NS2 will wither on the vine if the server isn't fixed

Security

<!--quoteo(post=1923565:date=Apr 9 2012, 06:33 PM:name=Techercizer)--><div class='quotetop'>QUOTE (Techercizer @ Apr 9 2012, 06:33 PM) <a href="index.php?act=findpost&pid=1923565"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->UWE are working on server performance as hard as they can; they know it's an issue and are working at it furiously.<!--QuoteEnd--></div><!--QuoteEEnd-->

We would be a lot less worried, if we got some more detailed insight about the attempts to solve this problem, as well as work and progress on it then.

UWE are usually very communicative about all kind of things, but neither changelogs, nor the developer tracker had anything about progress on server performance/multithreading/linux build for a long time now.
At the same time the problem doesnt seem to get better.

And with the release date getting close, people automaticly assume that it doesnt have a high enough priority.

Insane

<!--quoteo(post=1923578:date=Apr 9 2012, 06:33 PM:name=Security)--><div class='quotetop'>QUOTE (Security @ Apr 9 2012, 06:33 PM) <a href="index.php?act=findpost&pid=1923578"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->UWE are usually very communicative about all kind of things, but neither changelogs, nor the developer tracker had anything about progress on server performance/multithreading/linux build for a long time now.
At the same time the problem doesnt seem to get better.<!--QuoteEnd--></div><!--QuoteEEnd-->

The thing is, you'll never see anything like "Improve Server Performance" on the progress tracker. It's too big an issue; things that affect performance are broken into chunks and dealt with one by one. Just because you don't see an obvious indicator that an issue like server performance isn't being addressed, or isn't a high priority, doesn't mean that it isn't true.

Yuuki

I think the answer to this kind of topics is to give people something to chew on, technical details. Just saying "they are working on it", "it's top priority", "it will be ready this summer" doesn't have so much content, it doesn't feed you for very long.
But I know making detailed post or videos takes time, and time is a scare resource.

ScardyBob

<!--quoteo(post=1923590:date=Apr 9 2012, 01:29 PM:name=Yuuki)--><div class='quotetop'>QUOTE (Yuuki @ Apr 9 2012, 01:29 PM) <a href="index.php?act=findpost&pid=1923590"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->I think the answer to this kind of topics is to give people something to chew on, technical details. Just saying "they are working on it", "it's top priority", "it will be ready this summer" doesn't have so much content, it doesn't feed you for very long.
But I know making detailed post or videos takes time, and time is a scare resource.<!--QuoteEnd--></div><!--QuoteEEnd-->
Well, Hugh did make a video with UWE talking about <a href="http://youtu.be/Dt8_lShjatg" target="_blank">performance (multithreading)</a> back in January. Maybe we can suggest he do another. Unfortunately, videos about performance can be a bit boring because there's nothing visually (besides talking heads) to show.

Wheeee

i think it's just frustration mounting because the last ~6 patches have shown steadily decreasing performance for many people, and that is exactly the opposite of what we're told is supposed to happen, and what we're expecting. i think people would be much less worried if b203 performance wasn't the worst it's been in ~15 patches.

Zeno

Sorry if it's only semi-on-topic but I have a technical question to those who can code:
As far as I understand LUA is a main bottleneck for ns2 either for the client and for the server. Simply porting the lua part of the code to c++ won't work for many reasons and would revert the modability of the game of course... . Also the performance increase wouldn't be that great probably. But it was stated that the lua engine/interpreter is indeed slow and a massive bottleneck... would it be theoretically possible to create an automatic lua->c++ translator which could translate and compile the code for the server only? So servers could still work with lua but when started the admins would compile the game code first and then run the server lua-free - clients would still run normal lua.
Sorry if it's total BS, I have no idea of computer science :P

Yuuki

It's probably possible yes :

<a href="http://lua-users.org/wiki/LuaToCee" target="_blank">http://lua-users.org/wiki/LuaToCee</a>

But I guess it would require a lot of work. Some languages are based on translation to c but I don't really know the pros and cons compared to a virtual machine.

player

Nah, not a chance. First of all, the binding-layer would still be the Lua-API, which incurs significant overhead. Additionally, making C-API calls to Lua what would normally be a Lua-script compiled into bytecode may in fact be slower. Further more, Lua deploys many programming-techniques which cannot (sanely) be ported over to C. A bloody mess it would be.

I do agree with the sentiment that the server should've been operating on a C++-SDK, and the Lua-scripts be kept client-side, in order to achieve the maximum moddability, but still retain the sandbox which is an absolute must for the clients.

SentrySteve

The 'stop complaining' crowd needs to realize they're talking to Sycthe, who would only bring this up because he cares and there haven't been too many server performance related communications coming from UWE. He's right in every regard and, if most of the responsibility falls on Max, I'd hope other devs are able to assist in whatever way(s) possible. While I doubt the server issue has ever been ignored, it's time to do whatever it takes to fix it as soon as possible. If they've had someone going full steam on the server issue for the past few months then that would be pretty terrible news as no <i>major</i> progress has been made.

