fix(input): Fix double-click fast drive timing in ParticleUplinkCannon #2012

bobtista · 2025-12-19T17:11:18Z

Closes #2011

Testing (TODO)

Fast drive mode activation via double-click now works consistently at any frame rate and logic time scale setting.

stephanmeesters · 2025-12-20T13:25:35Z

I could do a bit of testing and review for this

For both 30hz and 60hz :

double click to activate fast beam should feel the same
after activating fast beam the trajectory should be the same
make a save game right after activating the fast beam then check if the trajectory is the same (for tsh regular and retail compatible versions)
load a retail savegame and check that the beam trajectory is the same as in the retail version

Does that seem right?

GeneralsMD/Code/GameEngine/Source/GameLogic/Object/Update/ParticleUplinkCannonUpdate.cpp

GeneralsMD/Code/GameEngine/Include/GameLogic/Module/ParticleUplinkCannonUpdate.h

GeneralsMD/Code/GameEngine/Source/GameLogic/Object/Update/ParticleUplinkCannonUpdate.cpp

GeneralsMD/Code/GameEngine/Include/GameLogic/Module/ParticleUplinkCannonUpdate.h

GeneralsMD/Code/GameEngine/Source/GameLogic/Object/Update/ParticleUplinkCannonUpdate.cpp

bobtista · 2025-12-21T19:56:35Z

Replicated to Generals in the most recent commit, which differs slightly and is worth giving a quick look

stephanmeesters · 2025-12-22T12:46:20Z

Tested this by making a savegame of a skirmish at max gamespeed, then loaded it back at regular gamespeed. Trajectory of the beam and beam duration appeared the same also using fast drive

Using the compatibility flags I loaded a retail gamesave, and also saved a game and loaded that in retail. That works fine

I thought it would be interesting to test 30hz and 60hz with the same gamespeed the way GO does it, stephanmeesters@7491c32

Double click felt the same, trajectory and duration the same

It did have problems with the replay as the beam duration was no longer the same, but the gametime was also reported wrong, so that's probably just because I hacked that GO code in

GeneralsMD/Code/GameEngine/Source/GameLogic/Object/Update/ParticleUplinkCannonUpdate.cpp

Caball009

Tested and works. Looks good to me.

Edit: tested only in skirmish.

xezon

I think this fix is unsafe.

The safe fix likely is to scale m_doubleClickToFastDriveDelay with LOGICFRAMES_PER_SECOND/BaseFps (and other scales, see FramePacer).

xezon · 2026-01-06T17:21:44Z

Generals/Code/GameEngine/Source/GameLogic/Object/Update/ParticleUplinkCannonUpdate.cpp

-#if RETAIL_COMPATIBLE_XFER_SAVE
-	const XferVersion currentVersion = 2;
+#if RETAIL_COMPATIBLE_CRC || RETAIL_COMPATIBLE_XFER_SAVE
+	const XferVersion currentVersion = 3;


This should be 2.

Fixed. Generals retail compat version is now 2

Generals/Code/GameEngine/Source/GameLogic/Object/Update/ParticleUplinkCannonUpdate.cpp

xezon · 2026-01-06T17:41:52Z

GeneralsMD/Code/GameEngine/Source/GameLogic/Object/Update/ParticleUplinkCannonUpdate.cpp

+				const Bool useFasterSpeed = m_scriptedWaypointMode || (m_lastDrivingClickFrame - m_2ndLastDrivingClickFrame < data->m_doubleClickToFastDriveDelay);
+#else
+				DEBUG_ASSERTCRASH(m_lastDrivingClickTimeMsec >= m_2ndLastDrivingClickTimeMsec, ("m_lastDrivingClickTimeMsec should always be >= m_2ndLastDrivingClickTimeMsec"));
+				const Bool useFasterSpeed = m_scriptedWaypointMode || (m_lastDrivingClickTimeMsec - m_2ndLastDrivingClickTimeMsec < data->m_doubleClickToFastDriveDelay);


I did not test this but just reading the code I expect that this can and will cause mismatch in Multiplayer. Mismatch probably will not happen always, but there is no guarantee that the delta between 2 timeGetTime calls on 2 different machines will always have the same delta between them, because timeGetTime is low-resolution and time itself is relative in the Network.

The timeGetTime values are no longer xfer'd. The double-click detection only affects the local player's beam speed selection and doesn't impact game state synchronization, right? Each player's clicks are processed locally before being sent as network commands.

Please test with 2 clients in Network match. The clicks are used to determine logic state. If these do not match exactly, then clients will use different laser beam speeds.

You're right, I ended up going with frames and scaling the comparison in frame count so it behaves the same.

This would work for GO 60 hz, but not for increased logic timescale, right? Maybe getActualLogicTimeScaleOverFpsRatio can help with that.

