On Fri, 30 Nov 2001, Jacob Meuser wrote: > > 1) Put the site under CVS, maybe it's already like that, but I didn't > see any tag in index.phtml. I recently did this with the site I'm > working on (mostly php, some static), and I'm very happy with it. CVS > might be clanky and old, but it's still a nice tool. >
I have done this, on several trees mostly in the neighbourhood of ~250k worth of text, it's pretty easy to set up and run, it's somewhat harder to persuade people that they need to go through this process rather than just drop their files into any convenient directory... ;-b > 2) An anonymous ftp upload area. Quiet down, if it's chrooted, noexec, > and users can't see anything's there, why would someone abuse it, and > what harm could they do? This would be for people to upload patches, > content, whatever for the site. Alternatively, a [www|wwwdev]-changes@ ml > or gnats or something similar could be used to suggest changes, but large > additions and graphics or other binaries might be more efficient with one > direct transfer. > Uhm, It would certainly be convenient, but only if you were sure that someone was actively maintaining it and straining out the cruft. > 3) An https/password protected (passwords for whoever the group decides can > have one or whatever) area that would have access to the ftp booty and a p* > script that was a frontend to a limited set of CVS commmands. > check out http://viewcvs.sourceforge.net The current version works fine without the database as long as you're using it read only. To take advantage of the advanced features it needs the db, currently it depends on Mysql, at one point I spent a fair amount of time reading the code with an eye toward converting it to postgres, but I was distracted by other more pressing needs. It wouldn't take that much mostly grep work but a lot of debugging and making sure that all of the queries had been correctly translated from one sql dialect to another. It's an hour or two for superdevguy, but it would probably take me a couple of full-time days (one to understand the code, fifteen minutes of typing and then a day of debugging :-\ ) > 4) read-only cvs access for the world, so people can make well-formed > patches :). > viewcvs can do that, although it's generally not considered a wise idea to put the sourcecode for activecontent live on the web... > 5) Of course, documentation for all of it =] > The documentation exists in a dank, dusty archive of an email list that is no longer active on a machine in norway that is connected to the net with a cable of rusty barbed wire. Bwahahahahaha!!?!?!! > Unless Larry already has something worked out ;) > More like, a vague intuition of what a good place to start with would be. > %%message to self: quit it with the emoticons%% > Awwww, but they're fun 8-}
