Dragonshard Multiplayer

[Sharpe]

Thread Subtitle: Now we know where Atari skimped when they hustled this beast out the door.

Alternate Subtitle: Gamespy is teh SuQ!

I was going to start a thread about Dragonshard strategies to both gush about how great and elegant the game design is and swap some builds and ideas, but first I have to say: the current Gamespy multiplayer sucks the giant weenie. I mean its really bad, worse than anything I've seen since the emergence of Battle.net and Gamespy.

First off the autopatcher isn't working fully: it updates players to version 1.00.0004 which is not the most recent. The full update is 1.01.0004 which you have to manually download from Atari. If your versions don't match you can't join games etc. Confusion reigns.

Then the Gamespy in-game interface is just absolute crap: it sets your game to use the map selected under your "Player Profile" which is accessible only from the main menu (ie you cannot change this while in Gamespy unless you log all the way out to the main menu). The default is a 2 player map and if you create a game with a 2 player map default setting only 2 players can ever be in your game --- *even if you upgrade to a larger map while in the multiplayer game setup screen* - for the past couple of nights only 1v1 games were being played for this reason. Now people are figuring it out but the interface is just asinine -- to change your player settings you have to log out to the main menu? And although the in-game screen lets you change maps it doesnt open up more online spots for players? (To make it worse the apparent number of open slots in the game lobby will reflect the map but the default number still controls the number that can join). Just based on that the multiplayer code is clearly a kludge.

But the real capper is this: the network code is a crippled, club footed Galapagos Tortoise on Vicodin. I mean the code is slooooooooooooooooow. Every game I've tried that was more than 1v1 ran like a lame dog. And, if you allow soldiers, once anyone gets a decent sized army, the game will slow to a literal crawl and often freeze or lockout players.

People were playing games with soldiers turned off (a major loss of game feature IMO) and still the sucker runs like poop.

It's weird b/c the game is decently stable and although a little slow (laggy) 1v1 quite playable. And I believe the game is playable 2v2 or 3v3 on a LAN. But on gamespy, if you want team vs team games, well I dunno.

I *really* like this game but team vs team online is IMO the core of good RTS gaming and right now I haven't been able to have a complete team vs team game online, despite almost a dozen tries.

What are other folks experiences?

[Merakon]

That's sad. This game wasn't on my radar until Tom's review. But since the most joy I get out of RTS games is multiplayer, it will be hard to justify picking this up until they fix the multiplayer problems.

I hope they patch it, or there is some sort of configuration solution for you.

[SlyFrog]

Ah, along with Rise of Nations and many other titles, another wonderful game fucked up the ass by GameSpy.

That's "GameSpy - Pissing on the Legs of Gamers for Over 10 Years!"

[Reed]

Why is it GameSpy's fault and not the publisher/developer? They are the one that chose to use it for their multiplayer.

All the big publishers seriously need to start up their own Battle.net already.

[SlyFrog]

Why is it GameSpy's fault and not the publisher/developer? They are the one that chose to use it for their multiplayer.

All the big publishers seriously need to start up their own Battle.net already.

Oh, there's plenty of blame to go around - I was merely focusing on GameSpy here. But it's hard to say, "Yeah, their product sucks ass, but it's not their fault, it's the publisher's because they used it."

It can be both, but in no way can GameSpy avoid blame. They have a history of shitty matchmaking that's older than Grandma's quilt (okay, it's at least a few years old).

[McBain]

Gamespy matching service = kiss of death for RTS games.

See: Dawn of War.

I really, really, really hate Gamespy.

[Ben Sones]

Everyone does. I can't fathom why any publisher/developer ever uses their matching software.

[Reed]

It can be both, but in no way can GameSpy avoid blame. They have a history of shitty matchmaking that's older than Grandma's quilt (okay, it's at least a few years old).
Right, and that makes it all the worse that publishers continue to sign on. It's ridiculous.

So why should GameSpy be concerned with improving when everyone continues to line up anyway? They need some competition, badly.

