Page MenuHomeWildfire Games
Feed Advanced Search

Today

elexis added a comment to D2084: Remove CMapGeneratorWorker struct AutoFree?.

For a C++ exception in Run() the following happens (with the currently committed code without this patch):

Tue, Jul 16, 9:17 PM
elexis added a comment to D1042: rename maps/ file names to use underscores (instead of spaces and capitalization).

1 what to do with the scenarios/Height Map *.png files

Tue, Jul 16, 8:56 PM
elexis added a comment to D2084: Remove CMapGeneratorWorker struct AutoFree?.

If a mapgen script throws an error, or if the user hits alt+f4 (SQL quit event), then LoadGlobalScriptFile (it's JS::Evaluate call`) and subsequently Run will return false.

Tue, Jul 16, 8:45 PM
elexis added a comment to D1904: Move fauna_arctic_wolf_attack.xml from trigger/ to gaia/.

Or fauna_arctic_wolf_domestic.xml?

Tue, Jul 16, 8:15 PM
elexis added a comment to rP22485: [Windows] Automated build..

Acknowledged!

Tue, Jul 16, 6:05 PM
elexis added a comment to D2083: territory pull again.

I tested the existing and newly proposed entity a bit.

Tue, Jul 16, 5:52 PM
elexis created D2085: Support RandomMap tests.
Tue, Jul 16, 5:10 PM
elexis updated the test plan for D2084: Remove CMapGeneratorWorker struct AutoFree?.
Tue, Jul 16, 4:50 PM
elexis created D2084: Remove CMapGeneratorWorker struct AutoFree?.
Tue, Jul 16, 4:48 PM
elexis added a comment to rP15569: Fixes small memory leak introduced in r15568..

From 2014-07-26-QuakeNet-#0ad-dev.log:

22:36 <@leper> CMapGeneratorWorker::Run() seems to leak stuff
22:37 < Yves`> leak? memory?
22:38 <@leper> the SAFE_DELETE is only reached when all checks succeed
22:39 < Yves`> hmm yes, looks like that
22:44 < Yves`> look like I can just remove the outer scope braces and have to keep the rest
22:45 <@leper> you could change the return false to goto err or add a struct that takes a pointer and frees that in the dtor
Tue, Jul 16, 3:24 PM
elexis added inline comments to D2080: Remove pointless Eval calls creating empty objects and arrays.
Tue, Jul 16, 11:13 AM
elexis planned changes to D2080: Remove pointless Eval calls creating empty objects and arrays.

As mentioned by Vladislav JS_NewPlainObject will not return a pointer in case of out-of-memory, then it will try to null-deref.
JS::ObjectOrNullValue can fix that up to the point of not adding a regression.
But the error should be reported rather, probably with some throw rather than some logerror or silent return.
That OOM check is already broken in the codebase it seems.

Tue, Jul 16, 12:37 AM

Yesterday

elexis accepted D2082: Fix undefined behaviour introduced in rP22473.

After applying this patch, my visual replay and nonvisual replay mode hashes match again.

Mon, Jul 15, 7:53 PM
elexis added a comment to D2070: Linting: Remove controversial ESLint style rules.

One of the ideas of the linting bot rules was to incentivize changes towards more standardized language use.

Mon, Jul 15, 5:37 PM
elexis updated the summary of D2070: Linting: Remove controversial ESLint style rules.
Mon, Jul 15, 5:34 PM
elexis committed rP22481: Provide microsecond precision in the RandomMapLogger too, refs rP19603..
Provide microsecond precision in the RandomMapLogger too, refs rP19603.
Mon, Jul 15, 4:45 PM
elexis updated the test plan for D2080: Remove pointless Eval calls creating empty objects and arrays.
Mon, Jul 15, 4:27 PM
elexis created D2080: Remove pointless Eval calls creating empty objects and arrays.
Mon, Jul 15, 4:23 PM
elexis added a comment to D1493: Camera view config options.

Therefore you either have to mention that it requires a restart, like we do with the other settings, or actually make them take effect.

The latter would be much better, especially to test out which values work best for one, adapting progressively. (It makes testing this patch much harder even, as one has to restart the game between value changes.)
I suppose it can be done in a separate diff if the c++ code is uncertain territory.

Mon, Jul 15, 2:15 PM
elexis updated the diff for D1754: Pass an argument to GUI events / SendEventToAll, remove loading screen progess workaround.

