1. TSR Structure (Bernie)
=========================
I've been considering the stuff that's been
discussed re shrinking EPPPD, and have come to the
conclusion that EPPPD is not properly structured.
I think Clarence has come to the same conclusion.
A TSR should have three sections :
1. A jump to the initialization code.
2. The TSR proper, inc. data and common code.
3. The initialization code.
After the initialization code has done it's job,
it then exits via interrupt 27 (I think), telling
the O/S to keep sections 1 & 2 resident, but
releasing section 3. This means that all the stuff
with configuration option processing should *not
take up any memory* once EPPPD has gone resident.
While this *should* be the case, it clearly isn't!
I don't know how easily the above is achieved in
C (that's what EPPPD is written in, I think),
however. Another thought I had was perhaps the
same thing could be achieved with overlays?
(does C support them?) ... I have not heard of
anyone using overlays for this purpose before,
but it sounds OK in principle (maybe ;-).
Sorry to not have "raised" this earlier, but I had
not realized before, that EPPPD's structure was to
blame for it's "excessive" memory usage ...
2. DPMS Documentation (Bernie)
==============================
If you want DPMS documentation, I can e-mail that
to you, else just enter "DPMS" at Novell's SEARCH
facility. I can't offer help to use this however,
as I have never used any protected mode stuff (not
even XMS) in my own app's ...
2. EPPPD Files (Michael)
========================
Michael, why is the "full" DOS PPPD stuff at
"arachne.cz" version 0.5, instead of 0.6?
Also, why isn't the source available at
"arachne.cz"?
See ya!
Regards,
Joe.