[Jakub]

Everyone does. I can't fathom why any publisher/developer ever uses their matching software.
Because it's easy, and simpler than trying to set up your own Battle.net-style network (which, btw, haven't proven all that successful in the first place other than Bnet itself and perhaps the old Boneyards).

As for the entire mess, I'm more inclined to blame Atari. Honestly, let's look at their track record. It's a miracle that there's more than one patch for Dragonshard in the first place.

[Sharpe]

Has anyone managed to get a more than 2-player game going online? I've now tried about 20 times and although we've managed to start some, we've not yet managed to finish a 4 player. Also, midway through tonight's efforts, everyone who was online became unable to join any games, at all. We all tried collectively logging out and back in, no luck.

Multiplayer right now is like this: go online, see a game, try to join. About 3/4 of the time, you cannot connect. If you do connect, most of the time the game is limited to 2 players only. You can choose to play a 1v1 which actually runs decently. If you want a bigger game you have to keep trying. Out of 20+ tries, I've managed to get into 3 2v2s. In one case we got pretty far into the game, with the outcome largely decided, when the thing locked up and froze. In the other 2 cases we got partway through the game (with MAJOR lag) and then ended up losing players to disconnection. Once players started to drop, things got unstable and eventually all players were disconnected. (I believe this latter problem was out of synch errors, but we didn't get any specific messagess).

Bottom line is: I can play 1v1 OK, and can play 2 humans vs a number of computers OK also. But anything larger is essentially F'ed at this point.

I'm still getting some mileage out of skirmish due to the really interesting and well done game design, but the single player campaigns are short (and its sort of incomplete given that one faction got totally stiffed - there is no campaign at all for the 3rd faction :( ), and the limited selection of skirmish maps suitable for small games is going to burn me out soonish. It's really too bad, the devs managed to do something at least moderately innovative and very elegant with the tired RTS archetype and yet the best part of the game, team based online multiplayer, is pretty damn close to non-functional right now.

I guess this is my first encounter with the cheese of Atari. Feh.

[Jack Black]

I guess this is my first encounter with the cheese of Atari. Feh.

Welcome to the disappointment train. Choo Choo.

Seriously. Atari has such a bad track record that everything released by them I expect it to not work for the first 2 patches. Then usually the bugs/issues will be cleared up to not notice them more than the game.

[Luke M]

The state of Dragonshard multiplayer according to this thread really bothers me. I was so looking forward to saving my pennies and getting this game, but without solid multiplayer support, I'm not so sure if I should bother. What a waste of a great design.

Here's what I want to know: how come these multiplayer problems haven't come to light in any reviews? Tom mentions that the game has "some sync errors" and bugs in his review, but nothing so serious as the issues enumerated in this thread. How many online games have these reviewers tried, anyway? How'd they managed to bypass all the typical Gamespy lobby crap and actually play a game with more than 2 players? I certainly am not trying to make any accusations, but, even still, I'm a little suspicious of the disparity between reviewer and customer experiences here.

[Tom Chick]

My review was submitted before the game went live, and I'd only tried online games 1v1. As Dan mentioned, they seem to work fine (he and I even had a few games the other night).

I've played 3v3 games on a LAN with no problems, but as I posted in another thread, I don't think it works particularly well as a team based RTS, mainly because there's no way to share resources and xp isn't split after a battle. These are pretty sad oversights, considering how well designed Dragonshard is otherwise.

As for the sync errors I mentioned, they've apparently been cleared up in the day zero patch.

It is pretty galling that there's no option to connect via direct IP. If I'm not mistaken, don't many games with Gamespy support also work direct IP? I wonder why that didn't happen here.

-Tom

[wildpokerman]

My review was submitted before the game went live, and I'd only tried online games 1v1. As Dan mentioned, they seem to work fine (he and I even had a few games the other night).

I've played 3v3 games on a LAN with no problems, but as I posted in another thread, I don't think it works particularly well as a team based RTS, mainly because there's no way to share resources and xp isn't split after a battle. These are pretty sad oversights, considering how well designed Dragonshard is otherwise.

As for the sync errors I mentioned, they've apparently been cleared up in the day zero patch.

