Date: Thu, 04 Jan 2001 16:20:48 -0700 From: Warner Losh <imp@bsdimp.com> To: Will Andrews <will@physics.purdue.edu> Cc: Garrett Wollman <wollman@khavrinen.lcs.mit.edu>, current@FreeBSD.ORG Subject: Re: cvs commit: src/usr.bin/apply apply.c Message-ID: <200101042320.f04NKm147924@harmony.village.org> In-Reply-To: Your message of "Thu, 04 Jan 2001 17:20:05 EST." <20010104172004.P86630@argon.firepipe.net> References: <20010104172004.P86630@argon.firepipe.net> <200101041905.f04J5ou82617@freefall.freebsd.org> <200101041909.OAA61522@khavrinen.lcs.mit.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
In message <20010104172004.P86630@argon.firepipe.net> Will Andrews writes: : What, exactly, are we trading off by making apply(1) a bit more : paranoid? A couple extra cpu cycles? Maybe you haven't noticed, but : these days there's almost nobody still using 100MHz chips. And out of : the ones that do, how many will use apply(1) more than once or twice in : the lifetime of the machine? First off, we do still have a significant part of the user base that is using older, slower machines. Many of them run embedded systems, some of which use apply(1). I have a big problem with this argument since it is doesn't understand the wide range of environments that FreeBSD runs in and runs very well in. Having said that, these patches are likely not to significantly slow things down. Apply tends to do a lot of I/O, which swamps the few extra instructions that the strl* or snprintf routines introduce. There's little reason to optimize things. I'm still not sure about the shell environment actually buying anything, but I could see how it might help. Warner To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200101042320.f04NKm147924>