Date: Wed, 22 Feb 1995 16:12:35 -0700 From: Nate Williams <nate@trout.sri.MT.net> To: Garrett Wollman <wollman@halloran-eldar.lcs.mit.edu>, Poul-Henning Kamp <phk@ref.tfs.com> Cc: current@freefall.cdrom.com Subject: Re: TRUE and FALSE Message-ID: <199502222312.QAA15832@trout.sri.MT.net> In-Reply-To: Garrett Wollman <wollman@halloran-eldar.lcs.mit.edu> "Re: TRUE and FALSE" (Feb 22, 5:27pm)
next in thread | previous in thread | raw e-mail | index | archive | help
> > If you develop kernel-dependent sw, you had better make sure that your > > development env is aligned with your kernel, ie something like: > > > cd /usr/src/include ; make all install > > cd /usr/src/sys ; make all install > > > or whatever the trick will be. > > And I'm saying that it's an incredible imposition to force such > developers to do this every time they make a change to a kernel header > file. I won't stand for it. I agree with Garrett here. It's silly to do this, and this will cause no end of problems when people forget to do this after they upgrade their kernel sources. Forcing a 'make world' every update is much too anal. People keep current with the kernel, and very often the kernel-specific utils needs updates. It's simple to do that now. Build and install new kernel re-build and re-install libkvm re-build ps By forcing the user to re-install the include files every time, he *may* be forcing things to break that are non-kernel specific. The LOCALE changes come to mind. I completely ignored them when they were being done but I was able to continue to do kernel testing. This would no longer be the case with your proposal. Nate
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199502222312.QAA15832>