phillip helbig writes:
> This is the way many software packages work. There's the main stuff
> from the main vendor (core) and layered products (contrib) which can
> always be expected to work in combination with each other and with the
> core. There are then retired (frozen) products which used to be part of
> the core or contrib (basic stuff or layered products) and can still be
> expected to work in the future (new stuff won't break them) but nothing
> new is to be expected. Then there is all the use-at-your-own-risk third
> party stuff.
the difference with tex is that diversity has taken hold to such an
enormous extent that even defining a useful core (such as the latex
team did 5 years ago) was a huge undertaking.
as things stand, we have 237 contrib/supported directories, 64
contrib/other (one of which is a `misc' directory).
i regularly find myself installing new `supported' directories, with
but the smallest inspection of what has been submitted -- there just
isn't the time to do things like that. occasionally, i find myself
installaing an `other' directory, which seems to me to cry out for
installation in the `supported' set ... but life's too short for me to
do other than i do -- i spend too much time on ctan as it is.
> In summary, I think we need
> o frozen as a new category
> o moving essential stuff to the core
> o making the distinctions clear to all
this, you perhaps (by now at least) realise, is an enormous set of
tasks. ctan runs entirely on the voluntary effort of people who are
paid to do other things; the latex team likewise -- i can't imagine
where the effort to do it all would come from.
if i could find a way to fund myself to sit at this desk for another 6
months, say, i guess i might make reasonable inroads into such a
programme. [this is essentially impossible, in fact.]