On 05 Dec 2005 21:09:58 -0500, Michael Poole wrote: > > Mikhael Goikhman writes: > > > This would be insane. You suggest that whenever someone requests a tree > > of 1000 files, 10Mb, tla should create million of files occupying 10Gb. > > [This is optimistic, since just nodes alone of typical 4Kb occupy 4Gb.] > > Wow! How many 1000-file, 10 MB projects have millions of file > revisions in a single category--branch--version _with no cacherevs_? > It sounds like a great example that could be used for evangelism and > demonstrating GNU Arch's scalability! > > (I cannot think of any. If you did not mean the numbers literally, > perhaps you should not have picked such an outrageously inflated > example. It makes it hard to take your argument seriously, especially > when you claim that's a conservative estimate.)
You are lucky if you didn't see a version of 1000 revisions without a single cacherev. I did it not once when monitored different archzoom setups on the net (tla performance is horrible in this case indeed). In any case, it is irrelevant whether your revlib grows to 1000 revisions after one "get", or after several more tla operations. What is important is that it will surelly grow one day. If you actually use tla, of course. Regards, Mikhael. _______________________________________________ Gnu-arch-users mailing list Gnu-arch-users@gnu.org http://lists.gnu.org/mailman/listinfo/gnu-arch-users GNU arch home page: http://savannah.gnu.org/projects/gnu-arch/
