Richard Fabian | 2 Apr 2007 13:13
Picon
Gravatar

tasking

I think this is probably the best place to submit this, as the audience is right, but something in me thinks that it's verging on "off-topic"

Our producer has asked me to help find out of there are any software solutions to a problem we are facing at the moment, and have faced on every project we've worked on.
The programmers are happy building lists of tasks, assigning staff based on capability and favourability, moving dates around after assessing critical paths, overallocating and looking after when things need to be done in time for other departments and requesting things to be done by a certain time from them too... but no-one else is happy doing all this "scheduling" and is saying that it is the producers job. Unfortunately this isn't possible as the producer is not "games industry", and doesn't innately understand some of the implications of certain elements of game development and we've seen some horrific gaffs because of an unnoticed dependency or a task that makes no sense (its like production through a "chinese room" engine).

so, the question really is: is there any software or peopleware process that will eliminate all the innadequacies of the system we have at the moment. We were hoping for some nice dependency driven todo list software or something, but we'd be equally happy with a peopleware method of fixing our problems.

--
fabs();
Just because the world is full of people that think just like you, doesn't mean the other ones can't be right.

_______________________________________________
sweng-gamedev mailing list
sweng-gamedev <at> lists.midnightryder.com
http://lists.midnightryder.com/listinfo.cgi/sweng-gamedev-midnightryder.com
Robert Blum | 2 Apr 2007 14:37
Picon

Re: tasking

> I think this is probably the best place to submit this, as the  
> audience is right, but something in me thinks that it's verging on  
> "off-topic"

It probably is - list-admins, shut us down if you think so ;)

> Our producer has asked me to help find out of there are any  
> software solutions to a problem we are facing at the moment, and  
> have faced on every project we've worked on.
> The programmers are happy building lists of tasks, assigning staff  
> based on capability and favourability, moving dates around after  
> assessing critical paths, overallocating and looking after when  
> things need to be done in time for other departments and requesting  
> things to be done by a certain time from them too... but no-one  
> else is happy doing all this "scheduling" and is saying that it is  
> the producers job.

That depends a bit on how your team is structured - but overall, it's  
the producers job to make sure there *is* a schedule. Creating that  
schedule is best left to the people who know how long things are  
going to take, IMHO. For us, this is usually the lead's task. He (or  
she) solicits times for tasks from the team members who are likely to  
work on them, ferrets out dependencies, and hands the whole thing to  
the producer.

Making those lists is not necessarily the producers task. Putting it  
into Project and making pretty GANTT charts *should* be his - and  
this is what most people really object to when they complain about  
"scheduling"

> Unfortunately this isn't possible as the producer is not "games  
> industry",

That might be part of the problem.... What exactly is your producers  
role? (I've been at places where it was boiled down to "produce me a  
coffee" ;)
>
> so, the question really is: is there any software or peopleware  
> process that will eliminate all the innadequacies of the system we  
> have at the moment. We were hoping for some nice dependency driven  
> todo list software or something, but we'd be equally happy with a  
> peopleware method of fixing our problems.

This does not sound like a problem you solve in software. It sounds  
like you need to find out what everybody thinks their  
responsibilities are, and then closing the gaps. I.e. "peopleware".  
Maybe you need to hire an asst. prod. who *is* familiar with game  
development. Maybe you need a "MSProject Wrangler". Maybe you need to  
have a word with the leads. It's hard to tell without being there...