Wheeee

i wouldn't mind waiting another year or more, if they needed a total rewrite of the server code to hit acceptable performance benchmarks on normal server hardware. it's that important.

Techercizer

<!--quoteo(post=1923637:date=Apr 9 2012, 07:01 PM:name=Wheeee)--><div class='quotetop'>QUOTE (Wheeee @ Apr 9 2012, 07:01 PM) <a href="index.php?act=findpost&pid=1923637"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->i wouldn't mind waiting another year or more, if they needed a total rewrite of the server code to hit acceptable performance benchmarks on normal server hardware. it's that important.<!--QuoteEnd--></div><!--QuoteEEnd-->
I think we all would. The only people I've seen pushing for a Summer release are UWE, but I don't think they're doing it for no reason.

Camron

I would love to see UWE write and use a Lua to C++ transcompiler. It wouldn't be that hard if they avoid some Lua features like closures. Then they could just use that against a C++ API. They could still continue to write code in Lua and even allow the client/server run it but also have the option of running the compiled-transcompiled code.

The current performance at this stage (a few months from release) is unacceptable. I still don't recommend NS2 to people yet because of it. I tell them how fun and the awesome the game is but I tell them to hold off on buying it <b>only because it's not guaranteed that the performance issues will be fixed.</b>

kabab

If they want this thing to have any e-sports future it will have to run at 60 ticks also 30 is just not enough for pro matches...

Quake wars came out with the servers capped at 30 ticks which really upset a large portion of the player base..

Hugh

We are absolutely committed to improving server performance and client fps. By large multiples. It is not anywhere near where we want it to be. We do have some big cards yet to be played, so keep the faith!

Wheeee

<!--quoteo(post=1923699:date=Apr 10 2012, 12:21 AM:name=Strayan (NS2HD))--><div class='quotetop'>QUOTE (Strayan (NS2HD) @ Apr 10 2012, 12:21 AM) <a href="index.php?act=findpost&pid=1923699"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->We are absolutely committed to improving server performance and client fps. By large multiples. It is not anywhere near where we want it to be. We do have some big cards yet to be played, so keep the faith!<!--QuoteEnd--></div><!--QuoteEEnd-->

thanks hugh. i needed that.

Alchemy

Over-clocked hardware as a requirement for servers is the number one issue currently for NS2 imo. The effect on the community in Australia/New Zealand is very noticeable.

Many (including myself) are investigating options which involve buying connections and hardware to fill the gap therefore can we please have some high level guidance on the time frame for these server improvements?

Is it worth the community spending money to fill the gap now?

Argathor

<!--quoteo(post=1923706:date=Apr 10 2012, 05:59 AM:name=Alchemy)--><div class='quotetop'>QUOTE (Alchemy @ Apr 10 2012, 05:59 AM) <a href="index.php?act=findpost&pid=1923706"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Is it worth the community spending money to fill the gap now?<!--QuoteEnd--></div><!--QuoteEEnd-->
I think this is important. In a situation where the community are having to spend huge sums of money just so we can actually play the beta, it would be respectful to keep us more informed than in other areas.

Without the community servers UWE couldn't even offer the chance to play the beta with pre-sales, it simply wouldn't be possible. Many of us are incredibly frustrated with the current performance, yet still considering investing large sums to create more quality servers. Without any idea of how long we might be expected to put up with the current situation it makes the decision difficult though. More communication would be appreciated, even if its that we will have to deal with this for months yet.

Argathor

<!--quoteo(post=1923699:date=Apr 10 2012, 05:21 AM:name=Strayan (NS2HD))--><div class='quotetop'>QUOTE (Strayan (NS2HD) @ Apr 10 2012, 05:21 AM) <a href="index.php?act=findpost&pid=1923699"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->We are absolutely committed to improving server performance and client fps. By large multiples. It is not anywhere near where we want it to be. We do have some big cards yet to be played, so keep the faith!<!--QuoteEnd--></div><!--QuoteEEnd-->
While I appreciate the effort (and <b>everything</b> you do and have done for the NS2 community), your reply is without substance. Please can we get some more information, especially regarding the expected timescale.

Deciding to spend £100+ a month, just to discover a decent server is released would be incredibly frustrating. Likewise, not doing so and allowing the current situation to not only aggravate current players but also put off new players would be equally bad. Not knowing how long we are likely to have to deal with the current situation is the problem.

Security

<!--quoteo(post=1923706:date=Apr 10 2012, 05:59 AM:name=Alchemy)--><div class='quotetop'>QUOTE (Alchemy @ Apr 10 2012, 05:59 AM) <a href="index.php?act=findpost&pid=1923706"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Over-clocked hardware as a requirement for servers is the number one issue currently for NS2 imo. The effect on the community in Australia/New Zealand is very noticeable.

