On Thu, Jul 03, 2003 at 06:37:25PM -0400, Todd Vierling wrote: > : Unfortunately, NIO is a critically needed feature. We will be better off > : to switch to NIO and lose those who cannot use NIO than to stay with the > : broken and overloaded freenet. > > Sounds like broken code. While NIO can improve performance, it's certainly > not "critical" from the perspective of my long-running nodes. My nodes' > load rarely went above 50%. > > I work on a large Java codebase at my office containing a couple dozen > programs that have very important duty (in fact, one of them is a BGP4 peer > that is currently talking to some of the Internet backbone routing peer > points). These apps do rather intensive network I/O without NIO, and yet > I've experienced zero "broken" I/O problems in this codebase.
They have dedicated servers, and they do very little in the way of expensive operations like crypto (which suffer substantially from having to timeslice with a bazillion I/O threads), and they are not remotely cross platform. Right? > > : The issue has just been broached and you are already to take your ball and > : go home? > > After getting *a complete brush-off* from Ian himself when I did indeed > offer to help (in a quite extensive context), why the hell not? It's not as > if he gives a damn either. -- Matthew J Toseland - [EMAIL PROTECTED] Freenet Project Official Codemonkey - http://freenetproject.org/ ICTHUS - Nothing is impossible. Our Boss says so.
pgp00000.pgp
Description: PGP signature
