feat(network): Send product information to distinguish clients by Caball009 · Pull Request #1404 · TheSuperHackers/GeneralsGameCode

Caball009 · 2025-08-01T20:16:20Z

This PR makes our clients send product information to distinguish them from retail clients primarily. Retail clients will ignore the product information data that's sent to them.

Product information is exchanged on demand and never broadcast. A client is considered 'patched' if it responds to a product info request (or, in pre-match, if it sends one). The implementation consists of three parts.

player - player in lobby:

When a player detects a new player in the lobby, it sends a product info request to that player.
If the other player responds with an acknowledgement, they are considered patched.

player - host in lobby:

When a player detects a new game host in the lobby, it sends a product info request to that host.
If the host responds with an acknowledgement, it is considered patched.

players in pre-match (game room):

When a player joins a match, it sends a product info request to all players in that match.
Existing players treat this request as confirmation that the joining player is patched (no explicit acknowledgement required).

Examples of future use cases, visualizing who's not using the retail client

LAN Lobby:

Pre-match / Game Options:

Status in match:

TODO:

Add code comments.
Update the title, this message and labels to reflect that the color change is no longer part of this PR.
Use value from feat: Implement math CRC calculation utility for logic mismatch detection purposes #2100 for variable fpMathCRC.
Replicate in Generals.

GeneralsMD/Code/GameEngine/Source/GameNetwork/LANAPI.cpp

GeneralsMD/Code/GameEngine/Source/GameNetwork/GUIUtil.cpp

GeneralsMD/Code/GameEngine/Source/GameNetwork/LANAPICallbacks.cpp

GeneralsMD/Code/GameEngine/Source/GameNetwork/LANGameInfo.cpp

xezon

Good effort.

This work relates to #16

I think the packets need redesigning.

GeneralsMD/Code/GameEngine/Source/GameNetwork/LANAPI.cpp

GeneralsMD/Code/GameEngine/Source/GameNetwork/LANAPICallbacks.cpp

GeneralsMD/Code/GameEngine/Source/GameNetwork/LANAPIhandlers.cpp

GeneralsMD/Code/GameEngine/Include/GameNetwork/GameInfo.h

GeneralsMD/Code/GameEngine/Source/GameNetwork/LANAPI.cpp

GeneralsMD/Code/GameEngine/Source/GameClient/GUI/GUICallbacks/Diplomacy.cpp

GeneralsMD/Code/GameEngine/Include/GameNetwork/LANAPI.h

GeneralsMD/Code/GameEngine/Source/GameNetwork/LANAPI.cpp

L3-M · 2025-10-23T13:37:09Z

What is the current status of this change? I think this would be a nice change for the first release.

Caball009 · 2025-11-01T15:05:53Z

What is the current status of this change? I think this would be a nice change for the first release.

The current implementation ought to be changed to make it more efficient / optimal, but that requires much more work. I did some of the work locally, but it didn't really work out iirc, so the PR is kind of stuck for now.

Caball009 · 2025-11-06T14:03:58Z

I've revisited this PR and was able to move away from the broadcast implementation. It feels better now. I put networking related changes in one commit and the color / status stuff in another.

What sort of information do we want to exchange just to establish which clients are on a patched version? Maybe it's a good idea to send the commit hash as well.

Caball009 · 2025-11-06T14:22:33Z

Short overview of the new implementation. Patch information is not broadcast and only sent when needed.

Interacting with players in the lobby:

When a player detects a new player in the lobby, they send a patch information request to that player.
If the other player responds with an acknowledgement, they are on a 'patched version'.

Interacting with game hosts in the lobby:

When a player in the lobby detects a new game host, they send a patch information request to that host.
If the game host responds with an acknowledgement, they are on a 'patched version'.

Interacting with players in the pre-match game:

When a player joins a match, they send all players in that match a patch information request.
If the other players respond with an acknowledgement, they are on a 'patched version'.
The other players consider the patch information request from the joining player as acknowledgement.

Currently, players in the lobby do not have a way to know which players in a match are on a 'patched version' except the host. It doesn't seem necessary to me to have the host communicate this information, and it also breaks the current pattern where each client only sends data about itself.

