Date: Fri, 11 Jul 1997 23:51:36 -0400 From: "Louis A. Mamakos" <louie@TransSys.COM> To: asami@cs.berkeley.edu (Satoshi Asami) Cc: jkh@time.cdrom.com, current@FreeBSD.ORG, hans@brandinnovators.com Subject: Re: Heads up and and a call for a show of hands. Message-ID: <199707120351.XAA18977@whizzo.TransSys.COM> In-Reply-To: Your message of "Fri, 11 Jul 1997 13:53:58 PDT." <199707112053.NAA27593@vader.cs.berkeley.edu> References: <199707112053.NAA27593@vader.cs.berkeley.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
> And Xi Graphics seems to have ignored my comment about just using the
> existing framework and adding /usr/{local,X11R6}/etc/rc.d/foo.sh to
> run ldconfig.sh and any other startup stuff. The -m flag to ldconfig
> was added for a reason.
I agree; this mechanism seems to work pretty nicely for both ports/packages
as well as local software.
> See what the modula-3-lib port does. I don't think an /etc/ld.so.conf
> is necessary (much less /usr/local/etc/ld.so.conf). I propose backing
> it out.
I've used it locally to handle [incr tcl] which is installed in a different
location and it sure beats having to edit a file, etc. And using the
/usr/local/etc/rc.d script, you can invoke arbitrary policy (see if a
directory that might be on removable media exists, etc) before adding
directories to the list.
It seems like to opportunity to render you machine inert with a badly
"automatically" edited /etc/ld.so.conf file compared with the existing
mechanisms argue against this new approach.
louie
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199707120351.XAA18977>
