Again, why is test here?

[Sv_Lor]

<p>*** Tradeskills ***- We have permanently increased tradeskill experience gain at all levelsby 50%.- Level 50+ imbued shields now have the appropriate effects.*** Desert of Flames ***- Urzyd will use its innate abilities without being interrupted,stifled, or stunned.- The Djinn Caretaker now awards Adept I spells instead of Master I.*** The Gigglegibber Goblin Gamblin' Game ***- Matching 2 numbers should now pay out winnings appropriately.- The correct rewards for winning are as follows:- Match 2 numbers and you win 10 silver.- Match 3 numbers and you win 50 silver.- Match 4 numbers and you win 2 gold 50 silver.- Match 5 numbers and you win 25 gold.- Match all 6 numbers and you win the jackpot!*** Gameplay ***- Rooted NPCs will now turn to attack nearby players.*** Epic Encounters ***- Spirits of the Lost no longer has a level cap of 52.*** Spells and Combat Arts ***Coercer changes:- The stifle portion of Harrowing Silence will no longer affect epictargets.Necromancer changes:- Death Rot, as well as other spells in the line, can now be cast bymore than one Necromancer and stack on the same target.Warlock changes: - The range of Nihilism is now 20 meters instead of 10 meters.Wizard changes:- The range of Frostshield is now 20 meters instead of 10 meters.*** Commands, Controls, and User Interface ***- The scroll bar in the member list of the Guild window works again.- Tooltips will no longer cause a crash after using /loadui.*** UI Files Updated ***eq2ui_mainhud_guild.xml </p> <p>  </p> <p><font color="FF0000">Sorry to offend anyone that is upset over my thread, or sees me as a silly n00b, but I am getting tired of test being last in line.  Are there any bugs with this update/patch?  Well, I guess that live will be able to tell testers...</font> </p> <div></div>

[Chog]

<BR> <BLOCKQUOTE> <HR> Sv_Lorax wrote:<BR> <P><FONT color=#ff0000>Sorry to offend anyone that is upset over my thread, or sees me as a silly n00b, but I am getting tired of test being last in line.  Are there any bugs with this update/patch?  Well, I guess that live will be able to tell testers...</FONT><BR></P> <HR> </BLOCKQUOTE><BR> <DIV>These are hot fixes.  They can not wait a week for the changes to go to test then Live.  There will always be hotfix issues in any MMORPG (any good MMORPG that is).  Test is used to test changes to the game.  Not all changes make it to test becasue of needing to be placed on Live servers as soon as possible.  You want every change to live servers to go through test first and that will never happen.</DIV>

[HanktheDwarf]

This patch was sent to Test last night at 11pm Pacific.  Certainly not enough time to test it anyway, but it was there first.  For hotfixes they rarely ever update the forums, they just update the notes on the patcher. <div></div>

[Daish]

