> It turns out, goh/h files are usually created from scripts that translate
> the def/uih files. Some things, however, just don't translate well, so
> they need to be translated by hand. This is especially true of the C stubs,
> which I think are always manually written. We do try to test all of our
Thank you for the info.
Of course, the next step would be extending this to the C stub code (which also
contains a few glitches), the corresponding prototypes and (gasp!) maybe even
the documentation (Objects, Routines, Structures, various quick references, each
for GOC *and* ESP) - in other words, defining the API in a sort of database
fashion... Just dreaming... :-) but I believe that in the long term, such an
approach could improve the quality of the documentation (which I think is rather
good, but quite often inaccurate or incomplete in details) and the effectivity
of its generation quite dramatically.
> Can you define "ceterum censeo"? My Webster's 10th doesn't list it,
I guess I've probably made it up. :-) Just looked it up in a German "famous
quotes" dictionary (Buechmann) to get the names right: It dates back to a Roman
politician named Cato, who is said to have finished all his public speeches with
the phrase "Ceterum censeo Carthaginem esse delendam" - "In addition, I believe
that Carthago should be destroyed."
ciao marcus