warzone2100/newnet.TODO

*********
* Notes *
*********
GAME_CHECK_DROID shouldn't do anything, unless the game is out of synch, now. And should print, if the droid isn't in _exactly_ the same place on all screens.
Lot of other messages shouldn't do anything, either.
GAME_BUILD should fix the structure ID of the structure being built. GAME_BUILD and GAME_BUILDFINISHED shouldn't do anything other than that, unless the game is out of synch, now.

**************
* Everywhere *
**************
Remove bMultiPlayer and turnOffMultiMsg() hacks, cleaning up any duplicate or dead code on the way.

Sanity check all use of selectedPlayer. It should be used for deciding how the interface works, not for changing the game state in random places for no good reason...

*************
* Somewhere *
*************
The GAME_FEATUREDEST should be turned into just synch messages, and features should be removed as soon as destroyed.

Fix production completed and research completed sounds.

Fix destroying droid templates. Can probably go out of synch, if destroying a template just before the droid is produced, due to trying to produce an unknown type of droid (so the other players would see the new droid, but the droid owner wouldn't).

********************
* src/structure.c: *
********************
Clean up structPlaceDroid.

Check what validLocation is used for, make sure that putting delivery points under locations where buildings are going to be placed can't cause synch errors.

In destroyStruct, check that psTile->illumination can't affect the game state. Or make it consistent, so all players see the same scorch marks.

*****************
* src/keybind.c *
*****************
Fix kf_CloneSelected.

*************
* src/hci.c *
*************
Check that this file doesn't modify droids directly...

Make modified code for placing structures in debug mode less hackish.

******************
* src/research.c *
******************
Check the researchResult function. Should a message be sent there, and if so, then should wait for it?

***************
* src/order.c *
***************
Just before transporterAddDroid, it changes the droid state directly. Should either not test for selectedPlayer, or should not change the droid state directly. See also: unloadTransporter.

*********************
* src/scriptfuncs.c *
*********************
Sanity check that this file doesn't do anything stupid, such as modifying the game state directly.
newnet: Synchronise build orders, against the will of whomever that wrote "//don't send BUILD orders in multiplayer". Disarm the GAME_BUILD and GAME_BUILDFINISHED messages, which are now for synch only. 2010-02-22 13:41:56 -08:00			`*********`
			`* Notes *`
			`*********`
			`GAME_CHECK_DROID shouldn't do anything, unless the game is out of synch, now. And should print, if the droid isn't in _exactly_ the same place on all screens.`
newnet: Make GAME_LASSAT synchronised. 2010-02-26 06:43:37 -08:00			`Lot of other messages shouldn't do anything, either.`
newnet: Synchronise build orders, against the will of whomever that wrote "//don't send BUILD orders in multiplayer". Disarm the GAME_BUILD and GAME_BUILDFINISHED messages, which are now for synch only. 2010-02-22 13:41:56 -08:00			`GAME_BUILD should fix the structure ID of the structure being built. GAME_BUILD and GAME_BUILDFINISHED shouldn't do anything other than that, unless the game is out of synch, now.`

newnet: Make GAME_DROID synchronised, now possible to build lots of trucks and stuff without going out of synch. 2010-02-22 04:29:28 -08:00			`**************`
			`* Everywhere *`
			`**************`
			`Remove bMultiPlayer and turnOffMultiMsg() hacks, cleaning up any duplicate or dead code on the way.`

newnet: Fix some selectedPlayer abuse. 2010-02-23 00:13:13 -08:00			`Sanity check all use of selectedPlayer. It should be used for deciding how the interface works, not for changing the game state in random places for no good reason...`

newnet: Synchronise GAME_STRUCTDEST and GAME_FEATUREDEST. 2010-02-22 14:17:10 -08:00			`*************`
			`* Somewhere *`
			`*************`
newnet: Don't wait for GAME_DROIDDEST or GAME_STRUCTDEST before destroying, since all players will do it at the same time, anyway. GAME_DROIDDEST and GAME_STRUCTDEST are a bit redundant, now... (Except for debug mode use...) 2010-02-26 04:03:29 -08:00			`The GAME_FEATUREDEST should be turned into just synch messages, and features should be removed as soon as destroyed.`
newnet: Synchronise GAME_STRUCTDEST and GAME_FEATUREDEST. 2010-02-22 14:17:10 -08:00
newnet: Try to synch GAME_RESEARCHSTATUS and GAME_RESEARCH. 2010-02-22 16:18:22 -08:00			`Fix production completed and research completed sounds.`

newnet: Make GAME_TEMPLATEDEST synchronised. Hope nothing explodes. Wouldn't surprise me, if something exploded, though. 2010-02-25 11:46:27 -08:00			`Fix destroying droid templates. Can probably go out of synch, if destroying a template just before the droid is produced, due to trying to produce an unknown type of droid (so the other players would see the new droid, but the droid owner wouldn't).`

newnet: Make GAME_DROID synchronised, now possible to build lots of trucks and stuff without going out of synch. 2010-02-22 04:29:28 -08:00			`********************`
			`* src/structure.c: *`
			`********************`
newnet: Mark GAME_DEMOLISH done, since it had no effect on the game either way. Remove GAME_SECONDARY_ALL, which wasn't useful for anything. 2010-02-26 04:41:54 -08:00			`Clean up structPlaceDroid.`
newnet: Make GAME_DROID synchronised, now possible to build lots of trucks and stuff without going out of synch. 2010-02-22 04:29:28 -08:00
newnet: Fix research causing power desynch, and due to power usage being random, random number desynch as well. 2010-02-23 00:58:19 -08:00			`Check what validLocation is used for, make sure that putting delivery points under locations where buildings are going to be placed can't cause synch errors.`

			`In destroyStruct, check that psTile->illumination can't affect the game state. Or make it consistent, so all players see the same scorch marks.`

newnet: Make GAME_DROID synchronised, now possible to build lots of trucks and stuff without going out of synch. 2010-02-22 04:29:28 -08:00			`*****************`
			`* src/keybind.c *`
			`*****************`
			`Fix kf_CloneSelected.`
newnet: Synchronise build orders, against the will of whomever that wrote "//don't send BUILD orders in multiplayer". Disarm the GAME_BUILD and GAME_BUILDFINISHED messages, which are now for synch only. 2010-02-22 13:41:56 -08:00
			`*************`
			`* src/hci.c *`
			`*************`
			`Check that this file doesn't modify droids directly...`

newnet: Make GAME_LASSAT synchronised. 2010-02-26 06:43:37 -08:00			`Make modified code for placing structures in debug mode less hackish.`

newnet: Try to synch GAME_RESEARCHSTATUS and GAME_RESEARCH. 2010-02-22 16:18:22 -08:00			`******************`
			`* src/research.c *`
			`******************`
newnet: Fix some selectedPlayer abuse. 2010-02-23 00:13:13 -08:00			`Check the researchResult function. Should a message be sent there, and if so, then should wait for it?`

			`***************`
			`* src/order.c *`
			`***************`
			`Just before transporterAddDroid, it changes the droid state directly. Should either not test for selectedPlayer, or should not change the droid state directly. See also: unloadTransporter.`
newnet: Fix production queues. Rename quantity variable which was used for two completely different things in a confusing manner. Disable stopping/resuming unsynchronised production and research. 2010-02-23 09:43:18 -08:00
			`*********************`
			`* src/scriptfuncs.c *`
			`*********************`
			`Sanity check that this file doesn't do anything stupid, such as modifying the game state directly.`