[pkg-GNUstep-maintainers] GNUstep fucking

[Gürkan Sengün]

--Signature=_Sun__27_Jun_2004_12_35_58_+0200_razwl72k7UwB2gY6
Content-Type: text/plain; charset=ISO-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable


> >>- openapp is a binary that people execute directly, so it must be in
> >> /usr/bin/; it is a serious bug if they are not.
> > i think we should use "open" from the Backbone project, that is the
> > successor of openapp. and put it in /usr/bin
> Too many tools are already named open.  If thats the case rename it to=20
> open.gnustep even.
many would at least be two. can you name the many tools please?
no i don't think open.gnustep is something that i or other gnustep users
want to type many times.

> > evan, i don't mind, either you or me (or together with co-maintainging)=
 we
> > use backbone cvs snapshots and generate four packages:
> > backbone-open, backbone-terminal, backbone-textedit, backbone-preferenc=
es
> > if you want we can also call them .app (but not the open tool, what will
> > them be called? (package name)). i would still prefer call it what they=
 are
> > called upstream, i.e terminal. or rather termanal textedid etc...=20
> > real gnustep people seem to use / as their root and ahve /System, /Volu=
mes
> > ... as defined in the standard doc of linuxstep....=20
> This is debian, not linuxstep.
would you mind reading it anyways? it's here:
http://www.linuxstep.org/documentation/LSFH-1.3/LSFH-1.3.html


> >>- architecture-independent files must be in /usr/share; it is a serious
> >> bug if they are not.
> > splitting and cluttering a gnustep.app/ package across the file system =
is
> > seriously wrong and against gnustep (openstep,nextstep) philosophy and =
idea.
> > there must be a way to follow the specifications and standards for GNUs=
tep.
> > users will be fucked when you give them something that is not GNUstep a=
nd
> > you call it GNUstep. if you really want to distribute the files across =
the
> > filesystem do so BUT DO NOT CALL THAT GNUSTEP!
> You are integrating GNUstep into an existing operating system.=20
> Following Operating System standards is optimal.
not if they break/change basic things that belong to the idea of it, that u=
sers of something are used to.

> > i recommend people source GNUstep.sh in /etc/profiles. if people do thi=
s,
> > do the gnustep-make wrappers scripts still work? i recall having had
> > problems with this. eric/doko can you check this?
> Sourcing scripts as described in Steve's first email is against policy.
it's recommendation and not a must. therefore *all* of my 20 or so packages
depending on gnustep *have* a wrapper script so nobody ever is forced to
run any script from /usr/lib directly.


> > some of my gnustep packages have wrapper scripts in /usr/bin or /usr/ga=
mes
> > some do not. the mass bug filing sucked. one mail for one message would=
 have
> > been enough.=20
> Sorry about the bugs reports that may have been filed against FHS=20
> compliant packages.  Based on the facts that most incompatible are=20
> dependant on gnustep-base1 or gnustep-make is how I chose packages to=20
> bug report upon.
i just checked for gods sake, all of my packages *in* debian have the wrapp=
er script.

the ones not yet uploaded have at least an entry in "menu".=20

g=FCrkan

--Signature=_Sun__27_Jun_2004_12_35_58_+0200_razwl72k7UwB2gY6
Content-Type: application/pgp-signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFA3qMPfI36WwmZVIsRAvnTAJsFBTRF1d2ypeXBWxSp7kBvFEBdJACglGxS
J3p+wECAKiWTyq5EZLt08iE=
=3v53
-----END PGP SIGNATURE-----

--Signature=_Sun__27_Jun_2004_12_35_58_+0200_razwl72k7UwB2gY6--