Archive for April 21, 2006

I was reading iamsure’s blog a short time ago and was disappointed to learn that the BNT Proton branch has suffered something of a setback. Not because of lack of development, not because iamsure’s PC crashed and he’s awaiting a new motherboard, not because there were major issues in the code – but because the code was firstly not backed up, and secondly not publicly available. I hate when that happens to me…and it has in the past.

It brought to mind the open source maxim “Release early, release often”. As I’ve hinted at in the past (and just pointing out a personal view), BNT is a project shrouded in a great deal of secrecy when you’re a non-developer looking in. It’s open source, yet not strictly open community. We all know there’s a BNT mailing list somewhere – but its apparently private and its existence rarely noted. I think I’ve only ever come across a handful of references to its existence though the forum may contain far more I’m unaware of. We all know there is (or was depending) some interesting Proton code in BNT Land which no one from the public could access. Its a shame really, BNT hangs out there like few other PHP games on the scene in terms of popularity and its user base.

It’d be interesting to see what would happen if someone magically turned the mailing list into a public free for all with archives all the way back to when it was established, and all code regardless of its status was always committed and publicly available from subversion. Suddenly every Tom, Dick and Paddy would have the ability to read all new code within hours of its writing and have access to all those private closed-door discussions (barring private email).

How would that impact BNT as a project? In QS our published policy (a mere formalisation of what pre-existed in our group tradition) is to delegate only disputes or sensitive (read: copyright) material to private forums or email. Everything should be open, and it should be static (no after the fact editing). Both of those have come under fire in the past – some people dislike publicly discussing their code, others dislike seeing public discussion of their projects unrelated to QS. In both cases I made it clear they were not debatable practices.

For QS the policy definitely shows advantages. Open communications, from bug reporting all the way up to inter-developer debates, when accessible to the public draws feedback – and more importantly suggestions and opinions. Open code, while of limited utility in QS Evolved where everyone is dealing with something 2 developers cooked up as individuals before we contributed it as OS, promotes peer review, alternate solutions, debate over approaches, and usually a great deal of valuable brainstorming.

Of course all individuals and groups have their preferences, and we all have reasons for them. BNT is hardly the only or even the rarity when it comes to keeping its communications and code to a more private level. Projects similar in organisation can be as fantastically successful as any other project. I just believe that in the PHP Game genre its more beneficial to give outsiders a window into the project’s administration and the ability to post their point of view, and review new code if they feel like it.

So long as BNT and similarly organised projects don’t take the path of Legend of the Green Dragon though everyone can safely ignore me…