Frank.Warren@f42.n161.z1.fidonet.org (Frank Warren) (11/05/90)
Despite what a snotty self-proclaimed MSCS weenie says, JPI has some
problems with memory allocation in V2.0. I was not aware that this was
still happening under 1.06 as I'd not heard much about it.
The old problem used to be a confusion on JPI's part about using
call 4A instad of 48 to DOS, and modifying everything while trying to
tell what was free and available.
Stony Brook is a *very* good alternative. Logitech has bought a
copy to distribute as their new 4.0 compiler and, happily, that means
that Stony Brook also works with things like CodeView and MultiScope
debuggers now.
My suggestion is to port. JPI won't be stable for a lon time yet.
It's been six months, their kludges aren't fixing the source of het
problem (a Modula-2 compiler cloned off the C model compiler they did
over a year ago). I've given up on JPI. Their tech support doesn't
know how to fix the problems, and has publicly invited everyone who has
complaints to either port to Stony Brook or use version 1.17 and kindly
don't mention embarassing things like bugs, or ask if there are reports
of them by others as they don't want to discuss it.
Overall the picture is pretty dismal.
Good luck, but whether it's Stony Brook or JPI's old 1.17, I think
you'll find that v2.0 isn't going to cut it.
--
uucp: uunet!m2xenix!puddle!161!42!Frank.Warren
Internet: Frank.Warren@f42.n161.z1.fidonet.org