I have a database which is fully generated by a (java) persistence layer; so
I have practically no layer where pure database modelling will be done, no
Erwin no nothing.
All I can do is giving the persistence layer hints on which objects actually
make persistent(own table), and other objects whose data will be kept with
parent or child objects(sub and super).

Now that makes tuning kind of tough, even to get the picture about the actual
data modell is rather difficult.

What I am thinking of, is that I use db2look to generate the TeX output of
the data model.
I have heard in TeX it can even return the graphic output returning also
constraints.

Two problems:
First big one: I do not get the teX output compiled with TeX. I have no TeX
on AIX, on windows it returns only error messages and an empty dvi file. TeX
itself works fine with any other document.  Anybody has an idea here ?

Second: Will informational constraints be included ? We have no constraints
implemented, generally the persistence layer takes care of that. But I was
generally thinkting to add informational constraints also to ensure indexes
are there etc.
Well, second does not help me if first one is not solved ...

Any other option on how to intelligently visualize the schema is highly
appreciated. I have no tools like Erwin or so at hand

I've never been in such an environment, any help appreciated.

Thanks

Juliane