On Sunday 18 January 2004 12:19, Joern Nettingsmeier wrote: > Larry Troxler wrote: > >>This is a questions of good indexing. Google is your friend. > > > > Speaking of mailing list archives, what happened to the good old days > > where you could just FTP a massive tarball of the last years messages and > > search through it on your own machine, using you own methods? > > > > It seems that for most mailing lists now, the only way to get at the > > archives is through a browser or search engine that is run on the server. > > Why is this? The information still has to be stored on the server, > > regardless. > > the probable reason is that ftp downloads of full archives put quite a > burden on the server, especially if someone has to pay per byte. > consider how many thousands of online queries correspond to the > bandwidth usage of just one full download. > Oh, duh! I'm sorry for missing the obvious reason - I guess you can tell that I've never hosted web space before, huh?? I guess the answer is for the mailing lists to cell CD's of archives. I'd gladly pony up for such a thing. > the webspace for the linux-audio-* archives is being donated by kai > vehmanen, and my guess is his bandwidth too is limited. > > that said, if someone does want to conduct some very special research > (such as indexing or statistical analyses of some kind that require > full-text archive access), something can be worked out with kai - just > drop him a note. > Nah, it was just a general note / question, not actually specific to the Linux audio list. > as a sidenote, the LAU traffic totals at around 55 megs uncompressed so > far (since its beginning in 2001), LAD traffic since 2002 is around 45 > megs, so the total is probably way over 100m. > > regards, > > j?rn That's all?? I actually thought it would be much more. Damn, then someone _does_ need to sell a CD filled with various mailing list archives, then. BTW with broadband accesss, I would gladly dl 100 megs, to escape the hell of the online search engine. My error, of course, was in ignoring the sever's bandwidth cost in providing such a tarball. Larry Troxler
