ferroz Posted February 6, 2012 Share Posted February 6, 2012 Bug hunting is VERY decieving. The process (from my experience) goes a little like this: 1) User sees bug and reports it 2) QA sees ticket and tests to confirm the bug exists (this can take a long time if the bug reproducible on only certain machine configs) 3) QA pushes to producers, which discuss how important the bug is 4) Producers then decide in which sprint/cycle the bug should be handled and by which team 5) The time comes for the bug to be fixed by the programmers. And when that time comes to fix the bug, a number of things can go wrong. - Dev can't reproduce the issue on their machine (quite common) - Dev discovers that the bug has to do with another module with a different team (also very common) Bugs that seem so simple to fix have taken me weeks to work out. Not to mention the tireless testing and deployment issues one needs to face before seeing their bug-fixing code in production. The bigger the team, the more cumbersome it can be to turn around a bug. That being said, some can be easy for a programmer to fix, might take a couple hours for him to track down the offending code and fix it. But even for an easy fix like that it can take weeks or months before it is deployed in game.Just a couple of comments: If you can spend a few minutes and try to reproduce the bug a few times, and include some information on it's reproducibility, that helps quite a bit. Even something as simple as "I tried it 6 times and it happened 3 times" really does mean something. This keeps the bug from incorrectly getting filed into a chaff bin (as in: cannot reproduce). If you take a few minutes and write up a very verbose set of steps to reproduce, that helps immensely. Again, it helps make sure the bug doesn't wind up in the circular file. Link to comment Share on other sites More sharing options...
Jeremy_Saenz Posted February 6, 2012 Author Share Posted February 6, 2012 I forgot to answer this one ~What would you tell the developers if you had the ability to speak to them face to face? I'm a workflow nut. So I would love to know how they manage their art and content pipeline for this game, and what changes they made to HERO to accomodate that flow. I'm hoping for very efficient content releases based on what they have been touting with their workflow Link to comment Share on other sites More sharing options...
Jeremy_Saenz Posted February 6, 2012 Author Share Posted February 6, 2012 Just a couple of comments: If you can spend a few minutes and try to reproduce the bug a few times, and include some information on it's reproducibility, that helps quite a bit. Even something as simple as "I tried it 6 times and it happened 3 times" really does mean something. This keeps the bug from incorrectly getting filed into a chaff bin (as in: cannot reproduce). If you take a few minutes and write up a very verbose set of steps to reproduce, that helps immensely. Again, it helps make sure the bug doesn't wind up in the circular file. Yes, Most QA people actually like this, it allows them to pinpoint the system it is in, and the risk that the bug adds to the overall product. Developing a game live like this is all about managing risk! Link to comment Share on other sites More sharing options...
Sabarok Posted February 6, 2012 Share Posted February 6, 2012 nice thread. I'm a programmer myself, so I can pitch in a little. I'm not a game dev and I've never worked in an office with more than 2 other programmers who worked the same code I did, so I don't have the same understanding of the process of larger groups. Based on your replies, I believe you when you say you do. It was nice to see your explanation on the bug handling from the perspective of a company larger than any I've ever worked for. I maintain software I wrote entirely on my own, and a handful of bugs have taken me months to track down simply because I could never reproduce them. I've even had it happen where I stumbled upon how to reproduce it while working on something else. Then you have the bugs with more than 1 cause. Developer gets a list of steps to reproduce the bug, which he tracks down and fixes and confirms it by reproducing the steps and sees that the bug doesn't happen, so the dev pushes it out as fixed. End users find the bug still exists through a different set of steps, and think that nothing was ever done. The rule of thumb when fixing bugs is this: If the dev can cause it, the dev can fix it. If it can't be reproduced on the dev's machine, the only thing the dev can do is take hours examining possible locations for the bug and hoping to spot something that looks out of the ordinary, and then crossing one's fingers and hoping the bug stops getting reported. Usually, the bug would just be sent back as "non-reproducable" in order to get more information on how to cause the bug. That's why we've seen some devs on the forums looking for information, such as the guard bug. Non-devs might be thinking that the only reason to reproduce the bug is to confirm its existence, but you'd be wrong. By knowing the steps it takes to cause the bug isolates with great precision where the bug is. But the most important reason to be able to reproduce it is so that once the dev does what they think is the fix, they can confirm its fixed by trying to cause the bug to happen again. How much QA ties into this process in a company as big as Bioware, I wouldn't presume to know. Link to comment Share on other sites More sharing options...
