Face it, there needs to be a test server.
Why would they bother doing a test server when people would use them to test everything BUT what anet put on there to test? No reason to use a test server when it wouldnt accomplish anything.
Guys, gotta stop listening to Marketing and do public tests on patches before they are released live. There is no benefit to not doing public testing, except that you lose some of the ‘ohh, shiny new things’ surprise for players, which I’m sure Marketing is telling you needs to happen. However, releasing things with so many bugs utterly destroys the positive effect this could have as right now, there are so many bugs, people are just so frustrated with everything.
There is another game out there that has had test servers since day one and as a former customer of that game I can say with absolute certainty it never helped one bit. As a player of both that game and GW2, patches in GW2 get released with fewer bugs and fewer major bugs than that game-that-shall-not-be-named that does have a test server.
People expect an MMO patch to be absolutely perfect upon release and it will never happen. There will always be a few nagging bugs here and there and so far Anet is doing a fine job of catching the most obvious before they are released. Never once has GW2 had to be taken down for a hours at a time due to a faulty patch.
Jaina Kitten, Level 80 Elementalist, Main
Yak’s Bend – Expletus
Yak’s Bend – Expletus
People expect an MMO patch to be absolutely perfect upon release and it will never happen. There will always be a few nagging bugs here and there and so far Anet is doing a fine job of catching the most obvious before they are released. Never once has GW2 had to be taken down for a hours at a time due to a faulty patch.
It’s not that we expect it to be perfect, it’s that we expect when they mess up random things that they at least let us know that they are aware of the problems and actually do something about it.
A public test server isn’t a bad proposal. Something needs to change with this system. It doesn’t necessarily have to be a public test server though.
(edited by Purple Miku.7032)
The amount of time a public test server would add to releases would be unacceptable. We already have people complaining releases come to infrequently. This is how the game is going to roll. Most of the worst bugs are fixed quickly anyway.
I’ve been in games with a PTS and I still found way too many patches to fix stuff after launch, and many of the worst bugs weren’t identified.
Why would they bother doing a test server when people would use them to test everything BUT what anet put on there to test? No reason to use a test server when it wouldnt accomplish anything.
" Solutions To A Problem Can Only Be Found, When You Want To Get Rid Of It "
Ankur
Ankur
Believe or not but it happens that some bugs never occur on testserver and only
when they are released on live servers.
In the end testserver are mostly used to know where you can make profit when
stuff is released, or testing out stuff there that would else cost you money on
live server.
EVERY MMO is awesome until it is released then its unfinished. A month after release it just sucks.
Best MMOs are the ones that never make it. Therefore Stargate Online wins.
Best MMOs are the ones that never make it. Therefore Stargate Online wins.
I would not make a test server to be a carbon copy of the game, elsewhere. Just make it one WvW map and one PvP map, you would find a vast majority of the skill and balance bugs with a couple hundred players there in a couple of hours. They are pretty glaringly obvious, I cannot see how they would be missed, unless you hardly do any testing or have, like, two people doing it.
(edited by Heimskarl Ashfiend.9582)
I would not make a test server to be a carbon copy of the game, elsewhere. Just make it one WvW map and one PvP map, you would find a vast majority of the skill and balance bugs with a couple hundred players there in a couple of hours. They are pretty glaringly obvious, I cannot see how they would be missed, unless you hardly do any testing or have, like, two people doing it.
I don’t think you’d get a couple of hundred players in there. That was the problem Rift had. They had a test server and no one used it, because you’re not advancing anything by using it.
Guys, gotta stop listening to Marketing and do public tests on patches before they are released live. There is no benefit to not doing public testing, except that you lose some of the ‘ohh, shiny new things’ surprise for players, which I’m sure Marketing is telling you needs to happen. However, releasing things with so many bugs utterly destroys the positive effect this could have as right now, there are so many bugs, people are just so frustrated with everything.
