Let me just add my 2¢ from someone who has had good experience with the Cathedral and is somewhat sceptical of the bazaar. A few, good releases as opposed to `release early, release often' seems preferable to me. Ideally, everything would be distributed with the OS and one would upgrade once a year. I like the idea of maintenance releases, if necessary, as opposed to individual patches. I want maintenance to be possible by someone who doesn't understand the details of the code he is maintaining. My experience is mostly with VMS and related stuff. There is some free VMS stuff, but this is also `cathedral mode' from what I can tell, being developed by the `select few'. I have the impression that a lot of LaTeX developers prefer VMS (including Lamport?) so I might be preaching to the choir but I'm sure many people on this list are fans of Stallman, gnu, Linu{x|s} etc. I think a bazaar approach to LaTeX would make updating even more difficult than it is now. (Standard LaTeX is OK, but stuff in contrib has to be done pretty much by hand.) And I think the documentation and user-level syntax could be more uniform than it is now. Although I enjoy system maintenance, I would prefer to do `real work' and not have to daily monitor the net for the latest updates. As a VMS fan, it would be interesting to see something which has, in my view, the advantages of VMS which was developed at a bazaar and not at a cathedral. And I understand that things like the compilers for VMS from DEC (Digitally Enhanced Compaq:) are developed by a really small group of people. And they are really good. Also in the VMS world is the OSU HTTP_SERVER, which is free and which many people say is better than commercial stuff. But it's the work of one person, so very much in the cathedral (or at least country church:) mode. The fact that linux is `as good as commercial unix' from my point of view doesn't necessarily mean that linux is very good; it could also mean that commercial unix is very bad. Probably, the truth lies somewhere in-between. Getting back to the original thread, from my point of view a goal in using LaTeX for journals is that one can use the full strength of LaTeX. In practice, one finds that one's own favourite packages are `not supported' at the journal end. In particular, this applies to natbib. So I think all essential stuff should be in the core part and not in the contributed part. Since the contrib stuff is not in all distributions etc it's almost as bad as having to find the various packages outside of CTAN---in any case, they have to be installed by hand, one can't count on a colleague/journal being able to process a document written using them etc. And keep in mind the colleague is often not a LaTeX guru but just a normal guy. To me, this implies cathedral mode. -- Phillip Helbig Email ......... [log in to unmask] Nuffield Radio Astronomy Laboratories Tel. .... +44 1477 571 321 (ext. 297) Jodrell Bank Fax ................ +44 1477 571 618 Macclesfield Telex ................ 36149 JODREL G UK-Cheshire SK11 9DL Web ... http://www.jb.man.ac.uk/~pjh/ My opinions are not necessarily those of NRAL or the University of Manchester.