--6pbY/KU4ayLo+qis
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Mon, Mar 21, 2005 at 10:07:38PM +0000, sigsegv@rambler.ru wrote:
  | Martin Husemann wrote:
  | >On Mon, Mar 21, 2005 at 09:25:34PM +0000, Matthias Scheler wrote:
  | >
  | >>Is there a reason why you don't use the safe road by building into a
  | >>destination directory and using "./build.sh ... install=3D/" afterwar=
ds?
  | >
  | >
  | >Because installworld is too clever and checks again.
  | >I tried to convince Luke to fix this a year or so ago, but he insisted
  | >it would only cause support trouble from people that accidently overwr=
ite
  | >their binaries with incompatible ones for another arch.
  | >
  | >FWIW: pkgsrc has a tool (libkver? I forgot the name) that allows faking
  | >arbitrary uname output.
  | >
  | >Martin
  | >
  |=20
  | Well I thought -E flag for build.sh was there to override such=20
  | assumptions, unfortunately for some reason it doesn't work in this case.

So instead build.sh sets (takes a trivial amount of time) and extract
them, instead of using installworld.

IMHO, -E must die.  It has caused more support headaches for me than
the `benefit' it offers the people clueful enough to use it correctly,
especially given the numerous other methods in which you can achieve
what -E gives you.

--6pbY/KU4ayLo+qis
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.6 (NetBSD)

iD8DBQFCP2pxpBhtmn8zJHIRAvVIAKC6CiGtdEujvymvCeMKsHVCXd9ZLgCgq7lf
sggDuacxdENZ7Z6WSVW8Tak=
=9JRa
-----END PGP SIGNATURE-----

--6pbY/KU4ayLo+qis--