Author Topic: 0.2.11 testing (renamed thread)  (Read 5947 times)

daniel.santos

  • Guest
0.2.11 testing (renamed thread)
« on: 13 December 2008, 11:23:32 »
Hello all.  I'm releasing a test build for 0.2.11 because I would like some people to try out the keymap functionality.  The win32 build is messed up right now (very slow, shift & control keys "stick", etc.) so please don't post bugs if you try that one.  But if you use linux and don't mind testing it out I would appreciate it.  Available here: http://glest.codemonger.org/downloads-t ... -test1.php (EDIT: Updated in svn already to fix a few bugs)

First off, the glestadv.ini file that comes with it only has the font definitions in it.  When you run it for the first time, it loads default values for any config setting that isn't specified.  Unfortunately, it doesn't detect when it had to do that yet, so you have to go into the options menu and change something to trigger it to write out the glestadv.ini.  The situation is similar for the keymap.ini file -- it doesn't need any of the settings, but it still fails to start (gives an error) if the file isn't at least present, so I've included an empty keymap.ini file.  At least that one will write out as soon as gae starts. (EDIT: I forgot to mention that I changed the names of almost all of the entries in the glestadv.ini file so they are all prefixed with a category, which makes it much easier to deal with now.  Also, you can just leave your old glestadv.ini file in place, make at least one change (back and forth will do) in the options menu and it will update it.)

So start it up and then exit and you should have a keymap.ini file with all of the default mappings.  Each command may have a primary and secondary mapping.  For instance, you can increase the speed with either the num pad plus key or the equals key on the regular keyboard (plus/equals).  
Code: [Select]
SpeedInc=Equals, KPPlusYou can use the modifiers "Shift", "Ctrl", "Alt" and "Meta" (I haven't tested the "Meta", but it's supposed to be the windows key) in addition to the key as follows:
Code: [Select]
SpeedInc=Shift+Equals, Ctrl+Shift+Alt+KPPlusThe following are valid commands although those in grey are not yet implemented (i.e., not "connected" to anything yet).
  • ChatAudienceAll
  • ChatAudienceTeam
  • ChatAudienceToggle
  • EnterChatMode
  • MenuMain
  • MenuQuit
  • MenuSave
  • MenuLoad
  • QuitNow
  • QuickSave
  • QuickLoad
  • PauseOn
  • PauseOff
  • PauseToggle
  • SpeedInc
  • SpeedDec
  • SpeedReset
  • NetworkStatusOn
  • NetworkStatusOff
  • NetworkStatusToggle
  • SaveScreenshot
  • CycleDisplayColor
  • CameraCycleMode
  • CameraZoomIn
  • CameraZoomOut
  • CameraZoomReset
  • CameraPitchUp ("free" camera mode only)
  • CameraPitchDown ("free" camera mode only)
  • CameraRotateLeft ("free" camera mode only)
  • CameraRotateRight ("free" camera mode only)
  • CameraAngleReset
  • CameraZoomAndAngleReset
  • CameraPosLeft
  • CameraPosRight
  • CameraPosUp
  • CameraPosDown
  • CameraGotoSelection
  • CameraGotoLastEvent
  • SelectNextIdleHarvester
  • SelectNextIdleBuilder
  • SelectNextIdleRepairer
  • SelectNextIdleWorker
  • SelectNextIdleRestorer
  • SelectNextIdleProducer
  • SelectNextProducer
  • SelectNextDamaged
  • SelectNextBuiltBuilding
  • SelectNextStore
  • Attack
  • Stop
  • Move
  • Replenish (heal/repair)
  • Guard
  • Follow
  • Patrol
The following are valid key specifications.  Not all keys are available on all operating systems and hardware.  When there's an actual UI, it will be easier because you can just press the key combination and GAE will capture it.  Until then, we'll have to specify the keys manually.
  • LButton
  • RButton
  • MButton
  • XButton1
  • XButton2
  • Backspace
  • Tab
  • Clear
  • Return
  • Pause
  • IMEKana
  • IMEHangul
  • IMEJunja
  • IMEFinal
  • IMEHanja
  • IMEKanji
  • Escape
  • IMEConvert
  • IMENonConvert
  • IMEAccept
  • IMEModeChange
  • IMEProcess
  • Space
  • Exclaim
  • QuoteDbl
  • Hash
  • Dollar
  • Percent
  • Ampersand
  • Quote
  • LeftParen
  • RightParen
  • Asterisk
  • Plus
  • Comma
  • Minus
  • Period
  • Slash
  • 0
  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • Colon
  • Semicolon
  • Less
  • Equals
  • Greater
  • Question
  • At
  • A
  • B
  • C
  • D
  • E
  • F
  • G
  • H
  • I
  • J
  • K
  • L
  • M
  • N
  • O
  • P
  • Q
  • R
  • S
  • T
  • U
  • V
  • W
  • X
  • Y
  • Z
  • LeftBracket
  • Backslash
  • RightBracket
  • Caret
  • Underscore
  • Backquote
  • Delete
  • KP0
  • KP1
  • KP2
  • KP3
  • KP4
  • KP5
  • KP6
  • KP7
  • KP8
  • KP9
  • KPPeriod
  • KPDivide
  • KPMultiply
  • KPMinus
  • KPPlus
  • KPEnter
  • KPEquals
  • Up
  • Down
  • Right
  • Left
  • Insert
  • Home
  • End
  • PageUp
  • PageDown
  • F1
  • F2
  • F3
  • F4
  • F5
  • F6
  • F7
  • F8
  • F9
  • F10
  • F11
  • F12
  • F13
  • F14
  • F15
  • F16
  • F17
  • F18
  • F19
  • F20
  • F21
  • F22
  • F23
  • F24
  • NumLock
  • CapsLock
  • ScrollLock
  • LShift
  • RShift
  • LCtrl
  • RCtrl
  • LAlt
  • RAlt
  • LMeta
  • RMeta
  • RSuper
  • LSuper
  • Mode
  • Compose
  • Help
  • Print
  • Sysreq
  • Break
  • Menu
  • Power
  • Euro
  • Undo
  • BrowserBack
  • BrowserForward
  • BrowserRefresh
  • BrowserStop
  • BrowserSearch
  • BrowserFavorites
  • BrowserHome
  • VolumeMute
  • VolumeDown
  • VolumeUp
  • MediaNext
  • MediaPrev
  • MediaStop
  • MediaPlayPause
  • LaunchMail
  • LaunchMediaSelect
  • LaunchApp1
  • LaunchApp2
  • Play
  • Zoom
Obviously, I wouldn't attempt to use any of the IME* keys or the power button (not sure why I included them).

