MegaGlest Forum
Archives (read only) => Vanilla Glest => Multiplayer => Topic started by: AF on 5 February 2008, 12:32:20
-
Since official support for external lobbies will not be supported, despite whatever controversy and issues with the evaluation of that decision, this does leave somewhat of a gap.
Namely if the glest team are not going to allow the community to pick up the lobby problem for them while they're figuring out an official method, what is happening? Is there a lobby client being developed? Is there an internal lobby client+server system being developed?
Or are glest players forever doomed to the current IRC channel + manually starting up and typing IP addresses? How can this situation be good for anyone? This is pre-1997 style multilayer matching.
Also, on a side note, is there any chance of reconsidering the offer of using AFLobby that actually involves evaluating AFLobby and not a buggy rival program?
-
I want af's lobby as the official one.
Irc to set up games is not cool
-
I would like something browser based... many computer-newbs don't know where to find a good irc program
Right now i'm in irc and no one is here...wait, someone just got here
Anyway, the irc is slightly cumbersome... what is the original thread about AF lobby again?
-
Wouldn't be surprised if that thread was simply removed.
Someone here has a strange attitude towards people who try to improve the game.
-
If AF's own words "Nobody uses his Spring lobby because it doesn't work". Nobody seems to like him in the Spring community either. So I can not take seriously a guy that doesn't have anything better to do with his life that spending it flaming other people's boards with statements like "I suggest you cooperate", "Your project will die" and similar (and that goes for you Duke too).
As a final statement, I will ignore any further flaming posts in this board, I won't be even be botherer removing them.
Kind Regards.
Martiño.
-
erm.,
Firstly, my lobby wasn't being used because of a major protocol change that caused a crash bug and I hadn't released a fix yet.
Secondly I do not appreciate the slur on my name, there are plenty fo people who like me on the spring forums, there just happens to be a small minority fo flamers who would say otherwise. If you actually bothered to check up on what they've posted on the spring forums you'll be sure to find such things as flame threads, pictures of naked women and goats, among other things, so I would not take them seriously.
Your also forgetting that spring lobby offered their services too.
-
You obviously haven't witnessed flame or you wouldn't call this flame. At the worst It's a sarcastic comment about my experience. Flame is pointlessly offensive and usually without reason.
I try to avoid that, as does AF. He's just direct (thats actually what I like about him) and I tend to get harsh if I'm offended, thats all.
And the fact that you are missqouting means that you either are doing it liberately or haven't understood what he meant. (or didn't even bother)
Scanning a text is not the best way to catch it's meaning, you have to actually read it.
Anyway, back to topic: here is the original thread:
http://http://www.glest.org/glest_board/viewtopic.php?t=3065&postdays=0&postorder=asc&start=15
daniel.santos said that he could create the interface AF requested.
Was this offer pursued?
-
Hailstone started and completed a working interface under linux, but hsi working copy lagged behind and didn't incorporate changes made int he official svn afterwards. He hasn't shown much interest lately either and I have been snowed under with university coursework and prior commitments to my 2 main projects to figure out where the glest compilation instructions went wrong and then make my own changes.
santos' email was never followed up by him and I've received no details since.
It may also be of note that my original question has still not been answered, your reply martino did not say if you or another glest developer had an internal lobby in development, or if an external lobby from someone else was being lined up, or if you had no intention of a lobby of any kind.
Its also worth mentioning that at the moment XFire or GameSpy Arcade style support for glest is impossible for the same reason AFLobby and Spring Lobby cannot be used. Glest provides no API for starting multiplayer games from another program, despite it being a relatively simple thing to do looking at what spring did, and what hailstone managed, or that nearly every RTS game made in the last 10 years does it too.
If you wish to see true trolling, I would advise you go to the spring forums and ask about smug clan or ranetti.. Better yet, go to 4chan, I assure you, if you are willing to act hostile to a blunt message, 4chan will make you want to commit suicide. Criticism != trolling, and treating it as such gives the false impression that this project is ruled by an oversensitive tyranical oligarchy, which is hardly going to attract interest.