Jeremy_Saenz Posted February 6, 2012 Author Share Posted February 6, 2012 Sounds like you'd fit in well with the SWTOR team. I honestly don't think you're qualified to take on the task you've set yourself, then again neither are most of the devs who respond here. Hope you do better! TBH I probably am not qualified to answer everyones questions, nor am I qualified to convince unsatisfied people to enjoy this game. Some people do enjoy getting clarification on how an industry works though, and I can answer some of the issues I have been seeing on the forums, so give me your best shot Link to comment Share on other sites More sharing options...
osuaaron Posted February 6, 2012 Share Posted February 6, 2012 (edited) Farmville and the Sims are two of the most popular and sustaining games ever made. Why are some of the 'social elements' from these games not considered when producing an MMO these days? For example, SWG had "Farmville like" mechanics with its factories and farms. You could go anywhere and put down a machine to farm vapor. You could also build a house and turn it into your museum - or your shop if you wanted to. Why are these elements gone from MMOs when they clearly served as good "end game" content to keep people busy and not demanding another raid or new loots? I mean - there are times I just don't feel like murdering everything, but I still want to log in. I avoid SWTOR when I'm not in a killing mood. Edited February 6, 2012 by osuaaron Link to comment Share on other sites More sharing options...
Jeremy_Saenz Posted February 6, 2012 Author Share Posted February 6, 2012 nice thread. I'm a programmer myself, so I can pitch in a little. I'm not a game dev and I've never worked in an office with more than 2 other programmers who worked the same code I did, so I don't have the same understanding of the process of larger groups. Based on your replies, I believe you when you say you do. It was nice to see your explanation on the bug handling from the perspective of a company larger than any I've ever worked for. I maintain software I wrote entirely on my own, and a handful of bugs have taken me months to track down simply because I could never reproduce them. I've even had it happen where I stumbled upon how to reproduce it while working on something else. Then you have the bugs with more than 1 cause. Developer gets a list of steps to reproduce the bug, which he tracks down and fixes and confirms it by reproducing the steps and sees that the bug doesn't happen, so the dev pushes it out as fixed. End users find the bug still exists through a different set of steps, and think that nothing was ever done. The rule of thumb when fixing bugs is this: If the dev can cause it, the dev can fix it. If it can't be reproduced on the dev's machine, the only thing the dev can do is take hours examining possible locations for the bug and hoping to spot something that looks out of the ordinary, and then crossing one's fingers and hoping the bug stops getting reported. Usually, the bug would just be sent back as "non-reproducable" in order to get more information on how to cause the bug. That's why we've seen some devs on the forums looking for information, such as the guard bug. Non-devs might be thinking that the only reason to reproduce the bug is to confirm its existence, but you'd be wrong. By knowing the steps it takes to cause the bug isolates with great precision where the bug is. But the most important reason to be able to reproduce it is so that once the dev does what they think is the fix, they can confirm its fixed by trying to cause the bug to happen again. How much QA ties into this process in a company as big as Bioware, I wouldn't presume to know. Nice post, I especially would like to elaborate on the One bug, Multiple causes effect. I pride myself on being a great programmer, but I have been chewed out before from a bug that I (truthfully) fixed days or weeks before the same issue came up again, only with a completely different cause. This is not necessarily my fault, because I did fix the issue that was causing the undesired functionality. Little did I know that there was another block of code that was also offending. Link to comment Share on other sites More sharing options...