GeneralsMD/Code/GameEngine/Source/GameClient/GUI/GUICallbacks/Diplomacy.cpp

xezon

This change is going into the right direction but needs more polishing and wider scope.

Are the pictures of the first post still relevant?

The terminology of the network data needs changing from "Patch" to "Product", because "Patch" is a much more limited term than "Product" is.

The ProductInfo message needs to contain all relevant information to describe the Product that the remote player is using. A version number alone will not be enough to properly describe a Product. Think of 100 different products that the user could interact with in the lobby.

GeneralsMD/Code/GameEngine/Include/GameNetwork/LANAPI.h

Caball009 · 2025-11-08T19:29:26Z

Are the pictures of the first post still relevant?

Yes, the colors are still like that; the status text will be slightly different. My current idea is that it shows the Git short hash string there (there isn't much room for additional text, unfortunately).

The terminology of the network data needs changing from "Patch" to "Product", because "Patch" is a much more limited term than "Product" is.

Fair enough. Everything should say 'product' instead of 'patch' now.

The ProductInfo message needs to contain all relevant information to describe the Product that the remote player is using. A version number alone will not be enough to properly describe a Product. Think of 100 different products that the user could interact with in the lobby.

I added a couple of stuff (e.g. exe and ini crc values), but I don't feel like this is final version yet.

Core/GameEngine/Include/GameNetwork/LANAPI.h

Core/GameEngine/Source/GameNetwork/LANAPIhandlers.cpp

Caball009 · 2025-11-30T19:15:53Z

I could use some feedback on this PR.

Core/GameEngine/Source/GameNetwork/LANAPIhandlers.cpp

xezon

Current thoughts on this. No exhaustive review.

Core/GameEngine/Include/GameNetwork/GameInfo.h

Core/GameEngine/Include/GameNetwork/LANAPI.h

GeneralsMD/Code/GameEngine/Source/GameNetwork/GUIUtil.cpp

GeneralsMD/Code/GameEngine/Source/GameClient/GUI/GUICallbacks/Diplomacy.cpp

Core/GameEngine/Include/GameNetwork/LANAPI.h

Core/GameEngine/Source/GameNetwork/LANAPIhandlers.cpp

greptile-apps

_{3 files reviewed, 2 comments}

_{Edit Code Review Agent Settings | Greptile}

Core/GameEngine/Source/GameNetwork/LANAPIhandlers.cpp

Caball009 · 2026-02-05T02:11:22Z

I'm slowly moving forward with this PR, though there are still several unresolved conversations. I've decided that the code for colored names is best suited for a follow-up PR.

xezon · 2026-02-08T10:23:44Z

Please set to draft if this is still WIP. It is not clear to me if Review can continue.

Caball009 · 2026-02-08T20:21:52Z

Please set to draft if this is still WIP. It is not clear to me if Review can continue.

This PR is ready for review. Perhaps there are a few minor details that may still change. I hope #2100 won't take much longer so I can include it in this PR.

For anyone who wants to test this, you can revert the reverted commit so that you can see the colored names and status again.

xezon · 2026-02-13T20:59:13Z

So #2100 is meant to be concluded before this change? I thought we would do #2100 afterwards.

xezon

Looks promising but there is still some polishing to do.

Core/GameEngine/Include/GameNetwork/GameInfo.h

Core/GameEngine/Include/GameNetwork/LANAPI.h

Core/GameEngine/Source/GameNetwork/LANAPIhandlers.cpp

greptile-apps

_{10 files reviewed, 1 comment}

_{Edit Code Review Agent Settings | Greptile}

Core/GameEngine/Source/GameNetwork/GameInfo.cpp

Caball009 · 2026-02-22T22:53:42Z

I just addressed most of the review feedback

I'd like to reduce the string buffer size from 200 characters to 190. For reference, we barely even use 50 characters depending on the user name: Community Patch�~1297�By Caball009�~8c71c3a61 I'm slightly worried that we'll want to add several additional fields later and not have enough room to do so in this type.

Caball009 · 2026-02-23T01:08:58Z

So #2100 is meant to be concluded before this change? I thought we would do #2100 afterwards.