Hmmm - this comment says "void setLogicTimeScaleFps( Int fps ); ///< Set the logic time scale fps and therefore scale the simulation time. Is capped by the max render fps and does not apply to network matches." and yeah we check

if (TheNetwork != NULL) { return TheNetwork->getFrameRate();

so multiplayer should be just using whatever this is.

Are you talking about playing single player games and manually changing the logic fps? Are you asking about getActualLogicTimeScaleOverFpsRatio vs getActualLogicTimeScaleRatio?

Did you test the current code in skirmish mode with high logic time scale? Like is there still a 500 msec window to double click at a logic timescale of 4?

Ideally this would also correct scale with logic time scale. For example set logic fps to 90 in skirmish, simulation will be 3 times as fast, and so double click duration needs to be trippled too I think (you can test it).

Ok I think I've got it right now, and it should work when we speed up logic in skirmish too. I can test soon

Ah, but this caused CI replay mismatches :(

Significant changes since approval.

xezon · 2026-01-08T10:43:34Z

Generals/Code/GameEngine/Source/GameLogic/Object/Update/ParticleUplinkCannonUpdate.cpp

-				const UnsignedInt scaledDelay = data->m_doubleClickToFastDriveDelay * LOGICFRAMES_PER_SECOND / BaseFps;
+				// TheSuperHackers @fix Scale delay threshold by frame rate and logic time scale to maintain consistent real-time feel
+				UnsignedInt scaledDelay = data->m_doubleClickToFastDriveDelay * LOGICFRAMES_PER_SECOND / BaseFps;
+				if (TheFramePacer != NULL)


Is not expected NULL

xezon · 2026-01-08T10:45:22Z

GeneralsMD/Code/GameEngine/Source/GameLogic/Object/Update/ParticleUplinkCannonUpdate.cpp

+				if (TheFramePacer != NULL)
+				{
+					const Real timeScaleRatio = TheFramePacer->getActualLogicTimeScaleRatio();
+					scaledDelay = (UnsignedInt)(scaledDelay * timeScaleRatio);


Curious why this causes mismatch. Perhaps check value in headless replay. It is expected 1.

Perhaps we must not scale it this way. It would mean we cannot use time scale to fast forward replays. Perhaps we must accept that double click will be jank then, or have an alternative code path that only works in Single Player and not in Multiplayer & Replay Playback, but this would increase complexity.

Checking for single player / skirmish is trivial, right? I think I'd rather have that than a poorly working double click.

Or how about we make it a proper game message instead of this frame trickery? Won't be Retail compatible, but much cleaner?

We could add a new game message eg MSG_DO_SPECIAL_POWER_OVERRIDE_DESTINATION_FAST
or we could add an argument to the existing one MSG_DO_SPECIAL_POWER_OVERRIDE_DESTINATION

// In CommandXlat: gameMsg->appendLocationArgument(*pos); gameMsg->appendIntegerArgument(SPECIAL_INVALID); gameMsg->appendObjectIDArgument(INVALID_ID); #if !RETAIL_COMPATIBLE_CRC gameMsg->appendBoolArgument(useFastDrive); // NEW #endif

I'll run with this - but it does mean we can't use this feature until we break retail compat

I think it is ok to fix for later. We do not need all fixes for VC6.

Ok, happy to close this if you like. I also took a stab at using MSG_DO_SPECIAL_POWER_OVERRIDE_DESTINATION

…l-time milliseconds

…leUplinkCannon

…non retail builds

…ing, add debug assertions

…UplinkCannon xfer version

… units mismatch

…or non-retail builds

…nstead of timeGetTime

…r fast-forward support

…t beam mode behind RETAIL_COMPATIBLE_CRC

bobtista force-pushed the bobtista/fix-particle-uplink-double-click-timing branch 3 times, most recently from bedd3a9 to 4718d05 Compare December 19, 2025 21:41

bobtista marked this pull request as ready for review December 19, 2025 22:24

Caball009 reviewed Dec 20, 2025

View reviewed changes

bobtista force-pushed the bobtista/fix-particle-uplink-double-click-timing branch from 4718d05 to fdb318e Compare December 20, 2025 20:26

Caball009 reviewed Dec 20, 2025

View reviewed changes

GeneralsMD/Code/GameEngine/Source/GameLogic/Object/Update/ParticleUplinkCannonUpdate.cpp Show resolved Hide resolved

Caball009 added Gen Relates to Generals ZH Relates to Zero Hour NoRetail This fix or change is not applicable with Retail game compatibility Input Minor Severity: Minor < Major < Critical < Blocker Bug Something is not working right, typically is user facing labels Dec 21, 2025

bobtista force-pushed the bobtista/fix-particle-uplink-double-click-timing branch from fdb318e to d3e0024 Compare December 21, 2025 19:53

bobtista force-pushed the bobtista/fix-particle-uplink-double-click-timing branch 2 times, most recently from 8e53756 to c01dbb2 Compare December 22, 2025 03:58