daniel.santos

  • Guest
Re: 0.2.11-test1 (Key mapping)
« Reply #1 on: 13 December 2008, 11:52:13 »
Follow up: I've added a "MiscDebugKeys" setting to the glestadv.ini file so you can experiment and see exactly what key you are pressing (according to the key mapper). I also fixed mangled language strings.  It's checked into svn (rev 247) if interested.  Just ignore the fact that it's redundant with modifier keys (i.e., pressing left control yields "Ctrl+LCtrl")

@kukac@

  • Guest
Re: 0.2.11-test1 (Key mapping)
« Reply #2 on: 13 December 2008, 13:34:33 »
Daniel, what about keyboards, which are different from the English keyboard? On the Hungarian keyboard, the Z and the Y keys are changed, plus we have some more, like Í É Á Ű Ú Ő Ó Ü Ö, and things like colon and semicolon are on the same button (you have to press alt to reach semicolon, and shift for question mark).

daniel.santos

  • Guest
Re: 0.2.11-test1 (Key mapping)
« Reply #3 on: 13 December 2008, 14:04:01 »
There's a lot I'm not quite certain of yet, but the MiscDebugKeys setting should help.  In fact, I don't think that "Colon" will ever get used on my keyboard, because I too have to hold down shift for a colon, so the key combination would come out "Shift+Semicolon".  As for keys like Í É Á Å° Ú Ő Ó Ãœ Ö, SDL provided a lot of "world xxx" key codes that the windows interface doesn't.  For windows, it has a more complex system where it has a "packet" key code and another message comes later with the rest of the key information.  I didn't map to the "world xxx" codes, but I can, especially if that turns out to be the keys you are talking about that a lot of people have on their keyboards. (they are defined as SDLK_WORLD_0 - SDLK_WORLD_95).

If it turns out that with MiscDebugKeys on, pressing these keys results in no output, then we'll definitely need to map to them.  Most of the game doesn't support unicode very well, that may have to change.

