Mark Crispin <[EMAIL PROTECTED]> writes: > On Fri, 10 Oct 2003, Chris Wilson wrote: >> squirrelmail says it'll improve my performance, kinda makes sense. >> http://www.squirrelmail.org/wiki/en_US/SquirrelMailPerformance > > That document is wrong. This is a case where a little knowledge is a > dangerous thing, since it transforms into myth.
Are there inetd implementation that doesn't result in a fork() + exec()? A standalone implementation might remove the need for a fork() + exec(), which would improve performance. The performance increase is probably not relevant, though, but may be on a loaded server if a poorly written web mail client is involved. Regards, Simon