Many (including myself) are investigating options which involve buying connections and hardware to fill the gap therefore can we please have some high level guidance on the time frame for these server improvements?

Is it worth the community spending money to fill the gap now?<!--QuoteEnd--></div><!--QuoteEEnd-->
<!--quoteo(post=1923734:date=Apr 10 2012, 10:10 AM:name=Argathor)--><div class='quotetop'>QUOTE (Argathor @ Apr 10 2012, 10:10 AM) <a href="index.php?act=findpost&pid=1923734"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->I think this is important. In a situation where the community are having to spend huge sums of money just so we can actually play the beta, it would be respectful to keep us more informed than in other areas.

Without the community servers UWE couldn't even offer the chance to play the beta with pre-sales, it simply wouldn't be possible. Many of us are incredibly frustrated with the current performance, yet still considering investing large sums to create more quality servers. Without any idea of how long we might be expected to put up with the current situation it makes the decision difficult though. More communication would be appreciated, even if its that we will have to deal with this for months yet.<!--QuoteEnd--></div><!--QuoteEEnd-->
<!--quoteo(post=1923735:date=Apr 10 2012, 10:17 AM:name=Argathor)--><div class='quotetop'>QUOTE (Argathor @ Apr 10 2012, 10:17 AM) <a href="index.php?act=findpost&pid=1923735"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->While I appreciate the effort (and <b>everything</b> you do and have done for the NS2 community), your reply is without substance. Please can we get some more information, especially regarding the expected timescale.

Deciding to spend £100+ a month, just to discover a decent server is released would be incredibly frustrating. Likewise, not doing so and allowing the current situation to not only aggravate current players but also put off new players would be equally bad. Not knowing how long we are likely to have to deal with the current situation is the problem.<!--QuoteEnd--></div><!--QuoteEEnd-->

Indeed. I've been pondering about wheter or not I should buy & overclock my own server blade, to colocate at a data center.

I just don't know if its worth it, or if UWE might suddenly come up with improvements that would make all those investments obsolete.

LV426-Colonist

I know they said this game isn't going to be optimized overnight -- but how long have they been saying that now?

The game is almost unplayable in it's current stage.

sherpa

Are there any ETAs on performance improvements? I've got my reciept from paypal for NS2, dated May 2009.

3 years ago.

Whilst performance improvements have been dramatic from the first beta to the current one, there still needs to be an improvement greater than what we've seen since 2009. That's kinda scary!

Xarius

Well, looking forward to those 'big cards' they have yet to play. I think the game is progressing at a good pace, though I can't imagine we'll have a fully optimised and more or less feature complete game in summer 2012. One can hope of course.

Wilson

<!--quoteo(post=1923637:date=Apr 10 2012, 01:01 AM:name=Wheeee)--><div class='quotetop'>QUOTE (Wheeee @ Apr 10 2012, 01:01 AM) <a href="index.php?act=findpost&pid=1923637"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->i wouldn't mind waiting another year or more, if they needed a total rewrite of the server code to hit acceptable performance benchmarks on normal server hardware. it's that important.<!--QuoteEnd--></div><!--QuoteEEnd-->

Unfortunately, working for 1 year costs money.

PogoP

It'd be nice to see a few more UK/EU-based servers! I keep joining US servers as they tend to have free slots, but my ping is always pretty bad.

Asraniel

<!--quoteo(post=1924063:date=Apr 11 2012, 08:14 AM:name=LV426-Colonist)--><div class='quotetop'>QUOTE (LV426-Colonist @ Apr 11 2012, 08:14 AM) <a href="index.php?act=findpost&pid=1924063"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->I know they said this game isn't going to be optimized overnight -- but how long have they been saying that now?

The game is almost unplayable in it's current stage.<!--QuoteEnd--></div><!--QuoteEEnd-->

This is really an exageration, or you didn't play the patches before the last one. The game became much much faster. Not long ago, servers couldn't even handle 6v6 games.
You mostly don't see the progress that is beeing made, because everytime the game gets faster, the servers increase the max player count, which of course drops performance again.

In reasonably sized games (7v7) and with a good ping, while not optimal, its far from unplayable.

acid_rain

Every game goes through this stage. You've just got to remember that's all it is... a stage.

Arkanti

<!--quoteo(post=1924144:date=Apr 12 2012, 12:52 AM:name=Asraniel)--><div class='quotetop'>QUOTE (Asraniel @ Apr 12 2012, 12:52 AM) <a href="index.php?act=findpost&pid=1924144"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->In reasonably sized games (7v7) and with a good ping, while not optimal, its far from unplayable.<!--QuoteEnd--></div><!--QuoteEEnd-->


On anything resembling a regular game server, it is.