Devs - I'd like to hear your tales of Crunch please



  •  For my MSc dissertation I'd like to hear you tales - good or bad - of
    crunches you've experienced. All completely anonymous, I'd just like to
    hear of circumstances surrounding the crunch; how it happened,
    duration, any similarities to other projects, how it could have been
    avoided etc.

    Any information you can give me would be extremely useful!

    Please feel free to contact me by email:

    b.worrall (at) warwick (dot) ac (dot) uk

    or post here if you have any questions.

    Thanks!





  •  In what way? The games industry has no official governing body, nor any real academic attention as demonstrated by lack of journals on this subject. Internet forums are an excellent source of data for my research; anonymous tales of crunches that let me tally causes and circumstances.

    If this is a problem then I apologise.



  • Crunch time is bad. It's a symptom of diseased structure.

    If your company regularly does "crunches" to the point where it's an actual thing that you call "crunch time", then you have a problem and you need to fix it. The problem is structural and likely pervades every level of the company, from sales' impossible promises, PMs not actually doing their job, programmers not paying attention to how time works in the real world, to crap resource allocation of the planning department.

    But don't forget to investigate programmer's ability to estimate time. They build it. They get to tell you how long it takes. Record what they say. Compare it to the real time it took. Ask what happened if it took significantly longer or shorter. Demand that they tell you when they hit 75% of their estimated time.

    Track, document, record, observe, reflect. Don't wing it.

    Also, we don't do crunches. Crunches are a bandage to patch up the flaws so the pus in the festering wound remains invisible.

    ugh **|:-|

    I have spoken.

    @jeffmcbiscuit said:

    The games industry has no official governing body, nor any real academic attention as demonstrated by lack of journals on this subject. Internet forums are an excellent source of data for my research

    This is for you. There are many episodes there on the practical development side of games. It's an excellent series.



  • @jeffmcbiscuit said:

    In what way? The games industry has no official governing body, nor any real academic attention as demonstrated by lack of journals on this subject. Internet forums are an excellent source of data for my research; anonymous tales of crunches that let me tally causes and circumstances.

    Wow I really hate you and it's only been 2 posts.



  • Thanks for that man.

    It's interesting you use the 'wound' analogy, as that runs throughout my dissertation. Thanks for the link, I'll check those vids out.



  • @jeffmcbiscuit said:

    In what way? The games industry has no official governing body, nor any real academic attention as demonstrated by lack of journals on this subject. Internet forums are an excellent source of data for my research; anonymous tales of crunches that let me tally causes and circumstances.

    Ok for a slightly less hate-filled response...

    1) Assuming that "crunch time" only exists for video game projects is both weird and wrong. Where did you get that?

    2) Teams that have "crunch time", regardless of product, are broken. The reason you see more crunch time in video games is because video games are usually written by shitty engineers who couldn't code their way out of a wet sock. (Seriously, download the Creation Kit from Steam and take a look at what Bethesda thinks is an appropriate quality level for public-facing tools. They wrote Skyrim with hundreds of multi-stage quests that all interact with each other in strange unpredictable ways, and no only did they not write any unit tests but their tools don't even have the capability of running unit tests on quests. We're not talking about the cream of the crop here, engineering-wise.)

    3) Modern development methodologies are primarily focused around getting rid of, or at least reducing, crunch time. So you won't see a team using up-to-the-minute methodologies suffering from crunch time, generally speaking. You will see the Bethesda's of the world, or the IBM's of the world, or the Siemen's of the world have crunch time because they're still writing code the way they did in 1985. Bethesda out of ignorance and hiring eager kids, the rest out of employing only useless morons who can wear suits but have few other qualifications.

    4) My only personal experience with crunch time is when I was at Microsoft doing network multiplayer testing for Xbox 360, and we had to give up our lab for a few days so the Gears of Wars guys could work through some last-minute critical issues. They all had terrible BO.



  • @blakeyrat said:

    Ok for a slightly less hate-filled response...
     

    This role reversal was fun!

    I was a bit pissed because three coworkers AROUND ME were somewhat pointing at eachother and at thin air because we have some process problems.

    Somehow, people think that a process problem is never their problem.

    @blakeyrat said:

    They all had terrible BO.

    At first I thought BO might stand for some kind of business practise or whatever and then I realized it's true meaning.

    @blakeyrat said:

    Seriously, download the Creation Kit from Steam and take a look at what Bethesda thinks is an appropriate quality level for public-facing tools.

    I watched some videos showing off the newest Unreal engine creation software, and it seemd pretty goddamn awesome, so I guess that's a good thing. Long, long ago, I tried the Unreal 2 editor, and it was complete junk that crashed a lot.



  • Sorry Blakey. I think.