Pass JS::HandleValueArray instead of a JS::Value so that the callee can support multiple arguments.
The creation of that JS::HandleValueArray might be further abstracted with template arguments similar to the nativewrappers.

Mon, Jul 15, 1:48 PM
elexis added a comment to D2042: Atlas variation submenu.

Perhaps it would be good to display the names of the variations. For example if one selects a cavalry unit, there are a range of colors settings, but one can't deduce which color refers to which property of the model.
(It's a bit unexpected that one can chose whether it's carrying meat or running, but maybe that's technically impossible to check)
Perhaps also more horizontal space for the panel, it looks a bit squeezed.

Mon, Jul 15, 1:43 PM
elexis commandeered D1754: Pass an argument to GUI events / SendEventToAll, remove loading screen progess workaround.
Mon, Jul 15, 1:13 PM
elexis added inline comments to D1991: simulation2: Remove Vector2D/Vector3D prototype workaround from EngineScriptConversions.
Mon, Jul 15, 12:58 PM
elexis added inline comments to D2042: Atlas variation submenu.
Mon, Jul 15, 12:53 PM
elexis added a comment to D2079: Adds snapping to edges for buildings.

This seems to me like desirable default placement behavior.

Being able to opt out with a hotkey?

Mon, Jul 15, 12:10 PM
elexis added a comment to D2054: Apply modifications to template values based on entity owner instead watching player.

"Apply modifications to template values based on entity owner instead watching player" sounds like #5109

Mon, Jul 15, 11:51 AM
elexis added a comment to rP22266: moves territory_pull.xml from other/ to special/ and updates the only file in….

That or keeping other/.

Mon, Jul 15, 4:20 AM
elexis committed rP22478: Moar lineendings..
Moar lineendings.
Mon, Jul 15, 4:20 AM
elexis accepted D1991: simulation2: Remove Vector2D/Vector3D prototype workaround from EngineScriptConversions.

Yes. Test plan I wrote in the commit myself includes those steps, and I've confirmed them locally as well before submitting. The game compiles, starts, and can create a game and select/move units without errors.

Mon, Jul 15, 4:10 AM
elexis added a comment to D1971: Support attack-ground for ranged units..

Didn't follow the discussion but I heard there is a vector serialization issue and I didn't see the reference #4698.
It's that the serializer only serializes objects, so you do get the { x, y } part but that object doesnt have the VectorXD object as the prototype, so use of these functions fails.
The workaround had been to use the static vector functions or an existing vector to do the math.

Mon, Jul 15, 3:57 AM
elexis added a comment to rP19154: Fix r19147. Bug noticed by mimo. Some objects are not instance of Vector3D….

refs #4698

Mon, Jul 15, 3:52 AM
elexis planned changes to D2058: FixedVector2D CompareLengthSquared.

As discussed with Vladislav on IRC in the past days, this iteration of the diff fails for fractional values, specifically this test:

		fixed n = fixed::FromFloat(1.0f / 10.0f);
		fixed n2 = n.Square();
		CFixedVector2D vec(n, n);
		fixed len = n2.Multiply(fixed::FromInt(2));
		TS_ASSERT_EQUALS(n2, len);
Mon, Jul 15, 2:44 AM
elexis added inline comments to D2079: Adds snapping to edges for buildings.
Mon, Jul 15, 12:57 AM
elexis awarded D2079: Adds snapping to edges for buildings a Orange Medal token.
Mon, Jul 15, 12:53 AM
elexis added a comment to D2079: Adds snapping to edges for buildings.

Very nice feature! From testing, feels perfect!

Mon, Jul 15, 12:51 AM

Sun, Jul 14

elexis added inline comments to D2042: Atlas variation submenu.
Sun, Jul 14, 9:59 PM
elexis added a comment to D1938: Easier introduction of new damage types..

It is important to analyze why the code was the way it was written - g_DamageTypes was introduced in rP20203 rP20625 (I can't read through everything now)

Sun, Jul 14, 6:56 PM
elexis added inline comments to rP22466: Cleanup MapGenerator.cpp, use VfsPath, create RegisterScriptFunctions, add….
Sun, Jul 14, 5:50 PM
elexis committed rP22476: Remove the unused variable proto from the ComponentManager in rP14877..
Remove the unused variable proto from the ComponentManager in rP14877.
Sun, Jul 14, 4:13 PM
elexis closed D1817: Remove unused variable and fail to register component if it has no constructor [ComponentManager].
Sun, Jul 14, 4:13 PM
elexis added inline comments to rP14877: Second (main) commit for the SpiderMonkey upgrade..
Sun, Jul 14, 4:11 PM
elexis accepted D1817: Remove unused variable and fail to register component if it has no constructor [ComponentManager].