<DIV>so...  chogar...  umm..  they DID put the stuff on test, but never really allowed any of the testers to test it....   and thats OK with you?</DIV> <DIV> </DIV> <DIV>I'm sorry, but maybe the problem is some people dont understand the concept of why one would want to have a test server, nor understand why every piece of code needs to be directly tested..   EVERY time....</DIV> <DIV> </DIV> <DIV>If a dev makes some changes, to lets just use stoplights for illustration...   the dev makes changes because the green light comes on to much.  Dev codes the fixes and a simple test reveales that now the green light no longer ever lights up.  Since the dev just made those changes 5 minutes or 10 hours ago... the Dev says 'ah ha!, ooops let me just fix that real quick'.  A second test shows that the green light does actually turn on this time, but now it does not turn on enough.   Once again, it is set back to dev to re-tweak and send it back to test.  The cycle continues until the fix actually DOES WHAT IS WAS SUPPOSED TO.  If they cannot get the code to actually work like planned, or accomplish what it was set out to do, they DONT RELEASE IT.  Odd concept, huh?</DIV> <DIV> </DIV> <DIV>That is the way software is developed and the way bugs are repaired.  No testing ='s stupid bugs.  Sure you may eventually find bugs in the live code, but, now, instead of immediately being able to remember what the changes were, what the plan was, and having a clear idea of how to fix it, the dev has moved on to 10 other things and has to start the whole process basically from scratch.  In fact, it may no longer even be the same dev or even the same dev team.</DIV> <DIV> </DIV> <DIV>Most production software companies have what they call a 'core' team and a 'release' team.  The 'core' team actually works with the designers and implements new code, system, content, etc.  The release team takes over after the code hits the live servers to do the supposedly 'minor' bug fixes like spelling changes and small math errors.</DIV> <DIV> </DIV> <DIV>Now, some of you can probably see a few easy ways that this system could lead to some of the more ridiculous bugs and messed up things we can see everyday in the world of software.  If the designers or original devs did not do a good job documenting what it was that was supposed to be built, or explain how it was actually accomplished when implemented, the poor 'release' devs now have to recreate all of that on their own as they try to decipher what, hopefully, is easily understandable code and try to fix the problem while maintaining the integrity of the original design, most of which they had to guess about...  </DIV> <DIV> </DIV> <DIV>If you think the 'core' devs have to push to release code in a timely fasion, what do think the 'release' dev's timetable is with thousands of unhappy customers waiting in the wings?  Guess what the regression rate (additional bugs created during a fix) is for these 'high profile' fixes are?  Let me give you a hint, it's higher....</DIV> <DIV> </DIV> <DIV>So, hopefully, you can see how simply putting stuff on the live servers with NO INTENTION of actually testing it actually CREATES more bugs than it fixes, the proof is in the pudding, you can see it as they work through all of the implementation problems they had with the combat revamp, and every other semi-major to major change they have made.  </DIV> <DIV> </DIV> <DIV>They have new 'wonderful' examples of this every day.... The new goblin gambling game didnt work as advertised when first released...  and why might that be?  well, apparently they never tested if the payout works if you get a two number match...     man, that sounds like something one would want to test before putting on the live servers, now dont it?  Or would you consider that just something you can tweak after the paying customers scream about it loud enough?  </DIV> <DIV> </DIV> <DIV>(this starting to sound a bit like the SOE provided 'joy' ride we are all experiencing in EQ2?)</DIV><p>Message Edited by Daishar on <span class=date_text>10-06-2005</span> <span class=time_text>11:02 PM</span>

[xOnaton1]

the goblin game did work fine on test a few days ago. I got 2 numbers matched and I won, but this was an earlier version. It didn't have the slot machine look and it only cost 1 s. So they changed it since then. <div></div>

[xenaphobia]

<P>Things like this kinda make you feel like you pay to test huh? Sure does when I played. I also have a huntch that one of the possible reasons they are rushing new things out so fast is because alot of people are quiting the game. Now those that love the game and still play don't throw a hissy fit, I didn't say everyone.</P> <P>As for the person that went to guild wars, good graphics huh? Y a I liked it but I don't play it much cause I miss a world with no instances. I just like running into people all the time, kinda fun for me. Well I hope you enjoy it.</P>

[lisasdarr]

<span><blockquote><hr>Daishar wrote:<div></div> <div></div> <div></div> <div>so...  chogar...  umm..  they DID put the stuff on test, but never really allowed any of the testers to test it....   and thats OK with you?</div> <div> </div> <div>I'm sorry, but maybe the problem is some people dont understand the concept of why one would want to have a test server, nor understand why every piece of code needs to be directly tested..   EVERY time....</div> <div> </div><hr></blockquote>You do know that everything is put through their internal testing systems before it hits any server, including test?</span><div></div>

[Daish]