@kukac@

  • Guest
Re: 0.2.11-test1 (Key mapping)
« Reply #4 on: 13 December 2008, 14:09:14 »
I will try the MiscDebugKeys tomorrow.

Err, a lot of things missing from the new version (like keymapping.ini, plus the new version of glestadv.ini contains only the letter type  :o )

daniel.santos

  • Guest
Re: 0.2.11-test1 (Key mapping)
« Reply #5 on: 13 December 2008, 14:36:49 »
Yea, this is just a test build.  I'll make a test2 build tomorrow as well (with the world0-95 keys in).  Also, the glestadv.ini will be filled out once you start gae and go into options and change something (I'll try to remember to do something about that as well).  I wanted to make it so that it could start even if the .ini files weren't there (and it would generate new ones) -- it's just not quite there yet.

@kukac@

  • Guest
Re: 0.2.11-test1 (Key mapping)
« Reply #6 on: 13 December 2008, 14:39:52 »
It needs a keymap.ini to start, which I don't have (would you attach it to the test1 download?).

titi

  • MegaGlest Team
  • Airship
  • ********
  • Posts: 4,239
    • View Profile
    • http://www.titusgames.de
Re: 0.2.11-test1 (Key mapping)
« Reply #7 on: 13 December 2008, 18:46:43 »
I cannot compile the current svn version ( branch 2.8.x )
I get some errors:

Code: [Select]
glest_game/../shared_lib/include/platform/input.h:273: Fehler: »SDL_BUTTON_X2« wurde in diesem Gültigkeitsbereich nicht definiert
shared_lib/sources/platform/input.cpp:50: Fehler: »SDL_BUTTON_X1« wurde in diesem Gültigkeitsbereich nicht definiert
glest_game/../shared_lib/include/platform/input.h:320: Fehler: »native2mb« wurde in diesem Gültigkeitsbereich nicht definiert
glest_game/facilities/pos_iterator.cpp:46: Fehler: »qsort« wurde in diesem Gültigkeitsbereich nicht definiert

Is there any chance to get around this?

Is this Version no longer compatible with ubunntu 6.06 (and it's old SDL version of course )?
I use this old ubuntu to compile because the version has an old glibc and so it will hopefully work on lots of linux versions out there.
Try Megaglest! Improved Engine / New factions / New tilesets / New maps / New scenarios

daniel.santos

  • Guest
Re: 0.2.11-test1 (Key mapping)
« Reply #8 on: 13 December 2008, 23:49:40 »
Actually titi, you're just helping me flush out bugs.  I can do a version check and define those two values if they aren't already present, thanks!  I'm going to try to get a "test2" out in a bit.

@kukac@, you can just do a "touch keymap.ini"

wciow

  • Behemoth
  • *******
  • Posts: 968
    • View Profile
Re: 0.2.11-test1 (Key mapping)
« Reply #9 on: 15 December 2008, 15:24:19 »
Probably a stupid question but can I bind keys from the command line?  i.e /bind MenuSave=F1
Check out my new Goblin faction - https://forum.megaglest.org/index.php?topic=9658.0

daniel.santos

  • Guest
Re: 0.2.11-test1 (Key mapping)
« Reply #10 on: 15 December 2008, 19:23:01 »
Not a stupid question at all, but no, you can't do it at the command line currently.  It makes me think of adding a console that comes down when you hit ~ though. :)  (Ahh, the good-ole Quake days!)

@kukac@

  • Guest
Re: 0.2.11-test1 (Key mapping)
« Reply #11 on: 15 December 2008, 19:35:09 »
I tried it, and it seems it has become a half-English half-Hungarian hybrid key mapping. Most of the keys work as normal Hungarian keys, like the Z and the Y, but several are recognized as English buttons, like Ö Ü Ó Ú Ő Ű Á É _?:-.,;*$ߤ÷×˝¨¸and maybe some others, plus it don't recognize í Í < at all, and I can't make ~ character in game (because on the English keyboard, it would be the same as 0 § Ö ˝, which makes the whole problem quite problematic... Oh why do humans have so many languages....)

titi

  • MegaGlest Team
  • Airship
  • ********
  • Posts: 4,239
    • View Profile
    • http://www.titusgames.de
Re: 0.2.11-test1 (Key mapping)
« Reply #12 on: 15 December 2008, 21:49:13 »
I didn't had time to test, but my boys did it. Here is what they told me:

1. connecting to a wrong ip causes glest to hang. There is a button that can be clicked, but nothing happens. Normal kill doesn't help, you have to kill -9 .
2. different keymap.ini files on server and clients are not allowed! Games crashes? stops? they don't tell me what happened exactly.
3. ingame crash: Here is waht they did:
-  they played a MP-game and saved it.
-  after this they reload the game and they play. The client has some "jumping" units but first everything was quite ok.
-  after a while the client starts a morph-command of the thor-statue ( to get a thor) . This command was only executed on the client NOT on the server !?!
  Some seconds later the game crashes with the message: Error cancelling upgrade, upgrade not found in upgrade manager"
  The error message tells something about an upgrade, so I'm not shure if it has something to do with the morph command.

Ok I will try MP now for my own to see what happens, but I think one man multiplayer testing is not very effective ....

UPDATE:
The game gets very fast completly out of sync!
Commands given on the server are NOT( or nearly never ) executed on the client!!
It's unplayable this way...
Commands are only executed by the updates the server sends to the client.

This bug is too worse, I will stop testing now.
Try Megaglest! Improved Engine / New factions / New tilesets / New maps / New scenarios

wciow

  • Behemoth
  • *******
  • Posts: 968
    • View Profile
Re: 0.2.11-test1 (Key mapping)
« Reply #13 on: 15 December 2008, 22:34:54 »
Well I don't know about the multiplayer problems but I can confirm that the new key mappings are working without any problems on win vista.  :)
Check out my new Goblin faction - https://forum.megaglest.org/index.php?topic=9658.0

daniel.santos

  • Guest
Re: 0.2.11-test1 (Key mapping)
« Reply #14 on: 15 December 2008, 22:48:10 »
Thank goodness for good news wciow!

titi,   :(  :(  :(  :(  :(  :(  :(  :(  :(  :(  :(  :(  :(  :(  :(  :(

If you had these problems, I would expect simlilar problems with 0.2.11.  I'll do some multiplayer testing and see what I can discover.  Thanks again to you and your sons.

daniel.santos

  • Guest
Re: 0.2.11-test1 (Key mapping)
« Reply #15 on: 15 December 2008, 23:18:31 »
titi,
1. OK, when testing before, I did find some type of problem that happened sometimes in the "crash handler", and network problems were always going there.  It's not supposed to crash IN the crash handler. :) ): throw runtime_error("Error canceling upgrade, upgrade not found in upgrade manager");  I'm glad they wrote it down or remember the message, that's VERY helpful!!!  Do your sons know why the upgrade was cancelled by the way?

titi

  • MegaGlest Team
  • Airship
  • ********
  • Posts: 4,239
    • View Profile
    • http://www.titusgames.de
Re: 0.2.11-test1 (Key mapping) (Note: 0.2.11 out now)
« Reply #16 on: 15 December 2008, 23:39:00 »
no they didn't do anything with an upgrade ( they say but I'm not shure here .) It's probabaly something the cpu player did.
When we test we always play coop vs CPUs and the computers screens next to each other to see every problem.


What about the commands from the server not executed on the client?



... and yes my boys simply wrote the error message on a piece of paper.
and sorry for all the bad news, but someone has to do this job  :( .
I don't know why I'm the only one who find these problems.
Try Megaglest! Improved Engine / New factions / New tilesets / New maps / New scenarios

daniel.santos

  • Guest
Re: 0.2.11-test1 (Key mapping) (Note: 0.2.11 out now)
« Reply #17 on: 16 December 2008, 00:22:04 »
w00t!!!!!!! I FOUND THE PROBLEM!!!!!!!!!!!

Sorry for the excitement.  Because I was doing testing and trying to generate problems before, I had the NetMinFullUpdateInterval set very low (like 4 seconds) so I wasn't seeing the missing commands from the server!  So it turns out that I forgot a stupid exclamation point!  Anyway, I also found a corruption issue with the new Command class I introduced in 0.2.10, but it's a very simple class so hopefully, this will all be working soon :)

Thanks again titi!

titi

  • MegaGlest Team
  • Airship
  • ********
  • Posts: 4,239
    • View Profile
    • http://www.titusgames.de
Re: 0.2.11-test1 (Key mapping) (Note: 0.2.11 out now)
« Reply #18 on: 16 December 2008, 00:57:06 »
Ha!  Good news!