There's been some discussion about what #2100 should do, but I think returning the hash can be done before this PR so it can be included.

xezon · 2026-03-02T18:55:24Z

I'm slightly worried that we'll want to add several additional fields later and not have enough room to do so in this type.

How much is left now? Are there not more than 200 characters of spare space?

xezon · 2026-03-02T19:03:49Z

Core/GameEngine/Include/GameNetwork/LANAPI.h

+		// - When a player detects a new game host in the lobby, it sends a product info request to that host.
+		// - If the host responds with an acknowledgement, it is considered patched.
+		// 3. players in pre-match:
+		// - When a player joins a match, it sends a product info request to all players in that match.


Do the previous players also receive Product Info from the new joined player?

Yes. Does the comment suggest otherwise to you?

Yes it is not clear to me.

It says:

Existing players treat this request as confirmation that the joining player is patched

This implies that existing player do not receive product info but just know that the requester is not a Retail guy.

Core/GameEngine/Source/GameNetwork/LANAPI.cpp

xezon · 2026-03-02T19:13:43Z

Core/GameEngine/Source/GameNetwork/LANAPI.cpp

 }

-void LANAPI::sendMessage(LANMessage *msg, UnsignedInt ip /* = 0 */)
+void LANAPI::sendMessage(LANMessage *msg, UnsignedInt ip /* = 0 */, Bool broadcast /*= TRUE*/)


The term broadcast is confusing. Because when it is set to FALSE it will still broadcast to game room participants.

Perhaps split this function into multiple to make its usage more explicit.

For example have a LANAPI::sendMessageToGameRoomPlayers(LANMessage *msg)

This is getting outside the scope of this PR imo.

Should be simple enough. Can write like so:

void LANAPI::sendMessage(LANMessage *msg, UnsignedInt ip /* = 0 */) { if (ip != 0) { m_transport->queueSend(ip, lobbyPort, (unsigned char *)msg, sizeof(LANMessage) /*, 0, 0 */); } else if ((m_currentGame != nullptr) && (m_currentGame->getIsDirectConnect())) { sendMessageToGameRoomPlayers(msg); } else { m_transport->queueSend(m_broadcastAddr, lobbyPort, (unsigned char *)msg, sizeof(LANMessage) /*, 0, 0 */); } } void LANAPI::sendMessageToGameRoomPlayers(LANMessage *msg) { if (!m_currentGame) return; Int localSlot = m_currentGame->getLocalSlotNum(); for (Int i = 0; i < MAX_SLOTS; ++i) { if (i != localSlot) { GameSlot *slot = m_currentGame->getSlot(i); if ((slot != nullptr) && (slot->isHuman())) { m_transport->queueSend(slot->getIP(), lobbyPort, (unsigned char *)msg, sizeof(LANMessage) /*, 0, 0 */); } } } }

This way you can simply call sendMessageToGameRoomPlayers(msg) instead of doing the argument tricks.

xezon · 2026-03-02T20:58:20Z

Core/GameEngine/Source/GameNetwork/LANAPIhandlers.cpp

+	}
+	else
+	{
+		output[ARRAY_SIZE(output) - 1] = '\0';


Would there not already be a guaranteed null via wcslcpy?

xezon · 2026-03-02T21:01:26Z

Core/GameEngine/Source/GameNetwork/LANAPIhandlers.cpp

+	// null terminate the output buffer to signal the end of the last input string
+	if (outputIndex < ARRAY_SIZE(output))
+	{
+		output[outputIndex++] = '\0';


Would there not already be a guaranteed null via wcslcpy?

Can we just do

return outputIndex < ARRAY_SIZE(output);

Core/GameEngine/Source/GameNetwork/LANAPIhandlers.cpp

xezon · 2026-03-02T21:13:09Z

Core/GameEngine/Source/GameNetwork/LANAPIhandlers.cpp

+		output[i]->set(input + inputIndex, length);
+
+		inputIndex += length + 1;
+		if (inputIndex >= ARRAY_SIZE(input))


This condition looks as if it returns true when the input is exactly 200 characters, perfectly fitting?