<DIV>Lisa....</DIV> <DIV> </DIV> <DIV>are you honestly argueing that the internal test team...  whatever/whomever they are..   you are honestly going to argue that SOE surely put all of these changes through it's extensive internal testing division....   </DIV> <DIV> </DIV> <DIV>You do understand that the code was BROKEN when they put it on the live server, right?  As the person a few replies stated, the game worked at one point on test, but then they changed/updated the code BEFORE they released to the live servers, and the one that actually hit the live servers was BROKEN.</DIV> <DIV> </DIV> <DIV>So, are you arguing that SOE internal test system is also broken, thus proving the need for additional testing on all code that goes into the game?</DIV> <DIV> </DIV> <DIV>Or, are you simply saying that as long as those titans of software testing in the SOE internal departments touch it, all code NEVER needs to go onto the test server?</DIV> <DIV> </DIV> <DIV>What they are doing is rushing code out the door, with no real intention of insuring it's quality or assesing the risk to the paying customer for broken code and/or server downtimes.  This is like a run away train, at some point, it will derail, a large crashing issue will get through (not that there havent already been some crashing issues for some players), and there will be a breaking issue that will affect you.  I hope you are still willing to defend the internal test teams on that day.</DIV>

[Chog]

<DIV>What people fail to realize all MMORPG's are a giant test server / paid beta / what ever name you want to call it.  There is not a MMORPG in production that is bug free or that has patches without bugs included (some more then others).  Does this make it acceptable?  No.  However, putting hot fixes onto a test server for any extended period of time (to correctly test the changes of course) would make the game worse.  Take the exploited expereince gain used by some killing Epic mobs, people would read the test server notes find out people are gaining more expereince then intended and go adventure in the respawning Epic zones with their guild.  A week or two later, the patch hits live and anybody in a raiding guild that wanted to be is now level 60.  Increase the healing rate of Wardens to make them better healers and desired in groups?  Nope, need to wait 1-2 weeks so the number changed on a spread sheet can be correctly tested, tough it out.</DIV> <DIV> </DIV> <DIV>Some changes have to either skip Test, not be documented on test, or be on test for a short period of time.  If you do not like the length certain changes are on test, that is your perogative.  What you need to understand is certain changes are either required on live as soon as possilbe or are trivial and a single person double checking the work can verify the changes.</DIV> <DIV> </DIV> <DIV>Would I love every patch to be bug free and no tweaks needed?  Yes.  However, I know that is unreasonable to ask from a MMORPG.  I do not care if changes where on the Test server for a month, bugs will still get though and complaints will still happen.  Either live with it or find an imaginery MMORPG that does not have bugs.  Maybe in 5-10 years bug free MMORPG's will be available, currently they are not.</DIV>

[Ram]

You are completly correct on this one Chogar. The only real way around the scenario which you laid out would be for EQ2 team to do their own internal testing... Not on the live test server.... <p>Message Edited by Ram on <span class=date_text>10-07-2005</span> <span class=time_text>07:52 PM</span>

[EtoilePirate]

<span><blockquote><hr>Ram wrote:<div></div>You are completly correct on this one Chogar. The only real way around the scenario which you laid out would be for EQ2 team to do their own internal testing... Not on the live test server.... <div></div><p>Message Edited by Ram on <span class="date_text">10-07-2005</span> <span class="time_text">07:52 PM</span></p><hr></blockquote>Chogar is correct, period.  The EQ2 team does lots -- lots and lots -- of their own internal testing.  On Test or in the DoF Beta some dev or other would often say, "hmm, that's not what happened when we tried it..."  And that's the point: each and every server -- Live ones as well as Test -- is a different environment, and things don't always work as intended or even as tested.</span><div></div>

[Daish]

<P>sorry... you guys are funny....   yes, Chogar has a valid point there...  and, I even agree with him.</P> <P>I agree that large game breaking, exploit capable, or crashing bugs need to be fixed as soon as possible, as they represent rather high priority issues....</P> <P> </P> <P>But, here is the point you guys are missing, if these changes were TESTED on the TEST server BEFORE they hit the live servers, there would be a great deal less high priority/profile, game breaking, exploit capable bugs on the live servers....</P> <P>SOE is the one that BROKE the code when they patched, it wasnt some natural event or hurricane that broke the code, it was SOE.  Software is only the sum of it's parts, and bad practices on top of shaky software is going to break..... period.</P> <P>You guys are arguing that there can be a nessesity for direct-to-live patches to solve urgent issues, but are neglecting to acknowledge the fact that there would be a great deal less of these urgent issues if the testing happened FIRST, rather than after the fact.</P> <P>Like the whole warden healing thing....  how did you manage to gloss over the fact that that is part of the combat revamp.  The combat revamp that was 4+ months in design/implementation.  The combat revamp that they NEVER HAD MORE THAN 1 DAY OF TESTING THE FINAL BITS before they released to the live servers.  The combat revamp that 4 WEEKS after release, they are still tweaking and still having to change almost daily.  It was the un-finished, un-balanced, messed up con system combat revamp that made wardens have jacked healing capabilities.  It was SOE's design/implementation for wardens to be that way, and since they never really tested it, now they realized that it is jacked and they changed it.</P> <P>At what part in that little warden history timeline do you guys believe that direct-to-live server patches is the ONLY WAY that SOE can/should build capable sofware and build a game we should be happy to continue to pay for?</P> <P> </P>

