I agree with Ubernostrum and kchrist. A project like this needs a clear vision, someone to keep things on track and prevent aC from becoming a Rube Goldberg device. One of the attractions of aC is that it's clean and lean. You need a strong, opinionated developer to control things and keep it that way.

I think Ilija is doing a great job. He's got a talent for fast, light designs that kicks butt. He's totally involved here on the forums. He's laid out his vision for aC. Why all the sour grapes?

Keep up the good work, Ilija. Good luck with the project launch. Sounds like you might soon have time to put some work into aC again.

Hey Ilja,

have you read "The Cathedral and the Bazaar"? here is a link to a pdf of the article http://gnuwin.epfl.ch/articles/en/cathedralbazaar/cathedral-bazaar.pdf

You can also read up on it at wikipedia and some other places through google. Its a bit of an old article, but its pretty easy to read and I think it really explains alot of the points that are trying to be expressed here. I hope you read it and would appreciate hearing your comments on it.

I know about the CatB though I never read it completely. Here is what I think about CatB (nothing more):

Its just one of many talks about software development. It is pretty influential in open source community, but there are actually programmers all around the world that haven't heard of it and they still write code (meaning - its not essential if you want to be a programmer).

Everything depends on what the problem is and how you look at it. For instance, take a look at BSD project. Cool stuff, solid, UNIX platform developed by many passionate programmers from all over the globe. Even this site runs on a FreeBSD powered server and its great. Bazaar approach rocks!

But then, how many people do you know that use any flavor of BSD for everyday tasks such as sending emails to their mom, managing holiday photos or producing a new logo for a company in some vector tool? None, one, two? More? Hm, I know two persons ;)

Now, take a look at something that one company with dedicated developers and a lot of invested resources can do from something like BSD. Yeah, development process is closed, decisions are made by a small group of people, it comes with a price tag and so on, but millions of desktop users simply DON'T CARE about that. From their perspective cathedral approach rocks (though they most probably never heard of it and do not know what it means).

Whole point of this post is that everything depends of your perspective. There is no silver bullet, there is no ultimate solution to bugs in software and there is no "the one and the only" way how you can develop software. You have a problem that you want to solve with your software, you have defined your target audience and now the question is how to approach the development with resources at hand. No 1997 essay will give you an answer to that question though it might help.

Hey Ilja,

Just a few comments.

1) I just wanted to point out that there is also OpenDarwin - which is the community side of OSX
2) I also understand your point of having a commercial company producing 'end user" products instead of a community of programmers scratching their own itches. I think the fact that there is money backing those "end user" products is a big difference here (unless you plan on starting to sell aC at some later point which i think i have read you don't). Also, I think its good to realize that even though it is a closed process it is still supported by a team of people, not just one person, and in reality its a group of small teams.
3) I also can completely take to heart your idea that the approach should match the goals of the project, and maybe that's where I'm having frustrations with this project.
4) Trying to focus just on that article, I pointed it out because as you mentioned not everyone has read it. I actually hadn't read the whole thing until I posted about it. I was surprised how easy to read and short it was. Also, I wasn't trying to suggest that this article would "solve" any of those issues but I think it provides some good insight into both open source projects, and projects as a whole. Another really good resource that CaTB references is "The Mythical Man-Month" by Fred Brooks. (You have probably heard of it too)

Related to your last point about matching approaches with goals, I would think that since aC is aimed at companies (or company-type groups) I would think a more collaborative approach would be useful. Every group has their own quirk that they will want to customize for, but there are also a lot of things that many people would like to use. Allowing the community to submit patches (and having more than one person that is able to make decisions about them) seems like a desirable approach compared to an "end user" product where there is not much customizing expected. Consider the difference between an iPod and FreeBSD (or OSX) in terms of customization.

@vicroy - didn't change my mind, just got busy. (perhaps another example of needing more people)

Hey Ilja,

something else that might help explain my points a bit are the various examples that I keep reading about in the forums. Below are just some examples of things that people are anxious to see in aC, and have resorted to coding themselves. If their changes were able to be submitted, reviewed, and contributed back to the code base, then everyone could benefit from one persons changes and we wouldn't be waiting for you to recode the same feature.

-Commenting on tasks ( http://www.activecollab.com/forums/post/6854/#post6854 )
-Writeboards/Wiki ( http://www.activecollab.com/forums/post/6701/#post6701 )
-Branding/Logo ( http://www.activecollab.com/forums/topic/185/ )
-Filter Tasks By Person ( http://www.activecollab.com/forums/topic/1303/ )
-Reporting/Full Activity list ( http://www.activecollab.com/forums/topic/1425/ )
-DashBoard Milestones/Calendaring ( http://www.activecollab.com/forums/topic/406/ )

yes, i'm aware that most of these are "hacks" but i'm sure better code would be produced if there was that option