But I have to sleep now. We will test more tomorrow. I checked out the current svn and everything builds.
So everything is ready to play and my little testers will do their job tomorrow again  :lol:
Try Megaglest! Improved Engine / New factions / New tilesets / New maps / New scenarios

daniel.santos

  • Guest
Re: 0.2.11-test1 (Key mapping) (Note: 0.2.11 out now)
« Reply #19 on: 16 December 2008, 08:04:12 »
hehe, awesome and sweet!! :)  Also, twice now, I've seen an annoying bug where I tell a builder on the client to start building and the building appears, but they just stand there, not actually "building" it, but on the server, they are building away and it will eventually get synchronized (I have the MinFullUpdateInterval set to 60 seconds now for my testing).  I know the basic issue, the builder is stuck in the "wait_for_server" skill, but that isn't supposed to last more than a few milliseconds (depending upon how bad lag is).  I've also seen some other minor issues, like units dying on the client because an update said they were dead caused them to not play out their death animation, they just were suddenly on the ground.  When playing with sending lots of commands, I had a unit that was in an incorrect location by about 5 cell spaces, it was interesting, because he obeyed the commands, but was doing it in different places on the client and server, but that eventually synchronized, so that's good.

So, from now on, I'm going to do testing where I play a cooperative game so I can test better.  I checked in these changes, but I haven't made a new release yet as I want to test more.  Also, I added some more garbage to the client and server logs when debug is on, but I don't think I need it anymore (I might leave it in or change it some, I just hate to bloat the code).  The current revision is 255 if you want to check that out and give it a try.  Thanks!!!

daniel.santos

  • Guest
Re: 0.2.11 testing (renamed thread)
« Reply #20 on: 16 December 2008, 17:52:55 »
ok, error when canceling upgrade is still present, but I have captured it in debug, so just to let you know, it's still not stable, hopefully I'll have something better soon.

titi

  • MegaGlest Team
  • Airship
  • ********
  • Posts: 4,239
    • View Profile
    • http://www.titusgames.de
Re: 0.2.11 testing (renamed thread)
« Reply #21 on: 16 December 2008, 18:30:51 »
We tried to play 2 times but ...

we had the same error like you:
Error canceling upgrade, upgrade not found in upgrade manager :/


The first time we played there was another hard crash:
Some seconds after the multiplayer game starts on the client suddenly all units disappear and there are only shadows visible.
3-5 seconds later we get a crash with a gae-crash.txt
Code: [Select]
Crash
Version: Advanced Engine v0.2.11-test2
Time: Mon Dec 15 18:41:07 2008
Description: SIGSEGV: address not mapped to object
Address: 0xaf03503c
Backtrace:
./glestadv [0x81635d9]
[0xffffe440]
./glestadv [0x814cdaf]
./glestadv [0x8151623]
./glestadv [0x815b32b]
./glestadv [0x806f4eb]
./glestadv [0x809f7de]
./glestadv [0x80bb5d9]
./glestadv [0x809fcd0]
./glestadv [0x809aee8]
./glestadv [0x809b075]
/lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xd2) [0xb767dea2]
./glestadv(__gxx_personality_v0+0xe5) [0x804e871]

=======================
Crash
Version: Advanced Engine v0.2.11
Time: Tue Dec 16 19:17:13 2008
Description: SIGABRT:
Address: 0x2d63
Backtrace:
./glestadv [0x81648b9]
[0xffffe440]
[0xffffe410]
/lib/tls/i686/cmov/libc.so.6(gsignal+0x51) [0xb76009a1]
/lib/tls/i686/cmov/libc.so.6(abort+0xe9) [0xb76022b9]
/usr/lib/libGLcore.so.1 [0xb70f2d85]

=======================



Something that really irritates me is the fact that the units on server and client always get out of sync much much faster then in original glest!
This is corrected by gaes update functions, but this causes unit jumping and sometimes really strange behaviour.
On the same platform the game should not get out of sync after some seconds ( 1 minute )!
Try Megaglest! Improved Engine / New factions / New tilesets / New maps / New scenarios

daniel.santos

  • Guest
Re: 0.2.11 testing (renamed thread)
« Reply #22 on: 16 December 2008, 23:36:47 »
Platform shouldn't matter, I've long since solved platform neutrality problems in GAE, it shouldn't be able to get this out of sync at all.  Part of the problem is the delay enforced on commands, they have to wait up to 250 milliseconds before they are transmitted because it waits for a key frame.  I will look into some more mechanisms to solve the sync conditions after the real problems are fixed (i.e., commands being sent by server properly, no crashes, etc.).  It may not be as clear from a user perspective, but the serious problems are those of stability.  Issues like game exiting for trivial reasons (wrong IP address entered, etc.) are easily solvable, but stability issues can be very tough.  However, I'm sure that 80% to 90% of the "sync" problems are due to some other bug somewhere, because as GAE is now, it should actually be more in sync than the original Glest.

