ed9b7b3eb1
Lua address several memory leaks - when a script stops, take any registered functions and spawn new threads for them to execute, and regularly dispose them. Also, the main lua thread needs to have scheduled disposes run on it, even if there is no active scripts (it was being run as part of disposing individual scripts, needs refactoring). All of this fixes #1920
2020-07-12 15:48:50 -05:00
9278c526d0
fix a few compiler warnings
2020-07-12 16:02:26 -04:00
1e3324cfab
Registered Lua Functions dialog - update when a lua script unregisters a file, make dupating more robust to catch other possible scenarios where it failed to update
2020-07-12 13:19:48 -05:00
c4bcb2451a
remove now unused class in RomLoader.cs
2020-07-12 13:03:32 -05:00
05709628e8
remove a jillion unused usings in RomLoader.cs
2020-07-12 13:02:12 -05:00
40ad7ed67e
Lua - account for callbacks with no parent script creating more callbacks by generating a mock lua file with a new thread, and register the thread with the sandbox. Fixes #1983 . Not super proud of this.
2020-07-12 12:58:43 -05:00
021b1639c2
Remove special code for UnsupportedGameException in RomLoader
...
This shouldn't be needed anymore
2020-07-12 13:48:46 -04:00
78bbc75f33
multi rom loading by xml all uses core inventory now
2020-07-12 13:45:33 -04:00
8d56b65734
Convert all single-CD loading to use CoreInventory
2020-07-12 13:21:10 -04:00
ab07e67a07
PCEHawk: Use CoreLoadParameters for cd mode
2020-07-12 13:06:15 -04:00
744a9a97d0
Remove custom loader code for ChannelF
2020-07-12 12:54:36 -04:00
5c5ffed5ff
Next piece of the puzzle
...
This probably breaks something. I am sorry, grab me and I'll fix it
2020-07-12 12:52:27 -04:00
33a0954087
lua console - fix recent scripts not updating when creating a new script, opening a script, or duplicating a script. Might be other scenarios that are currently missing (regression from some refactors about a month ago)
2020-07-12 11:12:29 -05:00
4ad89667d6
This will get simpler I promise
2020-07-12 11:37:06 -04:00
6f218ff20b
Remove a benign null reference exception
...
This was in a try catch so you got the null you needed anyway, but it could be confusing to see in a debugger report
2020-07-12 11:10:57 -04:00
06fa96e335
O2: finish g7400 graphics set
2020-07-12 10:44:11 -04:00
81d0b4ec9e
Revert "Replace dynamic type w/ casts in CoreInventory/CLP"
...
This reverts commit 9ee4821148
2020-07-12 10:18:08 -04:00
f41a70ef32
swish swish
2020-07-12 09:06:20 -04:00
f5efbff1d3
O2: more g7400 work
2020-07-12 09:03:28 -04:00
9072614dfb
romloader cleanup
2020-07-12 08:09:50 -04:00
dc8615ca06
fix compile, sorry sorry
2020-07-12 07:36:06 -04:00
80793836d3
Use CoreLoadParameters on dual gambatte
2020-07-12 07:28:06 -04:00
9d9dd8a0fd
More romloader stuff
...
It'll all come together eventually
2020-07-12 07:08:52 -04:00
9ee4821148
Replace dynamic type w/ casts in CoreInventory/CLP
...
dynamic introduced in cc9d7df9f
2020-07-12 17:30:10 +10:00
c712bde84e
remove firmware manager dependency from RomLoader, and a few other cleanups
2020-07-11 20:10:23 -05:00
467f0aff2b
Refactor 32x loading
...