Thanks for taking the review remarks here and on IRC into account, and extending the scope slightly beyond the removed variable.

Sun, Jul 14, 4:05 PM
elexis requested changes to D1707: Hard-counter tooltips.

Reading my old recommendation to use arrays and array functions and knowing that this function is at least called once per turn, once per selection, it's probably better to put this under scrutiny for performance issues.

Sun, Jul 14, 1:49 PM
elexis commandeered D1424: Fix pathfinding issue with units not garrison wallturrets due to a UnitMotion ObstructionFilter inconsistency.

(I wrote that)

Sun, Jul 14, 1:21 PM
elexis added inline comments to rP22346: Allow techs to affect unit counters stats..
Sun, Jul 14, 1:07 PM
elexis committed rP22472: Enable eslint rule no-floating-decimal and remove all superfluous floating….
Enable eslint rule no-floating-decimal and remove all superfluous floating…
Sun, Jul 14, 12:03 PM
elexis closed D2003: Linting: Enable 'no-floating-decimal' rule in ESLint.
Sun, Jul 14, 12:03 PM
elexis accepted D2003: Linting: Enable 'no-floating-decimal' rule in ESLint.

Correctness:
I agree that having this rule is nice to improve consistency, and in my practice since I'm here we always removed trailing zeroes and 'trailing periods'.
Then read the specs here https://eslint.org/docs/rules/no-floating-decimal.

Sun, Jul 14, 11:58 AM
elexis added a comment to rP22266: moves territory_pull.xml from other/ to special/ and updates the only file in….

@Nescio renaming to template_territory_pull.xml would solve the problem without adding the README.txt complexity, and it wouldn't be a cheat because delenda est can continue to use a custom other/ or rename to myfieldsthing/, and if someone was to add more territory pull child templates to vanilla, they may be added to a territory_pull/, which is still better than other/ it seems. Make this red go away.

Sun, Jul 14, 12:21 AM
elexis accepted D2020: Refactors Camera aspect ratio and calculation of plane points.

Compared the previous code against the new one and found no behavior changes.
Tested in accordance with the phrased testplan, didn't explode.
Didn't check the test, but it succeeded.

Sun, Jul 14, 12:14 AM

Sat, Jul 13

elexis committed rP22466: Cleanup MapGenerator.cpp, use VfsPath, create RegisterScriptFunctions, add….
Cleanup MapGenerator.cpp, use VfsPath, create RegisterScriptFunctions, add…
Sat, Jul 13, 11:41 PM
elexis closed D2068: Cleanup MapGenerator ScriptInterface.
Sat, Jul 13, 11:41 PM
elexis added a comment to D2068: Cleanup MapGenerator ScriptInterface.

May I use the phrasing "VfsPath change accepted by historic bruno" in the commit message?

Sat, Jul 13, 8:28 PM
elexis added a comment to D2068: Cleanup MapGenerator ScriptInterface.

On the VfsPath:

  • RegisterScriptFunction using VfsPath would be done consistently for multiple JS interface files, in particular for replays, this diff would only be one of them.
  • The pattern has been in use so far for LoadMapSettings which loads the XML map settings JSON data.
  • It is supported code-wise because the conversion explicitly exists in FromJSVal<Path> in ScriptConversions.cpp and in vfs_path.h: typedef Path VfsPath.
  • Tested on arch with gcc 9 and win 10 / vs 2015 in accordance with test plan.
  • On the tested platforms, using a non-ASCII filename such as african_plains🗡.js or fields_of_meroë.js did not work (file not found), not before nor after the patch. Using VfsPath will rather make it easier if all calls use VfsPath rather than all calls converting manually from wstring to VfsPath.
Sat, Jul 13, 7:53 PM
elexis created D2068: Cleanup MapGenerator ScriptInterface.
Sat, Jul 13, 2:40 PM
elexis added inline comments to D1817: Remove unused variable and fail to register component if it has no constructor [ComponentManager].
Sat, Jul 13, 1:56 PM
elexis added inline comments to rP14877: Second (main) commit for the SpiderMonkey upgrade..
Sat, Jul 13, 1:44 PM
elexis added inline comments to rP14877: Second (main) commit for the SpiderMonkey upgrade..
Sat, Jul 13, 1:28 PM

