ahn at orion.netrek.org (Dave Ahn) writes:
> Anyway, you are right, the old site could be brought online quickly, and
> a new wiki would take a much longer time to fully populate.

Not just populate. If it were that simple, I would have thrown stuff
into the dev wiki. But I do not want to increase the inertia of a 
system that I don't think looks good enough to attract newbs. 

We should have a design that looks as good as the genocide content.
And people who know how to maintain the design, and have access to do so.

(Do we even have anyone who knows how to make a wiki look good?
And how long will it take them to do so? Remember, unless we have a 
whole seperate test wiki, we probably don't want the sucker live until
the design phase is done. And the designers may well need some admin
capabilities; server config and restart.)

We should know we have wiki software that will be stable and
well-supported for a long time. Five years, minimum; perferably
longer. To me, that says WikiMedia. Which had 15 updates last year,
uses MySQL as a base (400 open server bugs, which isn't bad but does
represent possible admin needs), and I would expect needs database
dumps and not just simple backups.

Do we have the sysadmin resources for that when a simple restore took months?

We'll also need admins to do lockouts and reverts of vandalism. 

This is a lot more sysadmin work than what we do now, even though it does
distribute the content authoring and maint. 

On the other hand, put the genocide site into darcs and we have the
design already done, no wiki-design-aware people who have to climb the
learning curve and then stick around for when we need them again,
simple backups, far fewer vandalism worries (an entire security layer
less to be concerned about), no new admin procedures, designers
probably won't need shell access or test environment, very few
software updates to worry about. We're up in weeks, not months.

And if we then want to do a wiki as a primary interface, we can take
our time, without having to be concerned about having something in
place to attract and support newbies, especially during this rather
significant mac client rollout period, which is already underway.

Doing sustainable design of internet systems is not a matter of
throwing a wiki on the wall and seeing if it sticks unless we
really truly can't get it done any other way.