-
and treating it as such gives the false impression that this project is ruled by an oversensitive tyranical oligarchy, which is hardly going to attract interest.
It gets closer, but even this is no real flame material.
But I have to admit that I actually got that impression mentioned above.
-
hmmm, in hindsight maybe I should make clear that I'm not accusing glest devs of this, I'm just saying it 'looks' like it.
-
I'm just saying it 'looks' like it.
Actually that's more or less the definition of accusation I think (;
But it can't be helped I suppose, you can only judge a person based on their actions you witnessed, or to a lesser extend on those that are reported through a trustworthy source.
Since negative impressions usually stick about five times better then positive ones this is quite difficult a task.
-
Well if I said it 'was' or 'is' the case, that would be an accusation, instead I said it 'looks' or 'appears' to be the case, speaking in terms of impressions not actualities.
-
Since Daniel is doing a 'community' sort of Glest I'd be happy to work with him to incorporate what I've already done if that's what people want.
As a side note. I've been checking out Bos Wars (http://http://www.boswars.org/) to get some idea as to how they operate and some ideas they have put in their RTS game.
-
Since Daniel is doing a 'community' sort of Glest I'd be happy to work with him to incorporate what I've already done if that's what people want.
Well I think the project that Daniel and I are working on right now is almost certain to fork from the the original Glest. This is because Martinho and the other 'official' Glest devs don't have the time to officially check/support/update everyone elses code mods (which is understandable).
I think Daniel already mentioned somewhere that he incorporated your screen switch code into the patch so I'm sure any help would be appreciated and I'm sure he'll clarify this when he gets back from holidays.
-
If it's going to happen it had better happen soon.
-
If it's going to happen it had better happen soon.
Why?
-
While I agree that the current IP is nearly useless, you must understand what the Glest team have said. They never had, and never will intergrate other peoples version of Glests code with the official version to keep Glest reliable and bug free. This makes complete sence.
You must also realise that multiplayer has really only just been released and that Im sure in due time there will be a lobby. But the Glest people do have lives and cannot devote all there time to Glest.
-
I personally would employ what the open source community call peer review rather than blind inclusion of contributed code.So that way of explaining it is not a very tactful way of putting it, after all the glest developers and their code base is not exactly picturesque either, so why take the time to throw pot-shots at the open source community?
-
so why take the time to throw pot-shots at the open source community?
I don't think that was the intention. More that the Glest team wants to keep a version that is personal for them and to integrate other peoples code would take time they haven't got. Now you could be thinking that other people could do it, but then the Glest team would no longer have direct control and it would no longer be a small team.
Just my thoughts anyway. I think it is better to create a fork and respect the wishes of the Glest team. Possibly if they like a feature in a fork they could add it to the mainstream later.
I also agree with peer review.
-
AF i don't think the glest team is really indifferent(?right word) towards your work or general codework done by others. See at the new macintosh portation: There martiño is really interested to take over improvements of the code. Or see the "official binary" by titi. I don't think you should complain they don't care about other people's work, its just that programming is quite complex and so everything must be well planned.(i must add that i have no real understanding of programming ...)
-
No mention of 'indifference' or 'ignorance' was made, nor did I accuse anyone of anything.
Also please make clear the difference between describing impressions and an accusation. If I say someone looks sad, that's not the same as me saying someone is sad.
I don't think that was the intention. More that the Glest team wants to keep a version that is personal for them and to integrate other peoples code would take time they haven't got. Now you could be thinking that other people could do it, but then the Glest team would no longer have direct control and it would no longer be a small team.
Then it would be far easier to type out the following, as well as less ambiguous and less likely to cause offence because it doesn't imply that everyone else has bad coding skills:
"The glest team want to retain control over their engine, if you wish to make modifications freely you can distribute your own version."