Caball009 reviewed Dec 24, 2025

View reviewed changes

GeneralsMD/Code/GameEngine/Source/GameLogic/Object/Update/ParticleUplinkCannonUpdate.cpp Show resolved Hide resolved

Caball009 reviewed Dec 31, 2025

View reviewed changes

GeneralsMD/Code/GameEngine/Source/GameLogic/Object/Update/ParticleUplinkCannonUpdate.cpp Outdated Show resolved Hide resolved

Caball009 reviewed Dec 31, 2025

View reviewed changes

GeneralsMD/Code/GameEngine/Source/GameLogic/Object/Update/ParticleUplinkCannonUpdate.cpp Outdated Show resolved Hide resolved

Caball009 reviewed Jan 2, 2026

View reviewed changes

GeneralsMD/Code/GameEngine/Source/GameLogic/Object/Update/ParticleUplinkCannonUpdate.cpp Outdated Show resolved Hide resolved

Caball009 reviewed Jan 3, 2026

View reviewed changes

GeneralsMD/Code/GameEngine/Source/GameLogic/Object/Update/ParticleUplinkCannonUpdate.cpp Outdated Show resolved Hide resolved

Caball009 previously approved these changes Jan 4, 2026

View reviewed changes

xezon requested changes Jan 6, 2026

View reviewed changes

bobtista force-pushed the bobtista/fix-particle-uplink-double-click-timing branch from 871e244 to 99e68e1 Compare January 6, 2026 20:47

xezon reviewed Jan 8, 2026

View reviewed changes

fix(input): Convert double-click fast drive timing from frames to rea…

fe65f8a

…l-time milliseconds

bobtista added 15 commits January 8, 2026 16:26

refactor(input): Extract double-click condition to variable in Partic…

757a4d9

…leUplinkCannon

style: Fix spacing inconsistency in ParticleUplinkCannon xfer function

a40f288

refactor: Make millisecond variables conditional in ParticleUplinkCan…

750cf34

…non retail builds

fix(input): Fix double-click fast drive timing in ParticleUplinkCannon

69992b6

Replicate to generals

51afdcc

refactor(input): Remove unnecessary null checks from double-click tim…

4c8d39f

…ing, add debug assertions

Use RETAIL_COMPATIBLE_CRC || RETAIL_COMPATIBLE_XFER_SAVE for Particle…

55b8a0e

…UplinkCannon xfer version

fix(particle-uplink): use consistent RETAIL_COMPATIBLE macros and fix…

24fd78e

… units mismatch

refactor(particle-uplink): parse double-click delay as milliseconds f…

35efd03

…or non-retail builds

nit: update comments to match our format

2723f6e

remove redundant xfer load branch in ParticleUplinkCannonUpdate

d80e949

remove xfer of timeGetTime values and fix xfer version numbering

59b0786

refactor(particle-uplink): use frame scaling for double-click delay i…

d74ec2c

…nstead of timeGetTime

fix(particle-uplink): scale double-click delay by logic time scale fo…

52f5452

…r fast-forward support

revert xfer changes that were out of scope for this PR

c9b48ce

bobtista force-pushed the bobtista/fix-particle-uplink-double-click-timing branch from 43b4d61 to 31559b2 Compare January 9, 2026 00:37

feat(particle-uplink): add client-side double-click detection for fas…

6ea27b9

…t beam mode behind RETAIL_COMPATIBLE_CRC

bobtista force-pushed the bobtista/fix-particle-uplink-double-click-timing branch from 31559b2 to 6ea27b9 Compare January 9, 2026 00:45

fix(input): Fix double-click fast drive timing in ParticleUplinkCannon #2012

Are you sure you want to change the base?

fix(input): Fix double-click fast drive timing in ParticleUplinkCannon #2012

Uh oh!

Conversation

bobtista commented Dec 19, 2025

Testing (TODO)

Uh oh!

stephanmeesters commented Dec 20, 2025

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

bobtista commented Dec 21, 2025

Uh oh!

stephanmeesters commented Dec 22, 2025

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Caball009 left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

xezon left a comment • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Caball009 Jan 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

bobtista Jan 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Caball009 Jan 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

xezon Jan 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

xezon Jan 8, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Caball009 left a comment •

edited

Loading

xezon left a comment •

edited

Loading

Caball009 Jan 7, 2026 •

edited

Loading

bobtista Jan 7, 2026 •

edited

Loading

Caball009 Jan 7, 2026 •

edited

Loading

xezon Jan 7, 2026 •

edited

Loading

xezon Jan 8, 2026 •

edited

Loading