Fri, Jul 12

elexis committed rP22460: Fix lineendings..
Fix lineendings.
Fri, Jul 12, 7:41 PM
elexis committed rP22459: Unify the magic number indicating the number of impassable tiles at the map….
Unify the magic number indicating the number of impassable tiles at the map…
Fri, Jul 12, 6:39 PM
elexis closed D2061: Unify EdgeTile count in ICmpRangeManager, CCmpPathfinder, MapGenerator.
Fri, Jul 12, 6:39 PM
elexis updated the diff for D2061: Unify EdgeTile count in ICmpRangeManager, CCmpPathfinder, MapGenerator.

Rename to MapEdgeTiles

Fri, Jul 12, 5:52 PM
elexis added a comment to D2063: Unit motion - Check the return value of ComputeGoal and handle failure cases.

The two replays don't crash anymore for me, but I didn't check whether this is a false positive (since the replays don't store hashes)

Fri, Jul 12, 5:17 PM
elexis raised a concern with rP22450: Unit Motion - combine Goal computation logic from MoveToPoint and MoveToTarget.

Triggers a stack smash: #5496

Fri, Jul 12, 4:39 PM
elexis updated the diff for D2058: FixedVector2D CompareLengthSquared.

Remove commented out line.

Fri, Jul 12, 1:40 AM
elexis updated the diff for D2058: FixedVector2D CompareLengthSquared.

Revert cpp move.

Fri, Jul 12, 1:38 AM
elexis updated the summary of D2058: FixedVector2D CompareLengthSquared.
Fri, Jul 12, 1:37 AM
elexis added a comment to D2058: FixedVector2D CompareLengthSquared.

CompareLength is already comparing squared vectors, so I'm assuming you just want to avoid paying the cost of a multiplication every time ?

Yes

Fri, Jul 12, 1:11 AM

Thu, Jul 11

elexis updated the diff for D2061: Unify EdgeTile count in ICmpRangeManager, CCmpPathfinder, MapGenerator.

Move to EdgeTiles.h

Thu, Jul 11, 11:01 PM
elexis added a comment to D2057: GnuTLS configure requires pkg-config, breaking macOS build.

See #2780

Thu, Jul 11, 9:26 PM
elexis added a comment to D2062: [ATLASUI] Build atlasUI inside xcode missing libpng, libjpeg.

Fixed in rP22454.

Thu, Jul 11, 7:00 PM
elexis added a comment to D2061: Unify EdgeTile count in ICmpRangeManager, CCmpPathfinder, MapGenerator.

Bit weird for only one line, but Player.h set the precedent; static const ssize_t EDGE_TILES = 3; then I suppose.

Thu, Jul 11, 2:10 PM
Angen awarded D2061: Unify EdgeTile count in ICmpRangeManager, CCmpPathfinder, MapGenerator a Like token.
Thu, Jul 11, 7:05 AM
elexis updated the Trac tickets for D2061: Unify EdgeTile count in ICmpRangeManager, CCmpPathfinder, MapGenerator.
Thu, Jul 11, 3:11 AM
elexis updated the summary of D2061: Unify EdgeTile count in ICmpRangeManager, CCmpPathfinder, MapGenerator.
Thu, Jul 11, 3:10 AM
elexis updated the Trac tickets for D2061: Unify EdgeTile count in ICmpRangeManager, CCmpPathfinder, MapGenerator.
Thu, Jul 11, 3:08 AM
elexis created D2061: Unify EdgeTile count in ICmpRangeManager, CCmpPathfinder, MapGenerator.
Thu, Jul 11, 3:07 AM

Wed, Jul 10

elexis added inline comments to D2057: GnuTLS configure requires pkg-config, breaking macOS build.
Wed, Jul 10, 3:19 PM
elexis added a comment to D1846: Adds a "properties"-property to resources and let mods be able to prevent resources from being bartered, traded and/or tributed..