Perhaps do inputIndex += sizeof('\3') after this condition only?

Core/GameEngine/Include/GameNetwork/LANAPI.h

xezon · 2026-03-02T21:26:14Z

GeneralsMD/Code/GameEngine/Source/Common/GameEngine.cpp

 	m_logicTimeAccumulator = 0.0f;
 	m_quitting = FALSE;
 	m_isActive = FALSE;
+	m_launchTime = ::timeGetTime();


What is the purpose of the launch time? I suspect to see for how long the game has been running? timeGetTime() return the milliseconds from Windows start until now. This is not a useful information for others.

What we likely need to send is a UTC time. With c++20 that would be chrono::utc_time, but with VC6 we would need gmtime https://en.cppreference.com/w/c/chrono/gmtime

The only reason I went with launchTime is previous feedback from you tbh.

I'd rather go back to up time than doing much work for launch time. I see no benefit knowing the launch time, just the up time.

Yes but uptime is not particularly meaningful because it expires with passing time. Launch time is conceptually good but it needs the right value.

xezon · 2026-03-02T21:28:53Z

Core/GameEngine/Include/GameNetwork/GameInfo.h

+		};
+
+		UnsignedInt flags;
+		UnsignedInt launchTime;


If this is meant to represent unix time, then it needs to be 64 bits because 32 bits value wraps in 2038.

Same comment as above. That's just 4 wasted bytes as far as I'm concerned.

Hmm ok. You could lower the precision of the 64 bit time from seconds to minutes but that also seems a bit hacky. You could go back to uptime but then it is an accurate value, unless you also keep the arrival time of the uptime to add the delta on top.

Unclear to me what the most elegant solution here is. What is this time for? To see when someone has not restarted his game before a match?

If this is meant to tell whether a remote client has played other matches beforehand, then how about count the number of matches played instead of the uptime? Then the value can also be just 2 or 1 bytes. When match count shows 0, then we know it is a fresh client start. When uptime shows 3 hours, we can only suspect that the match count is larger than 0. It could be 0 even after 3 hours of uptime.

Core/GameEngine/Source/GameNetwork/LANAPIhandlers.cpp

Caball009 added Enhancement Is new feature or request GUI For graphical user interface Network Anything related to network, servers Gen Relates to Generals ZH Relates to Zero Hour labels Aug 1, 2025

Caball009 commented Aug 10, 2025

View reviewed changes

GeneralsMD/Code/GameEngine/Source/GameNetwork/LANAPI.cpp Outdated Show resolved Hide resolved

Caball009 marked this pull request as ready for review August 10, 2025 19:34

Caball009 requested a review from a team August 11, 2025 15:24

Skyaero42 reviewed Aug 11, 2025

View reviewed changes

GeneralsMD/Code/GameEngine/Source/GameNetwork/GUIUtil.cpp Outdated Show resolved Hide resolved

Skyaero42 reviewed Aug 11, 2025

View reviewed changes

GeneralsMD/Code/GameEngine/Source/GameNetwork/LANAPICallbacks.cpp Outdated Show resolved Hide resolved

Skyaero42 reviewed Aug 11, 2025

View reviewed changes

GeneralsMD/Code/GameEngine/Source/GameNetwork/LANGameInfo.cpp Outdated Show resolved Hide resolved

xezon reviewed Aug 11, 2025

View reviewed changes

L3-M added this to the Important features milestone Oct 23, 2025

Caball009 force-pushed the show_clients_with_sh_patch branch from 9fb6a01 to c4e3f1d Compare November 6, 2025 13:52

Caball009 requested a review from a team November 6, 2025 14:04

Caball009 commented Nov 6, 2025

View reviewed changes

GeneralsMD/Code/GameEngine/Source/GameClient/GUI/GUICallbacks/Diplomacy.cpp Outdated Show resolved Hide resolved

xezon reviewed Nov 7, 2025

View reviewed changes

Caball009 commented Nov 8, 2025

View reviewed changes

Core/GameEngine/Include/GameNetwork/LANAPI.h Show resolved Hide resolved

Caball009 changed the title ~~feat(gui): Distinguish patch players from retail players by a colored name and different status~~ feat(network): Send product information to distinguish clients (with a colored name and different status) Nov 8, 2025