It is pretty galling that there's no option to connect via direct IP. If I'm not mistaken, don't many games with Gamespy support also work direct IP? I wonder why that didn't happen here.

-Tom

I know you writers and gaming press people need to get paid and getting paid means people reading your review on release day but maybe some of you should think about reviewing the multiplayer interface or updating your reviews after the games go live.

I always laugh my ass off at all the high scores you all give to RTS games after playing the single player when come to find out it's a low scoring game because the multiplayer just doesn't work.

Seriously I don't know why you all give RTS games such a huge pass on the mutliplayer. No other game type gets away with this, would you give a MMORPG a high score and call it a review if you could play before the public can? If you play a game before it gets on store shelves and don't test the mutliplayer then what you're writing is a preview with a score, get honest already. At least leave yourself a big loophole and put it in the publishing contract that the score and review are subject to change when the game goes live.

[Graeme Dice]

Seriously I don't know why you all give RTS games such a huge pass on the mutliplayer. No other game type gets away with this, would you give a MMORPG a high score and call it a review if you could play before the public can?

MMORPGs are inherently multipalyer games. RTSs are not inherently multiplayer. There are many, many people who will never play more than a couple matches of any particular RTS online, but will still play through the entirety of any single palyer campaign.

[Tom Chick]

Seriously I don't know why you all give RTS games such a huge pass on the mutliplayer.

I don't want to get in the way of a good rant, but "huge pass on the multiplayer" is misleading. I tested Dragonshard's multiplayer quite a bit. What I didn't test was the online Gamespy support, which does indeed seem to suck. I just now tried to play online and couldn't even frickin' get in a game, 1v1 or otherwise.

It's a tough call, yurislave. Do you really want reviews to come out later, so a game has been tested "in the wild", or do you want them out on the day a game is relased? It's a particularly important issue for print media with their long lead times. But do you think PC Gamer is going to take any flak for reviewing FEAR so early? Certainly not enough to offset the advantage of an exclusive first review.

It's the nature of the industry, and while I agree that it sucks, it's ultimately driven by you guys, the readers. You want reviews fast, and you want them broad, and you want them to agree with what you think, which is often uncritical, and you want them delivered with platitudes in neat templates.

Peronsally, I'm trying to buck as many of those trends as I can, but I'm just one dude. :)

-Tom

[Sharpe]

To back up Tom, he did in fact test Dragonshard pretty extensively on his LAN with 2v2 and 3v3 games. He just didn't have a chance to test Gamespy extensively because of timing. Despite the historical Gamespy bashing (which is deserved) the current performance of Dragonshard on Gamespy is atypically bad. What makes it worse is that Dragonshard does not offer direct IP so there is no way to bypass Gamespy.

Given that I couldnt join games at all after 9pm last night and that Tom can't join games at all today, and all the problems that other Gamespy-dwellers had with joining recently, I'm beginning to think this problem is more than just bad code but is something with Gamespy and/or Dragonshard that is actually "broken".

Some type of patch, fix or at least an announcement is clearly required now. And don't dump on Tom, he does test all his reviewed games really extensively on his LAN (I know, I am a frequent guinea pig).

Criticism of Gamespy is overdue in general in the industry. For both AOM and RON, direct IP was far better than Gamespy. Gamespy was adequate for Kohan II but quite poor for Age of Kings, and totally crappy for Dragonshard.

Perhaps somebody could run a feature on online matchmaking comparision tests: Gamespy vs Bnet vs direct IP performance for a number of big-name games. I suspect Gamespy would come out looking like what it is: the early overloaded AOL of online gaming.

[Luke M]

moral of the story: Gamespy sucks.

sluggo, defend your honor!

[Gordon Cameron]

I thought the Zone was pretty good back in the AoK heyday... :P

[wildpokerman]

moral of the story: Gamespy sucks.

sluggo, defend your honor!

I bet sluggo has as much to do with gamespy arcade as Tom Chick has with the demo disk in CGM.

[SlyFrog]

moral of the story: Gamespy sucks.

sluggo, defend your honor!

I bet sluggo has as much to do with gamespy arcade as Tom Chick has with the demo disk in CGM.