Having dedicated functions would mean the JS complains about it when parsing the code

Actually not true, only happens when the code is run, and the only code that is run when the file is loaded is the creation of the prototype and registering of global functions.
So catching typos can happen in both cases only when the affected code is run during the simulation.
Actually the generic approach prevents testing for typos by design since it is undecidable whether there is just no resource with the tag in the current set of mods or whether there is a typo.

Wed, Jul 10, 2:34 PM
elexis added a comment to D1846: Adds a "properties"-property to resources and let mods be able to prevent resources from being bartered, traded and/or tributed..

"properties": ["barterable", "tradable", "tributable"],

I was wondering whether those should be strings passed around as opposed to booleans with dedicated functions.
Having dedicated functions would mean the JS complains about it when parsing the code, whereas strings are only detected when the specific line is run.
So in the first case you get an instant test on the entire code, the latter requires testing every line manually if one doesn't rule out typos by reading well.

Wed, Jul 10, 1:20 PM
elexis added a comment to rP22355: Unit Motion - Face point when calling StopMoving().

As reported in http://irclogs.wildfiregames.com/2019-07/2019-07-08-QuakeNet-%230ad-dev.log
Had this segfault on a Jebel Barkal match

#0  0x00007ffff61652e0 in __vfprintf_internal () from /usr/lib/libc.so.6
#1  0x00007ffff6178c5a in __vsnprintf_internal () from /usr/lib/libc.so.6
#2  0x0000555555831c83 in vsnprintf (__ap=0x7fffff7ff6d0, __fmt=0x555555c39d8b "%s", __n=255, __s=0x7fffff7ff684 "") at /usr/include/bits/stdio2.h:80
#3  CProfiler2::ThreadStorage::RecordAttribute (this=0x555555d907d0, fmt=fmt@entry=0x555555c39d8b "%s", argp=argp@entry=0x7fffff7ff7c0) at ../../../source/ps/Profiler2.cpp:398
#4  0x00005555556505dc in CProfiler2::RecordAttribute (this=<optimized out>, fmt=fmt@entry=0x555555c39d8b "%s") at ../../../source/ps/Profiler2.h:445
#5  0x0000555555668e49 in CComponentManager::PostMessage (this=0x555555ed52d8, ent=324, msg=...) at ../../../source/simulation2/system/ComponentManager.cpp:1021
#6  0x00005555556c71ce in CCmpPosition::AdvertisePositionChanges (this=0x555557971160) at ../../../source/simulation2/system/Entity.h:88
#7  CCmpPosition::TurnTo (this=0x555557971160, y=...) at ../../../source/simulation2/components/CCmpPosition.cpp:549
#8  0x000055555571415b in CCmpUnitMotion::FaceTowardsPointFromPos (z=..., x=..., pos=..., this=0x555557986000) at ../../../source/simulation2/components/CCmpUnitMotion.cpp:1242
#9  CCmpUnitMotion::StopMoving (this=0x555557986000) at ../../../source/simulation2/components/CCmpUnitMotion.cpp:495
#10 0x000055555573cfd1 in ScriptInterface_NativeMethodWrapper<void, ICmpUnitMotion>::call<void (ICmpUnitMotion::*)()> (fptr=&virtual ICmpUnitMotion::StopMoving(), c=<optimized out>)
    at ../../../source/scriptinterface/NativeWrapperDefns.h:106
#11 ScriptInterface::callMethod<void, &class_ICmpUnitMotion, ICmpUnitMotion, &ICmpUnitMotion::StopMoving> (cx=cx@entry=0x555555dd29e0, argc=<optimized out>, vp=0x7fffff7fff78) at ../../../source/scriptinterface/NativeWrapperDefns.h:145
#12 0x00007ffff77cbb12 in js::CallJSNative (args=..., native=0x55555573cec0 <ScriptInterface::callMethod<void, &class_ICmpUnitMotion, ICmpUnitMotion, &ICmpUnitMotion::StopMoving>(JSContext*, unsigned int, JS::Value*)>, 
    cx=0x555555dd29e0) at ../../dist/include/js/CallArgs.h:204
#13 js::Invoke (cx=cx@entry=0x555555dd29e0, args=..., construct=construct@entry=js::NO_CONSTRUCT) at /home/elexis/code/0ad-git-private/0ad/libraries/source/spidermonkey/mozjs-38.0.0/js/src/vm/Interpreter.cpp:498
#14 0x00007ffff77cc691 in js::Invoke (cx=cx@entry=0x555555dd29e0, thisv=..., fval=..., argc=argc@entry=0, argv=<optimized out>, rval=..., rval@entry=...)
    at /home/elexis/code/0ad-git-private/0ad/libraries/source/spidermonkey/mozjs-38.0.0/js/src/vm/Interpreter.cpp:554
#15 0x00007ffff7a9d954 in js::jit::InvokeFunction (cx=0x555555dd29e0, obj=..., argc=0, argv=<optimized out>, rval=0x7fffff8000e0) at ../../dist/include/js/Value.h:1475
Wed, Jul 10, 1:50 AM
elexis accepted rP22370: Refactors Developer Overlay using an object instead of global functions..