Caball009 commented Nov 9, 2025

View reviewed changes

Core/GameEngine/Source/GameNetwork/LANAPIhandlers.cpp Show resolved Hide resolved

bobtista reviewed Nov 30, 2025

View reviewed changes

Core/GameEngine/Source/GameNetwork/LANAPIhandlers.cpp Outdated Show resolved Hide resolved

xezon reviewed Dec 1, 2025

View reviewed changes

Caball009 commented Dec 4, 2025

View reviewed changes

Core/GameEngine/Include/GameNetwork/LANAPI.h Show resolved Hide resolved

Caball009 commented Dec 13, 2025

View reviewed changes

Core/GameEngine/Source/GameNetwork/LANAPIhandlers.cpp Show resolved Hide resolved

greptile-apps bot reviewed Feb 5, 2026

View reviewed changes

Core/GameEngine/Source/GameNetwork/LANAPIhandlers.cpp Outdated Show resolved Hide resolved

Core/GameEngine/Source/GameNetwork/LANAPIhandlers.cpp Outdated Show resolved Hide resolved

Caball009 changed the title ~~feat(network): Send product information to distinguish clients (with a colored name and different status)~~ feat(network): Send product information to distinguish clients Feb 8, 2026

Caball009 removed the GUI For graphical user interface label Feb 8, 2026

xezon reviewed Feb 13, 2026

View reviewed changes

Caball009 added 2 commits February 22, 2026 20:20

Merge branch 'main' into show_clients_with_sh_patch

765e5fa

Addressed feedback (part 1).

8c71c3a

greptile-apps bot reviewed Feb 22, 2026

View reviewed changes

Core/GameEngine/Source/GameNetwork/GameInfo.cpp Show resolved Hide resolved

Caball009 added 6 commits February 23, 2026 00:00

Updated comment for improved accuracy.

e834604

Removed temporary flags variable.

7a86f43

Improved comment.

51636b6

Added assertion for game slot transition.

fd39ae4

Added parentheses around assertion comment.

391bdf7

Synchronized comment with comment in PR description.

4302707

xezon reviewed Mar 2, 2026

View reviewed changes

Caball009 added 8 commits March 6, 2026 15:21

Merge branch 'main' into show_clients_with_sh_patch

7feddc7

Simplified return logic for 'setProductInfoStrings'.

ae11852

Set null-terminated flag.

1900b24

Simplified return logic for 'getProductInfoStrings'.

870bfd3

Changed 'createProductInfoMessage' to 'buildProductInfoMessage'.

759efbc

Set math crc field.

d53d412

Switched back from launch time to up time.

a8e1aa2

Fixed return logic for 'getProductInfoStrings'.

c4a34ce

greptile-apps bot reviewed Mar 6, 2026

View reviewed changes

Core/GameEngine/Source/GameNetwork/LANAPIhandlers.cpp Show resolved Hide resolved

Conversation

Caball009 commented Aug 1, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

TODO:

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

xezon left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

L3-M commented Oct 23, 2025

Uh oh!

Caball009 commented Nov 1, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Caball009 commented Nov 6, 2025

Uh oh!

Caball009 commented Nov 6, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

xezon left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Caball009 commented Nov 8, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Caball009 commented Nov 30, 2025

Uh oh!

Uh oh!

xezon left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

greptile-apps bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Caball009 commented Feb 5, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

xezon commented Feb 8, 2026

Uh oh!

Caball009 commented Feb 8, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

xezon commented Feb 13, 2026

Uh oh!

xezon left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Caball009 commented Aug 1, 2025 •

edited

Loading

Caball009 commented Nov 1, 2025 •

edited

Loading

Caball009 commented Nov 6, 2025 •

edited

Loading

Caball009 commented Nov 8, 2025 •

edited

Loading

Caball009 commented Feb 5, 2026 •

edited

Loading

Caball009 commented Feb 8, 2026 •

edited

Loading

Caball009 commented Feb 22, 2026 •

edited

Loading