A test server wouldn’t solve the issue. I’m not saying ANet has done a good job of preventing bugs; I’m saying that it isn’t as easy as setting up a test environment that players could use.
- Test servers increase the amount of change and configuration management required. Some of the bugs we’ve seen already indicate that this is problematic already, so adding a test server would make it worse.
- The time-to-fix the bugs they do find would increase, since now they have to go through an additional level of review. Even for the devs, destroying their test environment requires regrouping and starting some tests from scratch. For players in the test environment, it would either be a nightmare or it would further increase time-to-release for bug fixes.
- Keeping track of bugs would become harder, since there would be an increase in the number of situations in which they are found. Keep in mind that some of the bugs we’ve seen have come out of tinkering with mechanics ABC, without realizing they interfere with XYZ. Now we’d have A’BC and A’B’C’, too, making it harder to find root causes.
It turns out that test servers are great for situations in which the development process is already solid. If there are existing systemic problems, a test server makes things worse.
So, as always, my request is that ANet consider changing they way they manage builds and figuring out how they can run regression tests on more mechanics. Having a test server should be, imo, last on their priorities …at least for now.
John Smith: “you should kill monsters, because killing monsters is awesome.”
Why would they bother doing a test server when people would use them to test everything BUT what anet put on there to test? No reason to use a test server when it wouldnt accomplish anything.
WTF! Test servers are used successfully in almost every MMO.
Jade Quarry [SoX]
Miranda Zero – Ele / Twitch Zero – Mes / Chargrin Soulboom – Engi
Aliera Zero – Guardian / Reaver Zero – Necro
Miranda Zero – Ele / Twitch Zero – Mes / Chargrin Soulboom – Engi
Aliera Zero – Guardian / Reaver Zero – Necro
They honestly have made it pretty clear they don’t want, or at least have no intention to have, a public test server.
Honestly if they ever had on-going public testing of any kind it would probably be something like the sPvP side of the game only, definitely not WvW (technical reasons, judging from how the WvW stress test is being handled vs other recent beta events) and PvE they don’t want to give any spoilers on quite clearly.
Why would they bother doing a test server when people would use them to test everything BUT what anet put on there to test? No reason to use a test server when it wouldnt accomplish anything.
WTF! Test servers are used successfully in almost every MMO.
They’re used in almost every MMO. Successfully is a matter of opinion.
I think they need to focus on faster releases rather than more testing.
At the moment the community identifies the bugs hours after a patch anyway.
But then we have to wait weeks/months for fixes…
John Snowman [GLTY]
Space Marine Z [GLTY]
Space Marine Z [GLTY]
Why would they bother doing a test server when people would use them to test everything BUT what anet put on there to test? No reason to use a test server when it wouldnt accomplish anything.
WTF! Test servers are used successfully in almost every MMO.
I’m laughing. So hard right now. Every mmo I’ve played with a test server has 99% of the time never ended up seeing bug hunting or other such “Beta Testing” as the primary activity. In fact, almost every single time the first thread you’d see on their forums was about profiting off the changes in the best way possible. It was also the next 30-50 or so threads too.
Or test servers end up relegated to the role of “log in here to test character builds” and never serve any other purpose.
But then we have to wait weeks/months for fixes…
That honestly depends on how much of a clusterkitten the bug’s source is, among other things, one of which is impending changes. It wont matter a bit if they fix a bug while disregarding the fact that the next major revision going out is, without accounting for the fix just used, going to simply break what was just fixed.
I think they need to focus on faster releases rather than more testing.
At the moment the community identifies the bugs hours after a patch anyway.
But then we have to wait weeks/months for fixes…
The golden rule of Software development is that the total number of bugs never go down, as fixing one thing usually breaks 3 other things in the process. The intent of Testing is to identify the most significant issues, and deal with them before the user base gets their hands on it. Success varies by the nature of the issue, but in general you should NOT lose core functionality by the time it get released into the Live build.
