[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: FVWM 2.2 officially released



> Julian Gilbey wrote:
> 
> > FVWM 2.2 has just been released officially, so that FVWM 1.x is no
> > longer supported or maintained.
> 
> You mean by the upstream maintainers?  In any case, this leaves it free
> for someone else to adopt if they want to.  Free software doesn't die,
> it just fades away, replaced by something newer on most systems, but
> often lingering in hacked up and still *very* useful versions on odd
> systems here and there.

Yes, it's no longer supported upstream.  But you wouldn't expect much
support for libc4 nowadays either.  And for those individuals with
hacked up versions, they probably know what they are doing.  But I'm
not sure that it's really worth maintainers' time trying to maintain
an archaic version of a package when a new one is now released: no-one
really "maintains" teTeX-0.4 any more, for example: all of Thomas's
teTeX time seems to be devoted to teTeX-0.9-beta, which has been in
development for over two years, if I recall correctly.  And there is
no Debian teTeX-0.4 package any more -- it has long since been
obsoleted.

> > But when I do, I have an interesting problem: how to do the
> > conversion.  There is a script available for converting .fvwmrc files
> > to .fvwm2rc files
> 
> An ugly, frightening, and IMO extremely dangerous script.  One that may
> well fail miserably if it doesn't have write access to the home
> directories of users on remote NFS/NIS-linked systems.

There were two parts to the script: the script which did the actual
conversion, and the postinst section which offered to do this to all
users' files.  The latter has been removed.

> > and I can remove the fvwm package from ftp.debian.org.
> 
> NO!!!  Orphan it if you must, but leave the option open for someone else
> to pick it up and support it if they want to.  Maybe someone from Debian
> will become the new upstream maintainer as well as the Debian
> maintainer.

I have only just picked it up -- it's been orphaned for ages.  I
intend to upload a new Debian fvwm package which basically says: it's
orphaned upstream, the Debian package is still provided, I will accept
patches, but I'm not going to put in time into trying to fix essentially
upstream problems: the upstream authors are only supporting FVWM 2.

> I know it's hard for some developers to believe, but not everyone wants
> to live life on the cutting edge.  Some are even extremely conservative,
> and want to stick with something that they know works, and that they're
> comfortable with hacking.

Fine by me -- fvwm will continue to be provided, but I will not try to
solve non-Debian problems for people.  My work will be more than cut
out enough trying to solve the problems with fvwm2: there are over 45
open bug reports against it at present.

> Let me just ask something here:  fvwm1 is smaller than fvwm2, yes?  That
> alone is a reason why some people might prefer fvwm1, even with no
> upstream support.

Fine.  If people are prepared to use a package with no upstream
support, that's fine.  But it's had no Debian support for ages
either, although it's only been officially offered up for adoption for
about a month.

> > But how can I ensure that people know to use the fvwm2 package?
> 
> It's not a problem.  Don't.  You can suggest it, but don't try to cram
> it down people's throats.  Orphan it, and if it *really* is time for it
> to die, then no one will adopt it, and eventually it'll get dropped.

But I've already had requests for it to be kept, so I will keep it in
the distrib.  I guess I can get some idea of how many people are using
it by looking at the number of bug reports.

> > (1) Have fvwm become an empty package (priority extra) which Depends:
> >     fvwm2.  OK.
> 
> This is a HORRIBLE solution!  You're forcing people that don't WANT to
> switch to switch.

Accepted.

> > (3) Use some nice, new Replaced-by: feature or similar.  If it
> >     exists.
> 
> Did I miss (2)?  In any case, whatever you do, it shouldn't force
> anybody to upgrade unless and until they want to do so.

No, I did.  Sorry.  ((2) was originally an even worse solution.)

> How would you feel if the emacs maintainer decided that vi was now
> obsolete, and it was time to force everyone to "upgrade" to emacs?  I
> realize it's not a perfect analogy, but to people who still prefer fvwm1
> (few as they may be), it's probably a good one.  Just because the
> current upstream maintainers no longer want to bother with fvwm1, that
> does NOT mean that the users are ready to abandon it.

No, but how many users are protesting that we don't ship an "original"
vi, but only things like nvi.  And we don't ship a traditional sed or
awk for that matter either....

   Julian

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

  Julian Gilbey, Dept of Maths, QMW, Univ. of London. J.D.Gilbey@qmw.ac.uk
             Debian GNU/Linux Developer.  jdg@debian.org
       -*- Finger jdg@master.debian.org for my PGP public key. -*-


Reply to: