is a choice between preserving equal pixel size or aspect ratio because the squashed non-corrected versions tend to look odd, but since proper scaling requires ununiform pixel sizes it is an option.
- changed how status bar sizes are being handled.
This has to recalculate all scaling and positioning factors, which can cause problems if the drawer leaves with some temporary values that do not reflect the status bar as a whole.
Changed it so that the status bar stores the base values and restores them after drawing is complete.
This would cut off overlong names and the handling for status bars with protruding elements was far too simplistic and worse, making assumptions based on game mode.
It now uses a virtual function to query the status bar itself for returning this information so it can be overridden and uses V_BreakLines to split the text if it is wider than the display.
This is necessary because the hardware accelerated renderers will hide the problem, but with pure software rendering to a locked hardware surface, like DirectDraw can result in a crash.
Note that ANY mod that gets caught in this did something wrong!
Currently this is only being used for draw operations that are not automap related, i.e. DrawLine, DrawPixel and FillSimplePoly are not subjected to it.
All our continuous integration targets have no problems with C99 isnan() macro but on Ubuntu 16.04 compilation fails
It appeared that some implementation of C++ Standard Library may undefine bunch of C macros to avoid conflicts with own declarations
- finished work on the Doom status bar. I also took the opportunity to fix the layout of the inventory bar which is a bit broken in SBARINFO.
- tuned the selection rules for deciding what creates the status bar, so that the most recent definition that can be found is chosen.
- got rid of the image list in the Doom status bar. The cost of the texture lookup is mostly irrelevant here so clearer and shorter code is preferrable.
- moved the box fitting code from DrawTexture into the native function to have all coordinate calculations in one place which is necessary to implement proper alignment default handling. Without higher level functions altering positioning the default can be set to automatic alignment determination, i.e. the value's sign decides where something is placed. Of course for special cases this can be overridden.
- use ANIMDEFS to animate the inventory arrow,
Fullscreen HUD done with the exception of key and inventory bar. I also used the opportunity to make it a bit more resistant against badly designed inventory icons.
- better handling of ForceScale for the fullscreen HUD that doesn't mess around with CVARs.
- moved the mug shot into the status bar, because this is global state that needs to be shared between different pieces of code which want to display a mug shot.
- SBARINFO should work off the current status bar settings instead of the ones stored in its script object
Note that there is no direct access, all this exposes is a single function to get the current face's texture which then can be drawn using the existing functions.
Note that the pop screens are special because they are not subject to scaling - they will always be drawn with the current resolutions clean scale. As a result they cannot use the HUD drawers but instead continue to use the low level draw functions directly.
- replaced Inventory.DrawPowerup with a GetPowerupIcon method so that the calling code can handle the drawing and apply its own rules. This was a major design flaw of allowing the inventory items to handle the drawing themselves, because they were unable to adjust to different HUD frontends. Note that any mod that overrides DrawPowerup will not draw any icon that expects to be handled that way!
- the alternative HUD now has its own, separate drawer that obeys the AltHUD's rules, and not the ones of the normal fullscreen HUD.
- the standard drawer has been scriptified as a virtual function.
- Both drawers now handle positioning of the icon inside its assigned box themselves instead of trusting the powerup item to do it correctly.
- DTA_HUDRules and Screen.DrawHUDTexture are to be considered deprecated because both do not integrate into the redesigned HUD code.
- merged BaseStatusBar and CustomStatusBar back together.
Since the low level draw functions are better done in native code for both performance and debuggability the split has become pointless.
- decided to ditch the widget system I had started to lay out. As it turns out that would make things far more complicated and slower than they need to be.