  - Robert

_______________________________________________
sweng-gamedev mailing list
sweng-gamedev <at> lists.midnightryder.com
http://lists.midnightryder.com/listinfo.cgi/sweng-gamedev-midnightryder.com

Will Vale | 2 Apr 2007 14:45
Favicon
Gravatar

Re: tasking


I've seen the "that's the producer's job, just leave us alone and we'll  
make cool stuff" attitude before, and it's hard to get around. Even if  
your producer was very games-aware, they'll still need input and advice  
 from the different disciplines when working out schedules. After all, if  
your producer knew everything about the ins and arts of art, design and  
code you could fire your leads and let the producer do it all :)

On the software front, bug tracking software (something lightweight - we  
used PR-Tracker) can be a good way of managing task lists. It should  
reduce busywork for team members, by providing a simple way to receive  
tasks, send them on, comment on them, mark them as completed etc. For  
mangement, you can generate reports to get an overview of progress, enter  
high level tasks to be broken down into day-to-day tasks etc. You may not  
get explicit handling for dependencies this way, but (for example) an  
artist waiting on code could forward their task to the coders, getting it  
back when the necessary feature has been completed.

One advantage of this is that you may already have bug tracking software,  
and a team which is used to using it.

Depending on your publisher relationship the producer may need to turn the  
task lists and reports into a Gantt chart from time to time, which can be  
a pain, but that's the kind of thing producers are usually very good at.

I also recall that one of the other teams at Argonaut rolled their own  
system which revolved around having interactive task lists on everyone's  
desktop as a web gadget. Tasks could be ticked off when completed,  
commented on, etc. Apparently it was quite successful - a big advantage  
being that team members always had their tasks visible without having to  
remember to run a program. We did some experiments with automatically  
publishing our PR-Tracker-generated task lists to desktops as HTML, but  
without the interactivity it didn't offer that much benefit.

Whatever software you use, you still need to get your team to share  
responsiblity for the schedule, and that's hard. Getting the leads (or  
whoever the senior or inspirational people are in your team structure) on  
your side first is probably the place to start. It may also be worth  
thinking about splitting the lead or equivalent roles up into a creator  
and a planner role - doing both can be a strain on one person.

Reading back over this I'm not sure if it totally answers your question -  
I was thinking more about tracking and assigning work than planning future  
work - but maybe something there will be useful.

Cheers,

Will

_______________________________________________
sweng-gamedev mailing list
sweng-gamedev <at> lists.midnightryder.com
http://lists.midnightryder.com/listinfo.cgi/sweng-gamedev-midnightryder.com

Kent Quirk | 2 Apr 2007 14:48

Re: tasking

Richard Fabian wrote:
> I think this is probably the best place to submit this, as the 
> audience is right, but something in me thinks that it's verging on 
> "off-topic"
>
> Our producer has asked me to help find out of there are any software 
> solutions to a problem we are facing at the moment, and have faced on 
> every project we've worked on.
> The programmers are happy building lists of tasks, assigning staff 
> based on capability and favourability, moving dates around after 
> assessing critical paths, overallocating and looking after when things 
> need to be done in time for other departments and requesting things to 
> be done by a certain time from them too... but no-one else is happy 
> doing all this "scheduling" and is saying that it is the producers 
> job. Unfortunately this isn't possible as the producer is not "games 
> industry", and doesn't innately understand some of the implications of 
> certain elements of game development and we've seen some horrific 
> gaffs because of an unnoticed dependency or a task that makes no sense 
> (its like production through a "chinese room" engine).
>
> so, the question really is: is there any software or peopleware 
> process that will eliminate all the innadequacies of the system we 
> have at the moment. We were hoping for some nice dependency driven 
> todo list software or something, but we'd be equally happy with a 
> peopleware method of fixing our problems.
If you're in the mood to take a look at your process, you might look 
into Scrum. It's been known to work in a variety of situations.

Wikipedia has two articles on it -- they're different enough that 
they're both worth reading:

http://en.wikipedia.org/wiki/Scrum_%28development%29
http://en.wikipedia.org/wiki/Scrum_%28management%29

And if you're seriously interested, get Ken Schwaderer's book. It's 
short and simple, but it explains the process and the pitfalls pretty well.