Part of the problem sometimes has been that I hadn't understood the "contracts" between objects at all times, it's been a discovery and learning process as I've gone.  Then, I have changed some of these contracts in GAE.  I think I'm going to write up a specification that lays out these contracts between objects (classes of Unit, World, Game, UnitUpdater, UpgradeManager, etc.) both to make sure I remember and to help others who want to start programming GAE.  After those internal contracts are understood, then there's the entire networking layer on top of it, as client and server behave and interact differently.

Finally, I'm going to do an 0.2.12 that adds peer-to-peer communications so that commands created on one client are transmitted directly to all other clients (for games with 3+ players) rather than being routed through the server.  I'm also going to change the way commands are transmitted (on both clients & server) so that they are sent immediately, (in each world frame, which is currently 40 times a second) rather than only once per 250 milliseconds.  I'll still have the games sync on every "key frame" -- this essentially causes clients to wait for the server to tell them to proceed, assuring that clients are never in a frame ahead of the server.  Currently, key frames occur every 10 "world frames".  When testing that, I should be able to isolate any other game sync issues that arise.  I may also revamp the way world and rendering frames are executed.  Currently, animations only update a max of 40 times per second (in the world frames), even though all of the interpolation required to update them more is done during rendering.  By moving the update of "animation progress" to rendering, it can make animations smoother without extra costs.  But the main problem is that the world frames lag on slower computers, so I want to see how that can be fixed.  Perhaps after moving animation progress to the rendering frames, I can safely lower world frames to something like 15 times per second instead of 40, especially if I allow "excess" time spent executing a skill to be passed to the next skill's execution (that's complicated, maybe I shouldn't explain it :) )

Anyway, thanks titi, I'll post when I get something else out.

titi

  • MegaGlest Team
  • Airship
  • ********
  • Posts: 4,239
    • View Profile
    • http://www.titusgames.de
Re: 0.2.11 testing (renamed thread)
« Reply #23 on: 17 December 2008, 00:02:01 »
I don't think that its a good idea to let the clients talk to each other. I think its better the simple way it is done now.
Client server architectures are much more easy to implement and to debug! .
Everyone(including the server) sends his commands to the server. The server distributes a received command to every client( including himself) to be executed with the next command(key) frame.
So everyone should execute given commands in the same order at the same time(frame) .
This is also good for debugging because you can log every executed command on the clients(including the server).
these logfiles have to be everywhere the same. If they are different there is a problem.
This command distribution should happen for every key frame even if it doesn't contain any command ( or update commands )!
If a command doesn't reach the client in time, he has to wait!
These commands include gaes update commands which should be executed on every client too(including the server who send them)
All this should keep the game in sync( nearly) because all commands should be executed the same way on every client. ( I 'm ignoring floatingpoint problems here with different OS/compilers ...)

If one client is too slow, I think there is a way to speed up its execution ( that's what martino tried to do as far as i remember). A client is slow if he has received his commands for the next 2 or more key frames while working on the first one. Network Lag is a big problem here but this is another problem.
 
All this is "simple" and the clients don't have to setup their routers (open ports and all these problems)  

What do you think of this?
Try Megaglest! Improved Engine / New factions / New tilesets / New maps / New scenarios

titi

  • MegaGlest Team
  • Airship
  • ********
  • Posts: 4,239
    • View Profile
    • http://www.titusgames.de
Re: 0.2.11 testing (renamed thread)
« Reply #24 on: 17 December 2008, 00:18:55 »
Another problem I see with the "all clients send comands directly" is the following:
If I want to play with one of my boys and someone else from the inet we cannot do it. We are behind a router with NAT. You can forward the glest communication port to one computer but not to two!.So only one of us can play in Inet. The other one is hidden for someone from outside our private network :( .
Try Megaglest! Improved Engine / New factions / New tilesets / New maps / New scenarios

 

anything