(Illustration by Gaich Muramatsu)
Elliot Lee <sopwith_at_cuc.edu> ,in message <Pine.LNX.3.95.980605005037.15480A-100 000_at_helix.cs.cuc.edu>, wrote: > Peter said at his LinuxExpo talk that it trickles back according to the > available bandwidth... (in the voice of Mr. Burns:) Excellent. > > It's a grievous resource hog. > > Mostly an implementation issue - use of C++ doesn't help matters I > suppose, and it wouldn't hurt to run coda through purify once or twice... Well, let me say that I wouldn't use C for anything over 1000 lines anymore. It's much easier to write correct dynamic memory usage in C++ than in C. I'm also not familiar enough with the coda internals to tell you why they want >80M of VM, but I'm not arrogant enough to think that I could significantly reduce that without years of code study. I would be very surprised if it isn't a traditional time/space tradeoff. > The builtin security is waaay better than what NFS has - it's obviously > not perfect, but Well, if they're not subject to ITAR their security is still not good enough to use unassisted across the big bad internet. Volmes spread across the internet is where coda holds the biggest promise. I would really like to have a server at my house and coda export the source directories to the CEO over in Californa, 18 hops away. I'll have to do it through an IPSEC tunnel, though, because that information is way too valuable to have it snooped off the wire. I expect such a modular solution will be the best way to go for years to come. Why reinvent the wheel unless you absolutely need to attach a drive shaft? -- Bob Forsman thoth_at_gainesville.fl.us http://www.gainesville.fl.us/~thoth/Received on 1998-06-05 10:39:49