It's much better than
"They never had, and never will integrate other peoples version of Glests code with the official version to keep Glest reliable and bug free."
Which makes little sense at all unless you assume the glest developers are perfect coders and everyone else isn't.
While I can see that you guys would be just as happy if the process of building a glest with command line parameters took a year or two, I do not have that luxury, and I have two rival programs to compete against. Right now a working lobby would be a huge boost to any glest fork regardless of whose lobby it was and were it came from, but suffice to say, if you did actually take a year, me and my project would suffer greatly from that and the idea of 'glest support in the lobby' would become a running joke in the spring community. That initial positive wave of opinion in the spring community has dispersed and now people are starting to doubt that multiple engine support is actually possible to do despite the evidence to the contrary.
One note, I would say the glest developers are indifferent to my lobby. Why? Because martinho said so in this thread. Since the glest developers cannot even give a straight answer to the lobby problem in general ( e.g. XFire, or internal glest developments) never mind the specific case of spring lobbies, I do not believe there is any avenue for discussion with them right now, not that I wouldn't discuss things should they express an interest,
Any discussion I do have will be with you the glest community, as the community has expressed and interest in lobby matchmakers in general, and ultimately it is the players who will make the decision by switching to the lobby of choice or continuing to use the IRC channels.
-
AF: say exactly what you need in terms of command line parameters and we might do it.
-
glest.exe -h -p <portnumber>
glest.exe -j <hostIP> -p <portnumber>
The ability to specify a username/nickname parameter would also be useful but it isn't necessary, and it has not been discussed either.
-
AF, how does your game lobby work? Do we need a masterserver or is it like peertopeer?
But if its peertopeer there can't be any stats? On the other hand it would make glest independent from one individual machine which means it is more stable.
-
I hope that .exe is just an example? would be sad if it worked in windows only.
-
The lobby system itself is not maintained by any one person, he spring lobby system has TASServer at the core, and there are currently 3 actively maintained lobby clients, 2 abandoned clients, and a flotilla of bots and autohosts.
As such you would not need to do anything yourselves once you've implemented the command line parameters, users would just use the same server as the spring people, and the lobby would handle everything.
Most of the battle protocols are ignored when starting glest games as they do not apply, thus implementing lest support in a dedicated lobby would be easy.
-
Sounds not so bad. So there are already servers which would be able to handle with glest-multiplayer.
So there's the lobbyclient which works as a seperate executable and it is used to connect people to a game.
Some choose to host and other join listed games.
If everything finds together the little programm launches Glest with parameters so that glest knows the connection details.
Some issues are still left: I'm not sure about the server-client-syncronisation; The commandline can't tell glest which gamedetails (how many playerslots/cpus also map factions techtree...) should be used. If the server adjusts this in Glest there is the problem that Glest cannot tell the lobbyprogramm; means the clients who want to join don't know when they are able to join, if there are open slots for them.
But nevertheless the lobbyprogramm could work like this: The host does everything by hand and the game is ready to launch. Then the lobbyprogramm tells the details to the players which want to join. Would still be more comfortable.
If Glest gets some more parameters it could also be possible to adjust everything in the lobbyprogramm, the server is launched some seconds before the clients and then the server starts when everybody is in.
//EDIT: a simple idea how to implement serverdetail(map/players/factions) into glest could be via a textfile; the lobbyclient writes everything into a file before it launches glest and then glest automatically uses this settings to launch the server. A additional parameter in the -host commandline could turn this on/off.
-
I've implemented some basic command line options for Glest:
glest => starts game normally
glest -server => starts the game in the new game screen with all slots open
glest -client ServerIp => starts the game in the join game screen and connects to the server specified
The port is always 61357.
Everything is in the svn, we will include it in the next version, I hope this is the end of the lobby wars.
-
I will implement the changes into aflobby.
On a side note, the reason port numbers were originally part of the request was for NAT issues and allowing users to specify custom ports or make use of such things as holepunching.