Recent blog entries for janschrage

CaptainNemo:

PostNuke needs PostNuke users
[...]
Treat users like your greatest asset and they will turn around and become your greatest asset.

The gist may be the difference between a need and an asset, or possibly between users and contributors. Maybe I'm being picky here, but I'll stand by my words that a software project does not need users. Usually "a project needs users" is stated like a self-explaining mantra, which to me it is not. Let me elaborate: I am a user, like most of us here, to any number of OS projects. Not a single one of them needs me. And how should they ? To most of them I do not contribute, and they don't know that I'm out there using their software. If I emigrate to a lone beach after having written these lines and never again touch a keyboard for the rest of my natural life it will not make the slightest difference to them. And heaven forbid it should!

By contributing, as in "giving useful and usable feedback" I may turn into an asset for a project. Even a single accurate description of a bug will help. By contributing more I get more valuable for a project. I may become a developer on a project by contributing code, design, whatever. And I can still be replaced. If the project is interesting enough it will be carried through.If I don't report that bug, someone else will.

I'll go one step further: I may start a project, or make a unique contribution, something only I could have done. If I'm any good that contribution will still exist and be usable when I leave the project. The fact that individuals can be replaced is in my opinion one of the biggest assets of open source.

To return to my original point: Most users of software are not assets for the project or the team that produced it. They are just consumers, and thus are not needed at all. This is true for most of us, for most of the software we use. Some users will turn into assets - or let themselves be turned - and if you want to argue that then we need them, so be it. I rest my case. :-)

And as an afterthought...the yet-to-be-named ProjectX will be open to a community again. Of course.

A second afterthought on re-reading this...maybe I missed your point completely and the need for users is derived from some measure of success, or satisfaction in having done something that a lot of people make use of. This is a legitimate motivation, but it is not mine. If you draw motivation from a userbase in any way at all the argument put forth above becomes invalid for you and the need for users self-evident indeed.

movement, hit the nail on the spot about my attitude. And ain't bonuses nice, just once in a while ? :-)

I keep reading over at gmane on the PostNuke mailing lists how users are what really makes a project. Typically voiced by backseat coders who think they know better than those who do the actual work. And they are wrong. A project does not need users. It needs developers. Without users you can get a piece of software, without developers you can't. Now what on earth is so hard to understand about that ? I know for certain that I don't need users. I develop things because it's fun, and because I like working together with others on a project. And if I or the developers on that project remain the only users it ever gets that is quite enough. I still have fun and I have a good product afterwards. Now everybody is welcome to share my efforts in OS projects, and I actually value feedback. But still I guess I shall never know what makes some imbeciles think they have a) a right to my work and to that of my fellow developers and b) the right to tell me or others what to do and c) that clamouring loudly is a valuable contribution. All it achieves is to make them look bloody stupid and create a nuisance.

As reaper20 points out: "I swear, the next moron that wants mozilla to render ALT tags as popups, or ask for colored scrollbars should get drawn and quartered. We're screaming for standards and these guys think its their right to dictate what Mozilla should be doing."

That's very much what happened to PostNuke. The backseat coders and wannabes took over, the good devs walked out and now there's only fluff left. The code they have produced after the break is something I wouldn't touch with fire-proof gloves, they haven't got CVS working again (which I left in a working state and ready for taking over when I resigned two weeks ago) and the mailing lists are full of flame wars with the occasional group hug to go with it. Oh, and they lost several weeks worth of bugfixes upon an all-too-hasty release. (These, of course, are to be incorporated again "as soon as we've got CVS running, and we're working very hard on that", talk about a laugh.)

What's happening over there at the moment is quite a study in how people work and of what can happen to an OS project that once was good.

Resigned from PostNuke, with ProjectX now. Things look bright again. After all, almost all of the former PostNuke management and good devs are here. They're fun to be with, too. We're getting a lot of organizational and backend work done, and development's continuing at a decent pace. The XML parser is coming along rather well.

Our new servers are up and running. Very well, too, after the initial teething troubles that you always get. Now we all know that the internet is green, or maybe grey, and that servers do indeed cease to work when you accidentally pull the power chord. We have our first scapegoat of the week.

mrb and I are all set on making some fundamental changes to the way we handle development. He's written a good RFC on a new process and I'm redesigning the access control data model we have at the moment to fit that process better. It does now, of course. It just doesn't do it very well because I had something different in mind with the first design, so some enhancements are called for to reduce admin overhead and make the whole thing a bit more powerful. The new process addresses some issues common to projects of this size, like how to keep a stable trunk in the repository, how to have some quality checks within the development process itself etc. without making it a hassle for the devs. I'm rather looking forward to the day we can start on it.
Plus, we're evaluating subversion as a CVS replacement. We're both beginning to like the system; it works on different OSes and has some very good features for version control that have always been an issue with CVS. The first checkout is a pain on a large project, but after that it runs smoothly. I do hope the devs are going to fix the need for occasional db recoveries with the 1.0 release, because with the current need for manual maintenance it is not acceptable for us for daily use.

Of course, almost the first thing that happened after the first weekly tarball of ProjectX was released on dinerminor.com was that some dimwit started complaining about how it failed to install for him and the lack of documentation and things like that over at PostNuke. And it says in big letters right next to the download link "***NOT FOR PRODUCTION and barely for testing***" and "ONLY INTENDED FOR DEVELOPMENT USE". When a kind soul pointed this out to him he complained some more. There's just no helping some people. We'll have a user community to take care of again, eventually, but there is also that very special kind that I'll be happy never to have to deal with again. Ever.