9
0
I recall reading a comparison of NetBSD and FreeBSD a few years back, and in 2000 (okay many) it looked like NetBSD forked processes better, while FreeBSD threaded better, and so, for Apache 1 NetBSD seemed a little better.
Since then It seems that FreeBSD supported SMP first but then NetBSD claimed to support it better, until FreeBSD again improved. ETC.
Also I personally found that though people say FreeBSD supports the standard PC hardware better, NetBSD had wider device support at one time and more options for various clone type PCI cards, like BT848s or Tulip Ethernet Clones etc.
Has anyone seen updated comparisons?
I'm not talking about a user walk through of what X desktop system is default etc. I'm talking about talking points with graphs and charts about what happens to latency or memory use or speed when you fork a 1mb process N times. You know when N is < 15 one might look like the clear winner until you see a different curve after 15 and the other is the winner.
I'm trying to get a feel for the current stance between the BSDs. I get the odd feeling that NetBSD sort of shifted focus once they dropped the old logo of conquering a pile of old machines.
@GrahamLee I think reading a comparison is a first step over testing your application. And no, if you're interested in a particular application I wouldn't be first writing an test and article on a generic comparison. You can use existing generic comparisons to extrapolate about your application though. IE forking latency and Apache 1. – dlamblin – 2012-05-21T18:46:18.977
It'd be great to understand why the forking N 1MB processes microbenchmark is important; if you're wondering how NetBSD/FreeBSD work with your application then just test them with your application :-) – None – 2009-07-23T10:07:12.500
1This might be better off on serverfault.com. – ConcernedOfTunbridgeWells – 2009-07-23T11:20:44.600