[Aszuth]

<P>Daishar -</P> <P>You obviously don't have a clue how things work do you?</P> <P>For starters - TWO different environments, TWO different server (although similar) server builds between Test and Live make your "if they test everything fully there wouldn't be any issues" theory fall apart since any disparity can lead to an unexpected complication.</P> <P>On top of that, you've already said you don't care what an appropriate amount of time to test something is, you want it to be longer and longer and longer - 18 hrs of Test server testing spreadsheet entries wasn't enough time for you. (and it's about the most foolproof "patching" ever - about the same odds of it not working right as player data corrupting on a save)</P> <P>And yes, you got your point that you don't like the CU out a bushel of times already - it's not shocking that a healing type that was overpowered due to them expecting their abilities to mesh less well than they did started out lower powered when it went through while they shook out all the small changes needed from the CU.</P> <P>And even before the CU/DoF beta started patches usually happened on a similar frequency to current - the CU changed NOTHING about the patch cycle. (and in fact, if you look at a large amount of the recent patches, very few of them have much CU involvement)</P>

[Daish]

<P>Asz ....</P> <P>are you arguing that NO testing is necessary?  because, it looks like you are saying that if you put stuff on the Test server, there would be no benefit in testing it because the Live servers have different code on them....  Apparently, you never worked in software, there is a very systematic and understandable way you deal with version control when testing.  I'll give you the simple explanation, call the Live servers version 1.0 .. So, 1.0+ new bits is version 1.1.  Once 1.1 is finalized and tested, the live servers get update to 1.1, and the process starts again.</P> <P>And, it's funny that you can try to pull out a theoritical testing situation out of the air and say that no 'more' testing was necessary...   but, I can point to major changes that were put both on to the test servers, with little or no time to actually test,  or directly to the live servers, that were BROKEN when us paying customers got a chance to use them...</P> <P>So, you either have to argue one of these three points...</P> <P>- SOE's internal testing dept. and internal testing 'plan' does not work very well</P> <P>- SOE's lack of using the test server as a testing environment to evaluate the risk, let alone TEST the bits, is a bad idea</P> <P>- OR, it is simply NOT possible for SOE to release code without major game breaking and/or system breaking bugs, or code that will not need a successive set of further patching to actually make 'work'.</P> <P> </P> <P>Go ahead and explain how the gobblin gambling game was well served by the amazing amount of non-testing it was given?  </P> <P>If you actually look, something around 50% of all changes that SOE has put into the game have needed further work, patches, changes, or bug fixes..  just to make the code do what it was supposed to.  That rate is far above the 'normal' bug regression percentage the rest of the industry uses, and it is also further proof that thier methods of building software are to blame.</P> <DIV>Maybe you guys are forgetting that the GOAL of writing code is to make it so it actually WORKS CORRECTLY.  The goal isnt to just punch out code you can continue to work on for a few more months.  If the code never actually met the design requirements, then it should never have been released.</DIV> <DIV> </DIV> <DIV>Imagine if your supermarket cash registers got the wrong price on your groceries 30% of the time, and only that AFTER the person tried to scan the item for about 5 minutes.  Would you expect them to simply say...  well, as long as you continue to pay for our mistakes, we'll get right to trying to fix the problem, but it will take a few months...  come again!</DIV><p>Message Edited by Daishar on <span class=date_text>10-12-2005</span> <span class=time_text>03:49 PM</span>