[tex-live] free software, DFSG

David Kastrup dak at gnu.org
Sun Jun 6 19:51:52 CEST 2004

plaice at cse.unsw.edu.au (John Plaice) writes:

> On Sun, Jun 06, 2004 at 12:40:50AM +0100, Sebastian Rahtz wrote:
> > John Plaice wrote:
> > 
> > >I see a complete TeX rewrite, that goes well beyond TeX, and without
> > >the cult of 100% TeX compatibility, taking place within 3 years,
> > >at the very most.
> > > 
> > now, tell us, for how many years have you been saying that, John?
> Sometimes it takes years to really figure out what is wrong...
> With software, once a clear design comes up, the implementation
> moves quickly assuming you have skilled programmers around you.
> I believe that with Omega we have reached that point...

My outsider impression of the Omega project is that it is not as much
in need of skilled programmers as of a nanny.

Clicking on the "CVS" link on the Omega home page gives you a
non-HTML-recognized page.  Following the instructions for anonymous
CVS access gives:

bash-2.05b$ cvs -d :pserver:anonymous at omega.cse.unsw.edu.au:/home/cvs/root login 
Logging in to :pserver:anonymous at omega.cse.unsw.edu.au:2401/home/cvs/root
CVS password: 
Fatal error, aborting.
anonymous: no such user

That's not likely to get too many people interested.  One year ago, I
had managed to get hold of a volunteer for writing up user-level
instructions for Omega use which I considered sorely lacking (the most
simple example of writing a trivial Omega/Lambda document in some
encoding is just not there.  Instructions for switching encodings and
what they are are not there.  Instead some very basic instructions for
_writing_ OTPs and stuff.  That's not interesting to a user).  Well I
asked around several people and on the Omega list.  I finally had to
give up.  Nobody was interesting in even slightly coaching a volunteer
to write up sorely needed documentation.

The last mailing list entry from the Omega mailing list is from last
July.  The last entry from the Omega developers mailing list is from
last April.

John, for all practical purposes of normal users Omega is dead.
Whether or not there may be code backing up the design representations
on TeX conferences is irrelevant for the user as long as he has no
chance to ever see or use it.

If you take a project like the Hurd, it has a working web site at
<URL:http://www.gnu.org/software/hurd>.  It has cross references,
articles and stuff.  It has documentation.  It has a publicly
accessible CVS archive that actually works.  It has task files, TODO
lists, mailing lists.  Why, even in the June 2004 archive there are
already messages that are actually on-topic and not SPAM.

If you want to get anywhere in a free software project without a
particular large-scale sponsor with specific needs, things like that
are pretty much mandatory to keep the ball rolling.  The FSF learned
its lesson a bit late.

In contrast, we still have a gaping hole in the TeX world, for a
robust maintainable extendible code base.  I just doubt that Omega2
will ever make it there.  Not because it would not be needed or
technically inferior or whatever, but simply because it stays
completely out of any involvement with actual users and outside
developers.  It won't get friends if it doesn't play with the other
kids, staying indoors and planning the greatest presents to hand out
eventually.  It needs to get out in the streets.

Omega2 is not there, and the Omega developers refuse to even help
people writing generally usable documentation for Omega1's encoding
stuff because it will be obsolete anytime soon.  Or so we hear.

I am afraid that Omega has done its fair share of holding up TeX
development by now, probably surpassing the damage of the "Knuth
cult".  A lot of work has not been tackled since "it will become
obsolete with Omega".

Maybe we will see some usable code base in a tolerably manageable
language for extending TeX into the twenty-first century in the next
three years, like you claim.  But I'd not bet big money on it being
Omega, sadly.

David Kastrup, Kriemhildstr. 15, 44793 Bochum

More information about the tex-live mailing list