Jeremy_Saenz Posted February 6, 2012 Author Share Posted February 6, 2012 Farmville and the Sims are two of the most popular and sustaining games ever made. Why are some of the 'social elements' from these games not considered when producing an MMO these days? For example, SWG had "Farmville like" mechanics with its factories and farms. You could go anywhere and put down a machine to farm vapor. You could also build a house and turn it into your museum - or your shop if you wanted to. Why are these elements gone from MMOs when they clearly served as good "end game" content to keep people busy and not demanding another raid or new loots? I mean - there are times I just don't feel like murdering everything, but I still want to log in. I avoid SWTOR when I'm not in a killing mood. Great question! I honestly think it is all about politics. Financially, SWTOR has taken enough risks. And by "Risks", I mean anything that is not in WoW! The unfortunate thing is that the amount of risks these games take are not decided by the programmers or the designers, they are decided by the shareholders/investors. I would love to see more social elements in MMO's, I don't have numbers but I would be willing to assume that putting a ton of those features in may not make the most financial sense. We may never know Link to comment Share on other sites More sharing options...
Sabarok Posted February 6, 2012 Share Posted February 6, 2012 I pride myself on being a great programmer, but I have been chewed out before from a bug that I (truthfully) fixed days or weeks before the same issue came up again, only with a completely different cause. I don't think anyone can call themselves a programmer until they've had that happen to them I had it happen to me just last week. One of the employees was complaining about an issue, so I resolved it, or at least I thought I had. Later, he was still complaining about it wondering when I was going to get around fixing it once and for all and I looked over his shoulder and saw the very error I thought I had fixed. Same issue, different cause. If there's one thing you can count on end-users to do it's to figure out how to use the software in ways never planned for, though I do try and make it as hard on them as possible Link to comment Share on other sites More sharing options...
osuaaron Posted February 6, 2012 Share Posted February 6, 2012 Great question! I honestly think it is all about politics. Financially, SWTOR has taken enough risks. And by "Risks", I mean anything that is not in WoW! The unfortunate thing is that the amount of risks these games take are not decided by the programmers or the designers, they are decided by the shareholders/investors. I would love to see more social elements in MMO's, I don't have numbers but I would be willing to assume that putting a ton of those features in may not make the most financial sense. We may never know Thank you, but again, two of the most popular games ever designed. How on earth is that a "risk" in any form or fashion? Just a rhetorical question, thanks for your reply. Link to comment Share on other sites More sharing options...
Jeremy_Saenz Posted February 6, 2012 Author Share Posted February 6, 2012 Thank you, but again, two of the most popular games ever designed. How on earth is that a "risk" in any form or fashion? Just a rhetorical question, thanks for your reply. It's a risk because it is (In some ways, not all) a conflict of interest. This game was developed to be a Theme Park type mmo. Declaring that allowed the developers to scope out their content according to how the themepark mmo genre has defined it. The more dynamic model of having a sandbox is already a big undertaking in itself. For a dev team to try and develop a hybrid game would be a HUGE risk. I'm not claiming it couldn't be done, but I'm sure shareholders wouldn't like the idea when the Game Director saya "And it is only going to take 10 years until release!" Link to comment Share on other sites More sharing options...
MillionsKNives Posted February 6, 2012 Share Posted February 6, 2012 Thank you, but again, two of the most popular games ever designed. How on earth is that a "risk" in any form or fashion? Just a rhetorical question, thanks for your reply. It really comes down to design decisions. BioWare decided it would do a heavily story based game, which by definition has to be linear to a point. One cannot possibly design a story that has a player doing whatever they want. It's just impossible to manage that many branches. You do have choices, you can skip planets and decline or accept specific quests, as well as take them in varying directions, but it is all designed by someone. In order to get the story to the polish that it is (I am by no means saying that there aren't places to improve) that is one of the areas they had to spend a lot of their resources, between their programmers, designers, artists, writers, actors, etc. To make a polished portion of the game into a sandbox those resources would have to be taken from elsewhere. It's better to have one complete portion of a game than multiple diverse portions that are incomplete. What you're talking about is really what distinguishes a Themepark MMO with a Sandbox MMO. The former takes you through interesting places that the game has for you to see, and the other creates a more open world where the players choose what to do. This isn't to say that these things can't be added to the benefit of the game, even though it's designed as heavily a Themepark game. Because MMO's are constantly changing it could easily be worked in at some point. Whether BioWare feels this is important enough to add we'll have to find out. Link to comment Share on other sites More sharing options...
Jeremy_Saenz Posted February 6, 2012 Author Share Posted February 6, 2012 Bummer, it looks like this thread has been moved to the community forum. Link to comment Share on other sites More sharing options...
Recommended Posts