less code is good good
2020-07-11 20:38:09 -04:00
502aef73a0
Clean up ti 83 loading. Probably works, but I have nothing to test against it
2020-07-11 20:31:36 -04:00
8900618ce1
respect the Movie end type when loading .tasproj files outside of tastudio
2020-07-11 18:52:52 -05:00
1f39160999
give up on the idea of throwing an exception when attaching a core to a movie. The reason this doesn't work is that restart movie keeps the movie in memory (which was previously attached). This commit decrees that it is safe to keep a movie in memory, and attach to a new instance of the same core. The alternative is that movies would have to save and load on restart which is slow and may not be desired
2020-07-11 18:03:53 -05:00
86ac6eadcc
Romloader omega refactor: Use CoreLoadParameters for PSX
2020-07-11 18:47:56 -04:00
6657917281
Fix parse error loading CCD files
2020-07-11 18:47:56 -04:00
9d054d19f6
bk2 and tasproj movies - respect the preload flag, only loads header and inputlog. This speeds up the play movie dialog by 50-100%, a lot more if there is a large sizeable .tasproj files in the folder
2020-07-11 17:06:40 -05:00
a19791a212
Fm2Import - warn about ignoring the famicom port, instead of a full error, apparently movies can have this port set, but to none, so importing the movie can still be successful, for example: http://tasvideos.org/6739S.html , fixes #1989
2020-07-11 15:57:08 -05:00
7e2251b795
Fix the CoreLoadParameters <-> CoreInventory interop
2020-07-11 16:46:16 -04:00
dbb73805db
error message cleanup
...
Once everything is unified, we'll probably hit this path once in a while
2020-07-11 16:46:16 -04:00
cc9d7df9f7
WIP of some romloader cleanups
...
This one commit doesn't make things cleaner by itself, no. The rough idea is that eventually (R) we'll pass all cores that currently use mangled arse-custom constructors with custom code all o'er the place a single, awesome, CoreLoadParameters object.
Then the romloader can be changed to just synthesize this object on one common codepath, some bs for core preferences, and boom everything is perfect. Sort of.
2020-07-11 16:46:16 -04:00
3a820302e3
improve name of a method
2020-07-11 15:42:47 -05:00
e5dc473203
inputroll - make escape clear selected items, this was being done custom in ram search but makes sense to be a consistent feature. It's possible people will be upset with this in tastudio since this is new functionality (and not in taseditor either)
2020-07-11 15:41:45 -05:00
d44e96f535
remove some redundant code
2020-07-11 15:35:50 -05:00
aeaf6aa282
add some KeyCode extension methods to make KeyUp/Down event methods easier to read
2020-07-11 15:35:00 -05:00
e29d024a90
O2: graphics work
2020-07-11 16:31:42 -04:00
be93903c12
Play Movie dialog - properly show tooltips for detail values, also show expected vs actual (only works on the key column though, because winforms is poo), fixes #1888
2020-07-11 14:53:51 -05:00
196934bd56
Tastudio - fix TasSession loading, fixes #2063
2020-07-11 14:28:21 -05:00
cb575d36b4
add an option to InputRoll for "Selection when paging", as a bit of a hack to revert Tastudio back to its old PgUp/Dn behavior, while still keeping the current behavior for other input roll usages (as it makes more sense for traditional listview behavior). Fixes #2074
2020-07-11 13:49:00 -05:00
4343414937
Enable hypernyma for uhh, the thing. Yeah, CD games, that thing.
...
I had tested this locally some during original development, but turned it off at some point and forgot to turn it back on.
2020-07-11 13:59:05 -04:00
bd413738a1
Fix nyma core bug with layer settings
...
If you tried to load a nyma core with any of your layers disabled, you would crash
2020-07-11 13:56:43 -04:00
0578443338
Slight modification to previous commit: Keep config forward compatibility in this case.
...
That way, if somedev tries to bisect something, they don't get pooped on by their config.
2020-07-11 13:47:28 -04:00
bcbf124f21
Don't use fragile fully qualified typenames to resolve setting and sync setting types in config files.
...
Instead, the frontend must know what the type will be ahead of time.
TODO: Apply the same fix to mooovies
2020-07-11 13:40:25 -04:00
a28ca37cfb
Fix issue with settings adapter ( #2223 )
...
This would fail to work on any core that used the serviceprovider to provide a settings definition that wasn't on the iemulator itself? Guess we never did that
2020-07-11 13:35:09 -04:00
9820fb0879
lua - implement tastudio.submitclearframes(), fixes #2170
2020-07-11 12:19:27 -05:00
adelikat