    Kent

--

-- 
------------------------------------------------------------
Kent Quirk           I'm making a game about global warming.
Game Architect                        Track the progress at:
CogniToy                http://www.cognitoy.com/meltingpoint

_______________________________________________
sweng-gamedev mailing list
sweng-gamedev <at> lists.midnightryder.com
http://lists.midnightryder.com/listinfo.cgi/sweng-gamedev-midnightryder.com

Adam Martin | 2 Apr 2007 15:02

Re: tasking

On 02/04/07, Kent Quirk <kent_quirk <at> cognitoy.com> wrote:
> Richard Fabian wrote:
> > I think this is probably the best place to submit this, as the
> > audience is right, but something in me thinks that it's verging on
> > "off-topic"
> >
> > Our producer has asked me to help find out of there are any software
> > solutions to a problem we are facing at the moment, and have faced on
> > every project we've worked on.
> > The programmers are happy building lists of tasks, assigning staff
> > based on capability and favourability, moving dates around after
> > assessing critical paths, overallocating and looking after when things
> > need to be done in time for other departments and requesting things to
> > be done by a certain time from them too... but no-one else is happy
> > doing all this "scheduling" and is saying that it is the producers
> > job. Unfortunately this isn't possible as the producer is not "games
> > industry", and doesn't innately understand some of the implications of
> > certain elements of game development and we've seen some horrific
> > gaffs because of an unnoticed dependency or a task that makes no sense
> > (its like production through a "chinese room" engine).

The simple (read: cheeky) answer would seem to be (given you say you
have this problem on every project) - why do you keep hiring the wrong
person for the Producer role? You yourself say the team expects the
producer to do this, but you've got a producer who doesn't.

Either they're right, and you need to get an appropriate (not
necessarily better or worse, but just "correct for the job") person
instead, or they're wrong and you need to find out why they have this
belief and then deal with it. IME usually if the team is wrong about
something like this (e.g. being unreasonable) then there's a more
subtle underlying cause that you can find out within half an hour of
talking to people - the same issue will keep popping up from short
conversations with different people. Again, IME, that's usually an
inexperienced direct manager or a malicious/incompetent indirect
manager (e.g. "our team's fine, but every time we try to do this, we
get screwed by Team Y [who are doing it because their manager has a
secret agend of their own]") somewhere.

> >
> > so, the question really is: is there any software or peopleware
> > process that will eliminate all the innadequacies of the system we
> > have at the moment. We were hoping for some nice dependency driven
> > todo list software or something, but we'd be equally happy with a
> > peopleware method of fixing our problems.
>
> If you're in the mood to take a look at your process, you might look
> into Scrum. It's been known to work in a variety of situations.

(if you're still reading this, after my slightly tongue-in-cheek
comments above ;))

I second SCRUM as an excellent thing, BUT: it's very much *not* a
panacea, it only solves a certain set of problems, and even then it's
a very "junk in, junk out" kind of solution: you can follow every rule
to the letter, and yet find SCRUM makes things no better (in fact
worse even) if you fail to internalise it on an individual level. I
think the quote Schwaber uses often is "it's business as usual",
meaning that teams adopt SCRUM, but then carry on acting in the ways
they used to, thereby entirely devaluing SCRUM. SCRUM is more a
mindset than an actual PM methodology. That can be hard to accept, so
be careful not to adopt it naively.

Given the way you describe the team, it sounds like they may certainly
benefit from using SCRUM. Sounds like a good match. But ... if they
have a problem with a manager, that person could easily deflate SCRUM
if they don't themselves internalise it.

> And if you're seriously interested, get Ken Schwaderer's book. It's
> short and simple, but it explains the process and the pitfalls pretty well.

Yeah - the first one (there's a few others now, IIRC - all good, but
the first one is the one you need to read first :)).

Also, I recommend the IGDA Production SIG mailing list - free to join,
and then when you're on there go to the archives (have to be a list
member first) and search last month and the couple of months before
for SCRUM conversations. There's been a few recently, with a lot of
good links and commentary.

