On Sat, 2 Feb 2002, Jimmy Wales wrote:
Hr. Daniel Mikkelsen wrote:
What are the expenses involved? What is the bandwidth consumed? The storage space required? The load on servers?
- The physical expenses are minimal and easily absorbed. So none of
the things you mention are worth worrying about -- I can continue to pay for those things indefinitely. (If Wikipedia were to be 100 times more popular, though, these things would start to add up!)
Yes. Given the wonderfully light format (and nature!) of the project, I'd expect a normal small company could run it today.
I feel certain the project will become very large. When that happens, I'm thinking it shouldn't be too difficult getting a university (or at least some non-commercial entity) to host it, since it is makes very good use of bandwidth (compared to, say, radio streaming) and is Free. Also, although there would some consistency problems, making Wikipedia distributed (something like freenet/newsgroups) should be possible.
As for hardware, it's really cheap these days. I haven't scrutinized the PHP Script, but it occurs to me some creative cacheing (and at the other end, offloading history data to a less optimized method of access) could extend hardware life substantially.
- The programming expenses are drastically lower for Wikipedia than
they ever were for Nupedia. Volunteers are to thank for this. Nupedia was very expensive to build, because I paid programmers (chiefly Toan) to do it. Currently, Jason and I spent a few hours per week on Wikipedia.
Yes, Free Software has basically solved this problem. If the rest of the world's aspects could just learn from this model. :)
- The major remaining expense is Larry's editorial leadership, which
I believe has been a major factor in the success of the project to date. Fans of random anarchy may disagree, of course. And if they do, I'm happy to provide them (so long as the license is free!) with web space to have a totally unmanaged project. I think they'll get crap, though.
I agree that one should have a set of dedicated and involved editors, but do these really have to be hired? The project is way past critical mass, and I see a lot of very active contributors. A rotating voluntary post, or a comittee perhaps? Once enough people participate in something (read: once someone spends the effort and money to get something rolling, like you did), people find all kinds reasons to chip in. Status, the feeling of contributing to something worth wile, desire to influence - didn't ESR make a list of these?
Anyway, my 10c is that both volunteer editorship, and if that fails, random anarchy should be _tried_ before we (I hesitate to say "we" here, as I'm new to the project) try advertisments.
That said, I think you could make back a lot of your investment in effort and money by identifying Wikipedia more closely with Bomis and Nupedia. A lot of people visit these pages.
-- Daniel Mikkelsen, Copyleft Software AS