The frequency of the patches aren’t the issue… in fact, I’d argue the high rate of failures in code are a direct result of the short patch cycles they were pushing for with Living Story. The problem is the quality of QA testing that isn’t low level technical seems almost non-existent.
I’ve been having an extremely hard time differentiating the category of issues we’ve been encountering; and is something I’ve never encountered before in any games (including betas) over the last decade and half. Normally you’d be able to identify weak areas of a studio, as well as their stronger areas. But with Anet its about as confusing as it gets. It could just be that I’ve never encountered a game’s development that is so solid on so many levels, and then just completely botch something like art and animation when they have standardized skeletons and meshes for each race.
Whats weirder is that the game only has a few low level technical and game design issues, but mountains of problems with implementations, and even mathematics, that hamper advanced gameplay. Condition damage was a long standing issue, having chosen to ignore an obvious design implementation problem just because of a brute force technical limit. Whats even funnier is that they fixed the technical issue, but still have similar problems with the design and balancing that uncapping it was said to resolve.
Obviously progress was made. But it only served to verify another issue the community has long since been aware of….. conditions and its ramp up has nothing to syngerize with, leaving it unable to compete with crit burst damage. The only reason condi damage had a place in PvP and WvW is a disparity in the skills, which suffers from an identify crisis from the PvP/PvE unified balance method they insist on using. This wouldn’t be a problem had they centered PvE combat design to be a pure DPS race. GW2 had set itself out to be different…. so it always felt strange that mob fights were incredibly simple in nature, utilizing none of the finer points a balanced build could offer, nor intermingling any type of support role that wasn’t completely offensive focused.
can attest that EVE Online also has a test server, and that it was rarely used for its intended purpose
Karma Express – Norn Guardian commander
I remember the Beta Weekends and Stress Tests for GW2 Vanilla. At the beginning there were a few People who refused to participate because their Progress wouldn’t be safed to the Game itself. This side grew more and more until the Maps were really empty in the Stress Tests with so few participating that the Stress Test would not help to stabilize the Servers on Release.
First. Even with Test Servers there would still be big Bugs going into the Live Server because the Live Servers have many more Players and a different environment.
Second: Test Servers are rarely used for their intended purpose ( finding bugs ). Most of the People going to the Test Servers just want to play some content earlier than others
Third. There are many People unwilling to participate because it is “wasted time” that could be spend otherwise ( like farming Gold or other Stuff )
test server depends on 2 things
community finding bugs.
people fixing said bugs.
I already know who would be the ones finding the bugs and it isnt the vast majority of the players.
Fixing bugs. lol.
Why would they bother doing a test server when people would use them to test everything BUT what anet put on there to test? No reason to use a test server when it wouldnt accomplish anything.
Only really true for narrative based patches. Things like the trait overhaul however would fit nicely into the rest server.
Also you can always weed out people who don’t provide feedback… so I guess semi-public then? Definitely opt-in, since I don’t feel like testing something like the new traits.
I think they need to focus on faster releases rather than more testing.
At the moment the community identifies the bugs hours after a patch anyway.
But then we have to wait weeks/months for fixes…
But in theory (and I should put some emphasis on theory… in fact no this is more akin to blatant speculation, even below a hypothesis… still…) if we had proper testing in theory we could get content faster.
The golden rule of Software development
If it can be null it will be null?
test server depends on 2 things
community finding bugs.
people fixing said bugs.I already know who would be the ones finding the bugs and it isnt the vast majority of the players.
Fixing bugs. lol.
most players dont find bugs, but you dont need most players to find them as long as a dedicated % does. (with gw2 population, even a small % is probably enough)
The real problem is probably fixing bugs and reacting to feedback. Its possible that finding more bugs wont help if they dont have the infrastructure to fix bugs rapidly.
it may actually be pretty useful for design level changes though. Because once those go live they stay there for a long, long time.
