I posted this in a discussion on TheChaosEngine forums, but it sums up my position on overtime/fixed-working-hours quite succinctly, so I thought I’d re-post it out here in the real world. For reference, our team tries to work office hours of 9-5, rather than a flexi-time arrangement. This is, it seems, quite unusual in the games world, and there was a spirited discussion on TCE about whether or not it stifled creativity and/or leads to making the unpaid overtime situation worse.

[in reply to the implication that really good employees are the ones who work extra and late to deliver over and above expectations]

While I heartily acknowledge that it’s daft to just up sticks and leave at some arbitrary time and break your flow, I think that its still going to be the exception rather than the rule. You don’t always hit your peak productivity at 4:30 only to have to leave at 5. Certainly there are some days when I’m just getting going when I really should be going home – I find a good point to shelve it and go home. And when I come back in the morning, I’ve had a whole night to ponder the work and generally do a better job of finishing it up than had I just forged ahead.

That said, I don’t want my team feeling like more hours = better work. The only way I want to see productivity improved is for people to work smarter and harder for those 8 hours. I want people to go home feeling like they achieved lots, but still get home at a sensible time that leaves them free to enjoy their own time. I want people to come in and focus, so we’re all in that intense zone of getting s&*^ done at roughly the same time.

We all spend time during the day surfing the web, emailing others, etc., and it eats productivity, and doesn’t necessarily improve the creative atmosphere. I don’t want people staying late to get their stuff done because they have only done 5 hours of effective work in the 9-5 period. Then they end up getting home late, and blame the job for sucking up all their free time.

I want a team that burns bright during their work day, and finds that balance between producing volume, and feeling the creative spark that gets them producing quality. If the only way I can improve the team’s productivity or quality is to increase the hours worked, then I’ve failed. If we’re really at our peak productivity in that 8 hour day, then I’m due to hire someone else. I’m pretty damn sure that if we every got even close to that zone we’d be one of the most effective devs around.

Bah humbug. The new office on Rutland Square was looking better and better, and we were days away from signing the lease. Sadly though, delays in organising things amongst the three companies to be involved meant that someone else has seen the place and signed the papers all in the space of a week, and now we’re back to square one in our search for an office. Boo hiss. On the bright side though, the new office would have been a tad smaller than the current one, so at least now we’ll have the potential of finding somewhere with a bit more room to grow.

In other news, our prototype games on WiiWare are now looking like actual games, although I’ve left it in the capable hands of Pete and Tim while Charlotte and I have been taking care of our clients. So as not to be left out of all the fun though, I’ve taken on a bit of hobby coding and am converting our somewhat ungraceful build process (a combination of Lua scripts, makefiles for Wii/PSP and Visual Studio projects for Win32), and am converting it over to use JamPlus. There has been a lot of debate on the sweng-gamedev mailing list and elsewhere about getting build systems which cope with the rigours of game development. Jam had the most potential but development of the main version of it has pretty much died off, and the patches and work required to get a Jam build suitable (performance and flexibility-wise) for games development is enough to put everyone else off.

The helpful Joshua Jensen however has done some sterling work in putting together all that existing work in a practically off-the-shelf tool called JamPlus. I’ve been most impressed with its speed and flexibility so far, and thanks to Josh I’ve learned enough of the scripting logic to convert our quite specific build needs into Jam logic. Once it’s done I’ll publish our scripts to serve as an extra example for folks (minus all the Wii and PSP stuff that’s covered under NDA of course), hopefully that will help others and get it to be a well accepted tool. As always, I think anything that helps efficiency of games developers in general is good, but really it’s because as our prototype games have grown, it’s become clear that we need a proper dependency-checked build tool to manage our asset->game toolchain.