Finally ... anyone here who's doing SCRUM or thinking about it for
games dev, I'd love to talk to you. I've done it before for small
teams (<10 people) and I'm moving a whole studio over to it now - I've
spent a lot of time tracking other recent studio-wide adoptions (e.g.
Bioware, Sulake) but would love any additional advice or warnings, and
would be very interested in some mutual support with anyone else who's
going through / about to go through a similar conversion at the moment
:)

Adam
_______________________________________________
sweng-gamedev mailing list
sweng-gamedev <at> lists.midnightryder.com
http://lists.midnightryder.com/listinfo.cgi/sweng-gamedev-midnightryder.com

Richard Fabian | 2 Apr 2007 15:23
Picon
Gravatar

Re: tasking

I have the book, i'd love to implement it, but scrum got a bad name at our company when an attempt to try it out got assassinated by managment. That and the publishers we work with aren't flexible enough to try anything they perceive as "new".

On 4/2/07, Kent Quirk <kent_quirk <at> cognitoy.com> wrote:

If you're in the mood to take a look at your process, you might look
into Scrum. It's been known to work in a variety of situations.

Wikipedia has two articles on it -- they're different enough that
they're both worth reading:

http://en.wikipedia.org/wiki/Scrum_%28development%29
http://en.wikipedia.org/wiki/Scrum_%28management%29

And if you're seriously interested, get Ken Schwaderer's book. It's
short and simple, but it explains the process and the pitfalls pretty well.

    Kent

--
------------------------------------------------------------
Kent Quirk           I'm making a game about global warming.
Game Architect                        Track the progress at:
CogniToy                http://www.cognitoy.com/meltingpoint

_______________________________________________
sweng-gamedev mailing list
sweng-gamedev <at> lists.midnightryder.com
http://lists.midnightryder.com/listinfo.cgi/sweng-gamedev-midnightryder.com



--
fabs();
Just because the world is full of people that think just like you, doesn't mean the other ones can't be right.
_______________________________________________
sweng-gamedev mailing list
sweng-gamedev <at> lists.midnightryder.com
http://lists.midnightryder.com/listinfo.cgi/sweng-gamedev-midnightryder.com
erick jung | 2 Apr 2007 15:48
Picon

Re: tasking

Hi Richard,

The programmers has the responsability to schedule the technical tasks, and the answer is obvius... nobody knows how long time will be need from a specific task. I think that what you need is some kind of process (like scrum, etc). If you have a problem in scheduling, cuz in the final the time get from people will be more from the producer, than you need to talk with everyone to adjust something here or something there. Imagine in some cases (here in my job) that we need not to only has the time needed for a specific task, but we need to tell the number of loc (line of code) that we´ll need to accomplish this task. :/

I think you´ll need to create a new culture in the company, cuz schedulling isn´t an easy (or cool) thing to do, but it´s essential (and is a part of problem). Use any software that you want (project, excel, etc), cuz the diference isn´t in it.

[]´s
Erick

On 4/2/07, Richard Fabian <raspo1 <at> gmail.com> wrote:
I think this is probably the best place to submit this, as the audience is right, but something in me thinks that it's verging on "off-topic"

Our producer has asked me to help find out of there are any software solutions to a problem we are facing at the moment, and have faced on every project we've worked on.
The programmers are happy building lists of tasks, assigning staff based on capability and favourability, moving dates around after assessing critical paths, overallocating and looking after when things need to be done in time for other departments and requesting things to be done by a certain time from them too... but no-one else is happy doing all this "scheduling" and is saying that it is the producers job. Unfortunately this isn't possible as the producer is not "games industry", and doesn't innately understand some of the implications of certain elements of game development and we've seen some horrific gaffs because of an unnoticed dependency or a task that makes no sense (its like production through a "chinese room" engine).

so, the question really is: is there any software or peopleware process that will eliminate all the innadequacies of the system we have at the moment. We were hoping for some nice dependency driven todo list software or something, but we'd be equally happy with a peopleware method of fixing our problems.

