[Pkg-jed-devel] error building jed_0.99.16.pre.0.99.17.84-1

[Jörg Sommer]

--82I3+IH0IqGh5yIs
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

G. Milde schrieb am Tue 19. Apr, 11:40 (+0200) :
> On 18.04.05, J=F6rg Sommer wrote:
> > G. Milde schrieb am Mon 18. Apr, 09:50 (+0200) :
>=20
> > But jed will never buildable if a version lower than this is installed.
> > Think of the jump from sarge to etch. From this point of view our packa=
ge
> > will never buildable and I don't know if this is a policy violation.
>=20
> I suppose this is tolerable, as long as the packages are in experimental =
(if
> we do properly document it like: "uninstall jed-extra<=3Dnn before updati=
ng").

Are you shure, this is policy conform?

> I wonder, whether a pre-install script could do some manipulation or test=
ing,=20
> though.

For the pre-install script it is to late. The problem is while building.

> > > * if John doesnot like a new command line arg just for Debian, we cou=
ld
> > >   still "emulate" it in jed.conf by looking at __argv and killing the
> > >   spurious "argument buffer". As this argument is rather seldom used,=
 the
> > >   overhead of the spurious buffer would not seriusly harm.
> >=20
> > This is ugly.
>=20
> Agreed. But I wonder whether it would be less ugly to have a Debian
> specific patch in site.sl?

What is ugly on a patch? Currently, there are patches in the packages,
they change files in the jed source.

> > No. I see the bugreport coming: "If I have a local jed.conf in
> > /usr/local/etc/jed.conf (which is FHS conform) I can not build jed."
>=20
> This very much depends on what is in this file. If I remember

Right.

> right, the problem arose from code in /etc/jed_init.d evaluated by
> jed.conf. I doubt, a user or admin will replicate the jed_init.d
> screening in a local jed.conf.

This not, but he may place a .slc file in /usr/share/jed/... and
evalfile() it in /u/l/e/jed.conf

> Finally, we could also try an ERROR_BLOCK to keep jed running, but I am
> not sure whether it will help in this particular case.

Good idea. Because an admin may place a .slc file in the path that is not
byte compatible with a new jed version. This make jed crashes everytime
and a user has no chance to bypass this. (With --no-jed-conf he can. :)
If a an (syntax) error occure, in /etc/jed-init.d/, then it should go on.
Or what do you think?

J=F6rg.

--=20
Erfahrung hei=DFt gar nichts. Man kann seine Sache auch 35 Jahre schlecht
machen.
	   	     	     	      	    (Kurt Tucholsky)

--82I3+IH0IqGh5yIs
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: Digital signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.0 (GNU/Linux)

iQEVAwUBQmY/tIZ13Cz2nwVYAQIdKggAvaYGwDnXxGZopmbigdvFrdLCToJhJl8c
IibowlhmdRj+OhtG16PPZC4VZXE2wmlyoJ/gV33GFZDwSuaUYS71Bvi0LXDDepZT
0/w2GZO39LIk3V8CoyuYiN3L5ECaFzsXK6akgkS2dO7tv4mtHi2pY9FjRGJ3a61P
OM2ok15SI7xlaYJZSGmBCksKzU6JNU9PzTo0zOPF5KFSRd9kIvNQLtNE2M+Eru0u
ZzWwxZfN25RysVRurHC8n0Z2FlQRPADSvUrIsbKpGdMbE4ixqt+eHiVaq1mjoXGJ
yRt61FQHhT1Iq3P6cAYWmiuPwmo2qOGqW+T5fup2ZMrbMf5kAx0v+w==
=x+1U
-----END PGP SIGNATURE-----

--82I3+IH0IqGh5yIs--