    Thanks for your opinion though. I'm not saying crunch is only a game dev problem, but that it's particularly bad there. It's the reasons why that interest me. Your talk of outdated methods and tools is interesting. Some of what I'm finding is that people want to blame 'scheduling' and leave it there because then you can usually point to a person and blame them rather than admit to there being a problem with a process or even the whole methodology which is more work, and in the long run leads to a suspicion among teams that anyone even suggesting 'methodologies' just doesn't get it.



  • @jeffmcbiscuit said:

    It's the reasons why that interest me.
     

    AOL.

    I'd like to see:

    1. causes of crunch - some have been mentioned (over-commitment, poor project management, lack of progress tracking and control, unrealistic timescales, scope creep, crap time management) but it's the post-mortem of companies that experience it and their attitudes to it which I'm interested in. Does the videogame industry accept that crunches are the norm? 
    2. effect of crunch - from poor quality results, increased risk, bad reviews, impact upon business reputation to elevated stress levels, burnout and staff turnover. I know some individuals believe they produce their best results when
      under pressure, but empirical evidence often shows the results are rushed, of lower quality and sometimes cost more in the long run to fix than if they'd have put back delivery dates and "done it properly".
    3. reducing crunch - have any organisations learned by their mistakes? Who recognises crunch as being a bad thing and taking steps to remove it or reduce it? Which industries (catering, marketing, media) thrive on crunch? 



  • @Cassidy said:

    Does the videogame industry accept that crunches are the norm?
    I guess you could say so, yes.



  • @Ragnax said:

    @Cassidy said:
    Does the videogame industry accept that crunches are the norm?
    I guess you could say so, yes.

    There's more to the industry than just EA. I've worked in two British independent game studios and in neither did I work more than 40 hours a week with the exception of some paid overtime on two weekends.

    I hope the OP is categorising the comments received by country because US and EU labour law are rather different.



  •  Ragnax, I'm well aware of EASpouse. :)

    PJT33, I understand there are definitely differences in the way the industry operates in different locales, and labour laws reflect this to an extent. Do you think there was anything that your employers were doing that helped mitigate or avoid crunching?



  • @jeffmcbiscuit said:

    Do you think there was anything that your employers were doing that helped mitigate or avoid crunching?

    You just have to let the Grape Nuts soak in the milk for a few minutes.



  • @boomzilla said:

    @jeffmcbiscuit said:
    Do you think there was anything that your employers were doing that helped mitigate or avoid crunching?
    You just have to let the Grape Nuts soak in the milk for a few minutes.
    Was wondering whether I was the only one to think that way. My thought at the time was "Close the packet so they don't dry out to start with."



  • Just to drop in a link I've found two days ago... http://trenchescomic.com/tales - I don't give a crap about the comic, but if you think game developers are screwed up, check out the stories from the job Blakey once enjoyed (in the lone company that did not treat QA like trash, apparently).



  • The nice thing about my job is that I worked for Microsoft and not the game developer, and we basically had a "pass/fail" scoring mechanism. Some games had awful, awful, awful, awful netcode-- one 4-player party game wouldn't work with less than 256 kbit of bandwidth even though the screen was only showing something extremely extremely extremely similar to this. I won't name the game though! The minimum bandwidth for Xbox network certification was, IIRC, 64 kbit, but we tested down to 32k just in case people were running behind modems. Technically unsupported, but ~75% of Xbox games did ok anyway as long as there was no packet loss. I used to joke that that party game was sending a full copy of Stephen King's library alongside every game packet.

    Despite that, we'd occasionally test actual Microsoft-developed games, and in those situations we had direct access to the bug tracker and were encouraged to input as many bugs as we could, even if they had nothing to do with the netcode specifically. We got in the credits if we put in bugs that they fixed. (Although I got screwed on one very popular game, where I put in a critical bug they didn't previously have and my name's not in the credits, the bastards.)



  • @PJH said:

    @boomzilla said:
    You just have to let the Grape Nuts soak in the milk for a few minutes.
    Was wondering whether I was the only one to think that way. My thought at the time was "Close the packet so they don't dry out to start with."
    Well, the worst Crunch scenario was the time Jean LaFoote, the Barefoot Pirate, managed to capture the crew of the Guppy:




  • @jeffmcbiscuit said:

    Do you think there was anything that your employers were doing that helped mitigate or avoid crunching?

    They weren't working to deadlines set by marketing.



  • @boomzilla said:

    @jeffmcbiscuit said:
    Do you think there was anything that your employers were doing that helped mitigate or avoid crunching?

    You just have to let the Grape Nuts soak in the milk for a few minutes.

     

    I do that to the veal but it doesn't help.



  • @dhromed said:

    I do that to the veal but it doesn't help.
     

    The veal's gotta be dead first - else it escapes from your milk bath.

    Less fun, admittedly.. but a greater chance of success, nonetheless.