I remember the Beta Weekends and Stress Tests for GW2 Vanilla. At the beginning there were a few People who refused to participate because their Progress wouldn’t be safed to the Game itself. This side grew more and more until the Maps were really empty in the Stress Tests with so few participating that the Stress Test would not help to stabilize the Servers on Release.
First. Even with Test Servers there would still be big Bugs going into the Live Server because the Live Servers have many more Players and a different environment.
Second: Test Servers are rarely used for their intended purpose ( finding bugs ). Most of the People going to the Test Servers just want to play some content earlier than others
Third. There are many People unwilling to participate because it is “wasted time” that could be spend otherwise ( like farming Gold or other Stuff )
test servers dont need tons of people to participate to be effective, unless its stress test related. most of the problems the op are talking about arent stress related. just regular bugs, and balancing.
Shows what I know. I thought there was a Test Server.
Gone to Reddit.
Testing typically takes longer than dev time.
So again I don’t think the answer for Anet is more testing…we’ll get like one balance patch per year.
Maybe they should switch to one off skill/trait changes instead of doing every class at once.
John Snowman [GLTY]
Space Marine Z [GLTY]
Space Marine Z [GLTY]
I totally agree with this.
The game needs a lot more pre-live testing.
I don’t, honestly, care whether that testing happens via a PTR, or via some other method, but it needs to happen.
“Turns out when people play the game, they don’t admire your feet at all.” sephiroth
DICE who makes Battlefield has a CTE program that is actually well liked and gets the job done. If you limit space though, you need to make sure that the people who are there are there to help reproduce bugs, write reports, share feedback, etc.
I personally would love to lead a team of 20-30 people that could do this for every major release. The build wouldn’t take too long to make as most of the time its already being done but just kept on localized servers at Anet.
Anet needs to be in better contact with their hardcore players. The best GvG guilds like NS and Agg, The Abjured from PvP, and DnT for PvE, etc. Those are the people that can help with balancing the game.
Maguuma Guardian
The monk outfit is another great example of why a test server would be good. Then players wouldn’t have to deal with knee jerk fixes that alter items they paid for with real money. You are letting your playerbase pay to test these items and then altering them after the fact. People should not have to log on to find out their favorite new outfit just got altered, it would also save your support team some time, since they are probably dealing with a lot of refunds now.
Believe or not but it happens that some bugs never occur on testserver and only
when they are released on live servers.
Yep. And it’s not just true of MMOs either.
We had a similar problem at work when we updated our database. It was tested by all 7 end-users on the same PCs we would be using the final version on until we were all satisfied that there were no bugs (or at least no more bugs than the previous version, it’s a huge Access database so there will always be a few weird error messages).
The only thing that changed after that was where it was saved on the server and hooking it up to entire set of records instead of a small sub-set for testing. In theory since all the records are in the same format, saved in the same place etc. this should have caused no problems.
Since then we’ve had all kinds of weird errors, including one where if you click the empty space between two buttons it asks if you want to save and if you push yes it crashes the entire program. And 6 months on no one has been able to figure out what causes that one so we can’t fix it.
Why would they bother doing a test server when people would use them to test everything BUT what anet put on there to test? No reason to use a test server when it wouldnt accomplish anything.
WTF! Test servers are used successfully in almost every MMO.
I’ve played 2 MMOs with test servers : Ultima Online and Elder Scrolls Online and both still have bugs appearing after patches. Sometimes things that didn’t appear on the test servers or simply weren’t noticed at the time, sometimes things they couldn’t fix quickly enough (especially with players complaining that they’re sick of having to go to the PTS to try the new system and want the patch released yesterday) and sometimes things players deliberately didn’t report because it was part of an exploit they wanted to use when the patch was released.
Both also regularly had people telling each other the ‘best’ use of the test server was to try out new builds and character designs and just generally to mess around with the extra tools you were given (like instant max level characters) and it didn’t matter if you played the new content there or not.