I saw him looking shifty and pissing on a stack of Dragonshard CDs while twirling his moustache. I say we get a rope.

By the way, who is sluggo?

[Luke M]

moral of the story: Gamespy sucks.

sluggo, defend your honor!

I bet sluggo has as much to do with gamespy arcade as Tom Chick has with the demo disk in CGM.

Oh, I'm sure he's got nothing to do with it. But he does work there. I was hoping for an official response or an admission of suck or an apology for afflicting the PC gaming world with such crappy software or something like that. Just being goofy, mostly.

[sluggo]

I will defend our honor. :)

At times, blaming GameSpy for a lousy in-game browser is like blaming John Carmack for a lousy third-party Quake-engine game. It's not like there are guys here who actually build content within each game -- the tech is licensed to the developer, who then build a browser with it. So how much blame should be assigned to the tech itself, as opposed to the developers who build a browser with it, or the publisher who forces the game out the door without proper testing? The tech is there. It's been put to good use many times before, which is why so many companies continue to license it. But at some point, the tech is out of GameSpy's hands and it's up to the developer to do something intelligent with it.

In truth, I have zero official involvement with the technology side of our business, so I don't know if this is something that could be improved with better support. Maybe that's part of it. I dunno. But in talking with our SDK guys, they've often complained about the uphill battle it's been getting developers to implement GameSpy tech properly. At times, they get the impression that developers put their dumbest people on that part of the project.

I won't crap on any developers by name here, but there was one really high-profile title last year whose online launch was completely botched because the developer made the worst server browser I've ever seen and absolutely refused to follow a few simple guidelines with server reporting. It's like our guys said "this setting needs to be 1", the developer said "nah, we're going to make it 0," and then blamed GameSpy when everything crashed.

What makes this tough on GameSpy's end is that no one's going to publicly say "don't blame us -- the developers at XXX are a bunch of retards." After all, these companies are paying GameSpy to license the tech, and no one wants to piss on their customers. So the guys on this end just have to kind of grit their teeth whenever someone takes the tech and makes a lousy browser with it, and try to work with them behind the scenes to improve it.

As I said, I don't work with our technology licensing group, so there may be developers here who know more about working with GameSpy tech than I do. But I can promise you, every time someone releases a game and the browser isn't working properly, no one is sitting here and twirling a mustache and cackling about it -- odds are our SDK guys are even more frustrated than the gamers who are trying to use the browser.

[Marcus]

Very interesting.

[sluggo]

Additionally, if I can get on a moral high horse for a second: It thoroughly pisses me off when pubs -- especially print -- just flat-out guess that the multiplayer will work fine, then there are all sorts of in-game browser or Xbox Live issues (Madden, Tiger 05, NFL2K) that are nowhere to be seen in the review.

Will PC Gamer get flak for their FEAR review? If there are major problems in the retail release, they might. They certainly got some shit for a Half-Life 2 review that somehow ignored the existence of Steam. These pubs are essentially rolling the dice and hoping they don't get burned, which is kind of sad when you consider reviews are, at least in theory, supposed to fill some sort of service to the consumer.

On our end, we've gotten into the habit of holding off reviews until we can test games in a live environment -- something we did for both Dragonshard and Winter Assault last week. Instead of rushing the review up and guessing, we post some first impressions on day-of-release and take a few more days to test. Heck, when I reviewed Half-Life 2 last year, I flew to Valve to play it, but wouldn't post my review until I had tested both physcial copies of the game and the Steam download on my own hardware. Personally, in a web environment, I think this is the proper way to cover new releases, and wish more pubs would come around to this way of thinking ... or at least notice that someone's doing it.

[Linoleum]

I'll echo what Sluggo said in regards to in-game server browsers: SDK issues aside, this stuff generally gets underscheduled and thrown at junior programmers. The game industry has no real concept of proper UI design or usability testing.

That said, I hope I never deal with the GameSpy SDKs in any way shape or form ever again. :lol:

Thankfully GameSpy has some competition in that space now, and they are slowly gaining traction. The problem with the middleware market is getting momentum, it gets a lot easier after you sign your first couple shipping titles using your package.