False alert, the user used a mod.

Wed, Jul 10, 1:22 AM
elexis updated the diff for D2058: FixedVector2D CompareLengthSquared.

Missing parenthesis

Wed, Jul 10, 1:03 AM
elexis created D2058: FixedVector2D CompareLengthSquared.
Wed, Jul 10, 12:17 AM

Tue, Jul 9

elexis added a comment to D1501: Mapgen: library improvements.
In D1501#85699, @elexis wrote:
  • TileClass is to be changed more deeply, it's on our radar and there are various patches floating around

such as?

D1637 for instance

Tue, Jul 9, 9:48 PM
elexis added a comment to D1501: Mapgen: library improvements.
  • TileClass is to be changed more deeply, it's on our radar and there are various patches floating around
  • the library.js functions would be better moved to the constraint constructors IMO (new AvoidClassesConstraint)
Tue, Jul 9, 9:20 PM
elexis added a comment to D2057: GnuTLS configure requires pkg-config, breaking macOS build.

(See discussion on #0ad-dev, it will need an update for 3.6 and the 3.6 update ought to be before a release. The patch is intended for the current 3.5)

Tue, Jul 9, 9:17 PM
elexis raised a concern with rP22370: Refactors Developer Overlay using an object instead of global functions..

From https://wildfiregames.com/forum/index.php?/topic/26471-error-javascript-error-guisessioninputjs-line-799-referenceerror-g_developeroverlay-is-not-defined/

Tue, Jul 9, 6:25 PM
elexis updated the Trac tickets for D2053: Round value of capture attack.
Tue, Jul 9, 12:43 PM
wraitii awarded rP22443: Fix some gcc 8 and gcc 9 compiler warnings that were thrown 4500 times. a Like token.
Tue, Jul 9, 10:30 AM
elexis committed rP22443: Fix some gcc 8 and gcc 9 compiler warnings that were thrown 4500 times..
Fix some gcc 8 and gcc 9 compiler warnings that were thrown 4500 times.
Tue, Jul 9, 2:19 AM
elexis closed D2055: Fix 4500 gcc 8, gcc 9 compile warnings.
Tue, Jul 9, 2:19 AM
elexis added inline comments to D2055: Fix 4500 gcc 8, gcc 9 compile warnings.
Tue, Jul 9, 2:16 AM
elexis added inline comments to D2055: Fix 4500 gcc 8, gcc 9 compile warnings.
Tue, Jul 9, 1:30 AM
elexis updated the diff for D2055: Fix 4500 gcc 8, gcc 9 compile warnings.

Don't use the string() getter for path.h but the member variable directly so it's consistent for the assignment operator.

Tue, Jul 9, 1:12 AM
elexis updated the diff for D2055: Fix 4500 gcc 8, gcc 9 compile warnings.

whitespace, Sprite rework

Tue, Jul 9, 1:08 AM
elexis added inline comments to D2055: Fix 4500 gcc 8, gcc 9 compile warnings.
Tue, Jul 9, 1:05 AM

Mon, Jul 8

elexis added inline comments to D2055: Fix 4500 gcc 8, gcc 9 compile warnings.
Mon, Jul 8, 8:44 PM
elexis added a comment to D2042: Atlas variation submenu.

Thanks for having taken into account my IRC review remarks in the past days about the "|" concatenation in the attribute setting vs. individual XML nodes.

Mon, Jul 8, 5:50 PM
elexis updated the diff for D2055: Fix 4500 gcc 8, gcc 9 compile warnings.

Use initializer list for CSize

Mon, Jul 8, 4:30 AM