--
fabs();
Just because the world is full of people that think just like you, doesn't mean the other ones can't be right.

_______________________________________________
sweng-gamedev mailing list
sweng-gamedev <at> lists.midnightryder.com
http://lists.midnightryder.com/listinfo.cgi/sweng-gamedev-midnightryder.com




--
erick jung
blog <at> jung.inneninteractive.com
mobile <at> +55-81-99497608
_______________________________________________
sweng-gamedev mailing list
sweng-gamedev <at> lists.midnightryder.com
http://lists.midnightryder.com/listinfo.cgi/sweng-gamedev-midnightryder.com
Phlip | 2 Apr 2007 16:52
Picon
Gravatar

Re: tasking

Richard Fabian wrote:

> The programmers are happy building lists of tasks, assigning staff based on
> capability and favourability, moving dates around after assessing critical
> paths, overallocating and looking after when things need to be done in time
> for other departments and requesting things to be done by a certain time
> from them too... but no-one else is happy doing all this "scheduling" and is
> saying that it is the producers job. Unfortunately this isn't possible as
> the producer is not "games industry", and doesn't innately understand some
> of the implications of certain elements of game development and we've seen
> some horrific gaffs because of an unnoticed dependency or a task that makes
> no sense (its like production through a "chinese room" engine).

You have described the exact prerequisites for Scrum.

--

-- 
  Phlip
  http://c2.com/cgi/wiki?ZeekLand  <-- NOT a blog!!
_______________________________________________
sweng-gamedev mailing list
sweng-gamedev <at> lists.midnightryder.com
http://lists.midnightryder.com/listinfo.cgi/sweng-gamedev-midnightryder.com

Jon Watte | 2 Apr 2007 18:50

Re: tasking


Richard Fabian wrote:
> so, the question really is: is there any software or peopleware 
> process that will eliminate all the innadequacies of the system we 
> have at the moment. We were hoping for some nice dependency driven 
> todo list software or something, but we'd be equally happy with a 
> peopleware method of fixing our problems.

There is no software that will fix a process problem. And there is no 
process that will fix a people problem!

It sounds like the main problem is that your producer doesn't understand 
what the contributors do. That's a real problem. Perhaps you can make a 
lead from each group help him out on the scheduling and dependency part. 
Maybe the art lead doesn't like doing that, but it's part of the job -- 
any creative work is 95% drudgery anyway. (Note that art team lead and 
art director are two different roles, but can be held by the same person 
in small groups)

The other question is: why are you building large schedules? It's my 
opinion that they never work. It sounds like you're having a 
communication problem within the company. I would try to solve that by 
holding short, daily status meetings, and grouping the work in frequent 
mini-milestones (say, every three weeks) where a cross-functional team 
delivers some specific feature for the game. (Yes, I think SCRUM is a 
good process, if you want to go "all out" agile) The daily status 
meetings, coupled with the short and reachable deadlines, will serve to 
keep the team honest about estimates and work needed.

Cheers,

          / h+

_______________________________________________
sweng-gamedev mailing list
sweng-gamedev <at> lists.midnightryder.com
http://lists.midnightryder.com/listinfo.cgi/sweng-gamedev-midnightryder.com

Sylvain Vignaud | 3 Apr 2007 02:56
Favicon

Re: tasking

Jon Watte wrote:
> The other question is: why are you building large schedules? It's my 
> opinion that they never work. It sounds like you're having a 
> communication problem within the company.
I would definitely have the big points over the months in my schedule, 
plus the detailed tasks for the next month. I would not like do have 
only a 3 weeks vision. Well maybe you implied this.
_______________________________________________
sweng-gamedev mailing list
sweng-gamedev <at> lists.midnightryder.com
http://lists.midnightryder.com/listinfo.cgi/sweng-gamedev-midnightryder.com


Gmane