[Sharpe]

I don't know or care where blame lies between Gamespy, Atari, the devs, etc. But after a week of trying, I basically cannot play multiplayer. With some effort of joining and rejoining, a 1v1 can be had, but even then there are problems with lag and disconnects. And anything bigger is a no go.

The online multiplayer is just borked. The interface is weak, the feature set is seriously lacking (cannot adjust starting game settings or chat settings from within multiplayer, no friends list, very little info given on available games, etc), the join/start game stuff is working in a very poor and unreliable fashion, and then once you get in the lag is poor. And worst of all, games of 3 or more players are essentially a no-go.

It's too bad. The non-online game is decent but short and fairly limited (only 2 short campaigns, only 10 total skirmish maps). In theory the online game with users using custom maps, team vs team games etc would give the game some legs but that's not happening. I've seen a few custom maps but given all the connectivity troubles, no one trusts them.

This is very frustrating in a game with such solid and elegant design. The short list of maps, the lack of a campaign for the 3rd race, and the seriously incomplete and inadequater multiplayer support are strong evidence of a release that was just too early. The core design is so good, and the execution is good enough on what is finished that with another 2 to 3 months of development, this could have been a seriously excellent title. As it is, yet another 2-week-wonder. :(

Edited due to dyslexia

[Luke M]

Thanks for being bold and responding, sluggo. Your responses make sense, and I will echo that Gamespy has been a charm in many FPS games. It's just that it seems to be the kiss of death when it comes to RTS multiplayer game matching. We all know that Dragonshard is completely screwed up, and indeed that may be the developers fault, but Dawn of War, Winter Assault, and C&C Generals aren't spectacular, either. At least they're functional. To put it another way, Gamespy + FPS = gold, but Gamespy + RTS = iffy at best. I've listed only a paltry 3 examples, so perhaps the multiplayer devs are morons on all projects, but maybe it's also possible that the Gamespy SDK just needs to beef up its RTS support/documentation. Of course I could be completely off here, seeing as I'm not a dev and all. But I think there is definitely room for improvement.

[Wheelkick]

Will PC Gamer get flak for their FEAR review? If there are major problems in the retail release, they might. They certainly got some shit for a Half-Life 2 review that somehow ignored the existence of Steam. These pubs are essentially rolling the dice and hoping they don't get burned, which is kind of sad when you consider reviews are, at least in theory, supposed to fill some sort of service to the consumer.

On a tangent, I think the gamespy server browser works great in the FEAR MP demo.

And my guess on who's fault it is: A bad SDK or a bad implementation? Dandylife and Linoleum probably summed it all up. The truth lies somewhat in between: a less then perfect SDK gets implemented as low prio work by junior programmers.

[wildpokerman]

Can anyone speak to the online multiplayer? Right now AOE III is just lousy, I'm hoping some good balance patching and possibly some kind of rating system might make this fun.

Is there enough of a community to make Dragonshard a worthwhile purchase at this point or is it one to skip because it won't ever reach the critical mass to be called a "community".

It seems like the problem with RTS games is that the community of people willing to play rts games online is stretched too thin and concentrated in the two or three at a time that are good to make buying multiple games a worthy persuit.

[Mark Asher]

It probably deserves mention in reviews if the multiplayer was tested only on a LAN.

[wildpokerman]

It probably deserves mention in reviews if the multiplayer was tested only on a LAN.

That was my earlier point exactly. Testing the multiplayer balance does no good if the average joe without a home lan can't get a game on...

[Aszurom]

*why* in the hell would a company develop a game and put all the effort into multiplayer and NOT just have a simple direct-ip connect/host option? Is there something in the gamespy contract that denies them the ability to put just a simple input field in their UI instead of forcing you to deal with the matching service? I mean, if you HAD to go there to meet people at least you could exchange IPs and avoid the gamespy issues anyway.

I think that tying multiplayer to a particular service without possibility of direct IP connection is quite a bullet point in a review that bothers to discuss multiplayer at all.

If that service sucks, or goes bye-bye, then you'll not be playing online - and I think that's pretty major in a fair assessment of a title's features.