  • @pjt33 said:

    @jeffmcbiscuit said:

    Do you think there was anything that your employers were doing that helped mitigate or avoid crunching?

    They weren't working to deadlines set by marketing.


    This. The root of the problem is always marketing.  It's especially notable in the games industry because there's so much pressure to ship your game yesterday, although I don't quite understand why.  It's not like there are three companies all working on different versions of FarmVille and the first one shipped wins.  Sure, there are always copycats, but from what I can see games win market share not for being first but for being best.  The players know what's fun and great, and they talk and get their friends to play the best games.

     Outside of the game industry, where I work, I still see this problem.  I'm currently working for a pretty good company with internal IT that supports the business (not selling software) and we have pretty good software practices.  We have QA and risk review meetings and great teamsmanship and collaboration between teams.  But starting this year management decided to try to sell a new web service to the general populace (B2C), and wouldn't you know it?  They refuse to give us written requirements because "there's no time" and instead shout requirements over the wall to us on a daily basis.  They also call meetings to ask us if features are possible - but don't start work because we don't know if we want it yet - and the next thing we know they're demanding to know why the feature isn't ready for tomorrow's release.  They're planning releases - I swear to god - two days in advance and expecting QA to be able to do full integration and regression testing.  Basically a bunch of assholes and it's all going to go down in flames because they're using the excuse of beating the market to throw away decades of proven software development and project management practices.

     



  • @jetcitywoman said:

    Basically a bunch of assholes and it's all going to go down in flames
     

    So.. we expect you to be chasing snoofle's tail with your tales soon, then?



  • @Cassidy said:

    So.. we expect you to be chasing snoofle's tail with your tales soon, then?
     

    I think she meant it more generally.



  •  I used to work at a place where I could post fantastic WTF's to match Snoofle's, except the code samples were all in Cobol or assembler (specifically, Vax Macro) so two thirds of you all wouldn't have a clue.  :-)  But now in my current gig, I'm a business analyst so they don't let me even glimpse the code.  All my WTF's will be in the business end of things, like my little rant about that project.  And as I said, for the most part we do things well.  



  • @jetcitywoman said:

     I used to work at a place where I could post fantastic WTF's to match Snoofle's, except the code samples were all in Cobol or assembler (specifically, Vax Macro) so two thirds of you all wouldn't have a clue.  :-)  But now in my current gig, I'm a business analyst so they don't let me even glimpse the code.  All my WTF's will be in the business end of things, like my little rant about that project.  And as I said, for the most part we do things well.  


    Hmm....

    Do you live in central Pennsylvania, by any chance? If you do not, then I think I have met your career doppelganger. Isn't that frightening!



  • @jetcitywoman said:

    so two thirds of you all wouldn't have a clue.
     

    I'm sure someone could make a sexist remark about your divisions. We're not all young in here, you know.

    (and yes, I knew VAX/VMS - used it at University. Loved Unix when it replaced our ageing VMX system)



  •  @Xyro said:

    Do you live in central Pennsylvania, by any chance? If you do not, then I think I have met your career doppelganger. Isn't that frightening!

    Nope, northern Virginia.  Not surprised there are more out there like me.  I'm your average "senile" (in my late 40's) former software engineer.



  •  @Cassidy said:

    and yes, I knew VAX/VMS - used it at University. Loved Unix when it replaced our ageing VMX system)

    Unless you used it as recently as two years ago, I'm not impressed.  Everybody and their dog has told me "I used VMS at university".  99.9% of them went on to admit they'd not seen it since, and so were all completely surprised to learn

    a)  It's still "out there" being sold (I won't go so far as to say marketed, because it pretty much isn't)

    b)  Is used in many banking/finance and other critical systems industries

    People tend to think of slow systems running Cobol or Fortran with dumb terminal interfaces when they think of VMS.  Everybody is shocked to learn that there are Java and Apache builds for VMS and you actually CAN use it for everything that you can use *nix for.  It just seems to be an embarrassment to the HP corporation (and Compaq before them) for reasons nobody in the VMS world has ever been able to figure out.  Best we could figure, and it's weak but all we have, is that since it's not Windows, HP doesn't want to admit anything to do with it.  But then, HP has many other issues with their business, and the VMS situation pales in comparison to the rest of it.



  • @jetcitywoman said:

    99.9% of them
     

    Erm.. yup. I'm included in that bunch. Mainly because my (only) experiences of VMS were terminal-based, and my first experiences of Unix were graphical-based.

    In recent times (read: 10 years ago) we had a Unix Operator that was a VMS person pressganged into operational support and used to moan and swear about the shortcomings of Unix. Given the shit system we used, I didn't blame him.


Log in to reply
 

Looks like your connection to What the Daily WTF? was lost, please wait while we try to reconnect.