I've moved new ole to hbwin, because otherwise we will very very
slowly adapt to
it, since very few users would test it.


Not exactly. It just broke all the linking process itself, simply
unusable to test either.

You still didn't post what error you're getting. First you
wrote that you cannot compile, now you cannot link.

Again, we're in unstable state, you can go back to last working
hbwin.lib anytime, it's not obligatory to use the latest version,
if you do though, we'd like to hear about problems.

hbwin.lib - New implementation, with clean concepts, names, OLE, etc.
hbwinold.lib - Old implementation kept for compatibility until 3rd parties
can adapt. Code frozen.


Why to force existing scripts to be changed?

Nobody is forced to use the current SVN, this means nobody
is forced to rename anything. We hope to make hbwin
compatible on the mid run, so for Harbour users who can
wait a little bit, nothing should be changed. If we don't
manage or don't decide to create fully compatible hbwin
implementation, we can think about how to name them to
encourage users to switch to new, yet not break their
scripts.

It should be like:
hbole.lib - As current hbwin - make sure that it compiles and at least links
fine stand alone.
hbwin.lib - As standard old one - and frozen for new work.

I mean two parallel libs ao that user could switch over to other
for testing purposes. Once matured, we could face out hbwin.

hbole and hbwin aren't two parallel libs, they don't provide
the same functionality, they are not plugin replacements for
each other, so what you say isn't an option.

Anyhow I've now read messages from Przemek and Mindaugas targeting
your specific AX problem, I hope it will be committed and we
can move on to next issues. IMO your GTWVG AX code can also quite
easily changed to use real pointers, or call an API. What we
should still keep in mind is that 3rd party code may well have
numeric pointers for some time, so some additional tweaks will be
needed to give a compatibility layer. Such stuff is currently
gathered in contrib/hbwin/legacy.c.

Brgds,
Viktor

_______________________________________________
Harbour mailing list
Harbour@harbour-project.org
http://lists.harbour-project.org/mailman/listinfo/harbour

Reply via email to