|--==> "SJT" == Stephen J Turnbull <stephen(a)xemacs.org> writes:
>>>>>"Jerry" == Jerry James
Jerry> Steve Youngs <youngs(a)xemacs.org>
>>- Autoloads are built for *everything* in ./modules
SJT> This is what _should_ happen. The autoloads should be there,
SJT> and they should generate an informative error message if the
SJT> .ell isn't present.
I really disagree quite strongly here. IMO, the autoloads should only
be built for modules that are installed. Jerry's solution of wrapping
them in a '(when (file-exists-p "module.ell"))' does the next best
thing and works well.
SJT> Any modules that are distributed with XEmacs should be considered
SJT> standard (eg, md5, gzip).
Except for PostgreSQL and LDAP any other module in core is really
nothing more than sample code. Eventually the other modules (base64,
zlib) will undoubtedly "come online" as will other, yet to be dreamed
of, modules. In the meantime, maybe we should move the currently
"unused" modules into ./modules/sample/ until such time that somebody
breaths life into them.
SJT> The fact that today's sources include clearly optional stuff
SJT> like postgresql and ldap should be considered transitional
Why? The core is full of optional stuff that isn't considered
transitional. I don't think I quite understand what you're getting
Jerry> On the other hand, this doesn't support installing new
Jerry> modules later on that did not exist when the XEmacs binary
Jerry> was compiled. On the gripping hand, that would still be a
Jerry> problem if auto-autoloads.el contained only autoloads for
Jerry> modules installed at build time.
Yeah but it wouldn't take too much time and effort for some smart
cookie to come up with a way for autoload.el to overcome this. :-)
|---<Steve Youngs>---------------<GnuPG KeyID: 10D5C9C5>---|
| XEmacs - It's not just an editor. |
| It's a way of life. |