It also brings in other issues. For example even though it’s completely optional it’s not uncommon for people on the Elder Scrolls Online forum to demand exclusive rewards, preferential treatment or even to be paid real-life money for being “forced” to “waste” time playing on a server where their progress doesn’t affect the live game. (I don’t know if this happens with other games but I was pretty surprised to see it.)
You’ve also got issues like what happened with the Edge of the Mists map. That one did have a public test server, a lot of people used it and actually were testing the new map. And the reports were that it was brilliant for WvW gameplay and there were some really interesting fights going on. Then it was released and pretty much instantly became a karma/XP train map for PvE players which WvW players won’t use.
I’m not quite sure why, but the player behaviour and therefore the effect of the mechanics, content design etc. was totally different on the test server and on the live server.
Shows what I know. I thought there was a Test Server.
Well technically there is. There’s probably even more than 1. But they’re only accessible to Arenanet’s staff (mainly the Quality Assurance department and developers I assume).
What people are asking for is a public testing server where anyone can log in and play new content before it’s released. A bit like the Heart of Thorns betas but for all new releases.
Danielle Aurorel, Dear Dragon We Got Your Cookies [Nom], Desolation (EU).
“Life’s a journey, not a destination.”
“Life’s a journey, not a destination.”
test server depends on 2 things
community finding bugs.
people fixing said bugs.I already know who would be the ones finding the bugs and it isnt the vast majority of the players.
Fixing bugs. lol.
most players dont find bugs, but you dont need most players to find them as long as a dedicated % does. (with gw2 population, even a small % is probably enough)
The real problem is probably fixing bugs and reacting to feedback. Its possible that finding more bugs wont help if they dont have the infrastructure to fix bugs rapidly.
it may actually be pretty useful for design level changes though. Because once those go live they stay there for a long, long time.
Wont matter a single bit. We’ve proven that there are bugs they cannot find on internal testing or small scale testing. In fact, Anet’s declared they’ve fixed bugs before (teq flying away for good on a “too stronk” burn), only for those able to definitively prove it true or false (the communities of TTS, ATT, TxS, etc) do the actual work. Guess what? Within 5 minutes of teq starting, you’d have map commanders announcing teq got done borked and to go post on the forums to let anet know. It happens not once, not twice, but THREE times before teq got it’s current iteration.
Why would they bother doing a test server when people would use them to test everything BUT what anet put on there to test? No reason to use a test server when it wouldnt accomplish anything.
Only really true for narrative based patches. Things like the trait overhaul however would fit nicely into the rest server.
Also you can always weed out people who don’t provide feedback… so I guess semi-public then? Definitely opt-in, since I don’t feel like testing something like the new traits.
The “testing” of traits would involve large scale numbers in a live environment. A test platform would be too small to be idea. SA might not have been glaringly obvious if it wasnt in such a large scale. Unfortunately, I also think you also give too much credit to players.
I’m going to chime in with those that say a public test server isn’t necessarily the way to go.
I’m on a project currently where we (the developers) code and test our own work in a ‘dev’ environment, then it gets pushed to a ‘test’ environment where our QA team checks it, then it gets pushed to a ‘prod like’ environment where the ‘business’ (we’re contracted out to) does their testing, and then finally gets pushed into the actual production environment. We’re currently facing an issue where our code works in dev and prod like, but is buggy as all hell in test, and we have yet to figure out why (considering the primary differences between the environments is the physical box and who has access).
Beyond that, I have played (currently play) a couple of games that have public test servers, and in some instances it helps, but for the majority it really doesn’t. People aren’t very thorough in what they are checking (they don’t think about the fact that not only do you need to check the new content’s functionality, but you also need to check for any issue with existing functionality that may interact with the new stuff – ie, regression testing), or they aren’t checking the new content at all. Instead they are playing with all the freebie stuff that comes with the test account trying to determine how best to utilize it (when they get it) on their real account.
Don’t look at me like that. Whatever you’ve heard, it’s probably not true.
Anet just need a QA department.