Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

open-source / freeware / abandoware Topspeed distribution

0 views
Skip to first unread message

Paul Attryde

unread,
Oct 4, 2002, 2:29:31 PM10/4/02
to
"Jim Granville" <jim.gr...@designtools.co.nz> wrote in message
news:3D4081...@designtools.co.nz...
> mark...@cix.compulink.co.uk wrote:
> >
> <snip>
> > What became of the sundry attempts to get
JPI/TopSpeed/Clarion/whoever
> > to do some sort of free-distribution or open-source deal?

Unlikely to happen, or at least not anytime soon :(

Bob Brooker from Softvelocity (the company that currently has the
rights to the TS compiler) can sometimes be found on the Wedneday
Clarion IRC chat.

In an IRC chat at the end of August (apologies for so taking so long)
I asked him what the chances were of getting some sort of open-source
/ freeware / abandoware release. Apparently they're still selling the
JPI compilers, mostly to handheld/embedded developers, so they're not
likely to do anything soon.

Paul

Session Start: Wed Aug 28 09:55:03 2002
Session Ident: #cw-talk
[10:09] <PaulAttryde> Bob, any decision on if what you're going to do
with the old DOS compiler re: making them freeware, abandonware, open
src etc?
[10:09] <svbrooker> The concept of JNI calls would permit reuse of a
ton of existing code as many clarion ide "features" are dialogs
[10:10] <TheSylkie> Now THAT's cool.
[10:14] <svbrooker> Paul: We still sell new licenses for CPD and CDD.
If we were to "open source" CPD our copyright on the Clarion language
becomes less certain.
[10:15] <PaulAttryde> Bob: Sorry, my mistake, I meant the JPI CPP/M2
stuff :)
[10:18] <TheSylkie> I still use 2.108, on occasion. <G>
[10:18] <svbrooker> Still sell a fair number of new seatrs for those
too, mostly cpp
[10:18] <svbrooker> The JPI (thin) memory model and compiler
efficiency is still in demand for handheld developers
[10:20] <PaulAttryde> OK, I didn't realise you still sold any. I
assume anyone could call sales and get a price?
[10:21] <svbrooker> Yes, with the understanding that it is
"unsupported". Most recent purchasers are corporate types OEMing the
compiler
[10:22] <PaulAttryde> Well, that's why the M2 folks want you to open
it up, because they can't get any support on it :)
[10:22] <PaulAttryde> Thanks anyway.


Jim Granville

unread,
Oct 4, 2002, 4:08:53 PM10/4/02
to
Paul Attryde wrote:
>
> "Jim Granville" <jim.gr...@designtools.co.nz> wrote in message
> news:3D4081...@designtools.co.nz...
> > mark...@cix.compulink.co.uk wrote:
> > >
> > <snip>
> > > What became of the sundry attempts to get
> JPI/TopSpeed/Clarion/whoever
> > > to do some sort of free-distribution or open-source deal?
>
> Unlikely to happen, or at least not anytime soon :(
>
> Bob Brooker from Softvelocity (the company that currently has the
> rights to the TS compiler) can sometimes be found on the Wedneday
> Clarion IRC chat.
>
> In an IRC chat at the end of August (apologies for so taking so long)
> I asked him what the chances were of getting some sort of open-source
> / freeware / abandoware release. Apparently they're still selling the
> JPI compilers, mostly to handheld/embedded developers, so they're not
> likely to do anything soon.

Thanks for asking..

I would use more carefully chosen semantics :)
'Open source' triggers flags in the corporate types, as they worry
about copyright etc. 'Abandonware' give them no brownie points.

He is also a little confused about the old products, thinking
JPI == cpp :) - so I would mention exact models
JPI TSM2 V1.17 / V3.01.

I believe better is Borlands pathway, of a community/museum,
where the 10+ year old SW is placed. I believe even microsoft does
something similar. Not open source, but gets it into the student
base, where it can still do usefull work.

On that line, I did ask Borland to place the Turbo Modula2 on
the museum, with positive feedback, but no sign yet.
Turbo Modula-2 is CP/M, so needs a CP/M emultor, but I did
find a couple of sites teaching using this.
The Z80 is having something of a comeback, as the eZ80 looks to
drag it into this century.

Better than BTM2, for the Z80 might be this TML2, close to
Modula-2, and a little more open/active than BTM2.
see :
http://home.swipnet.se/~w-68269/tml2/tml2.htm

Both would have their place - TML2 (cross compiler)
would be good for deeply embeded eZ80 systems, without OS,
and BTM2 (or even TP3 :), could appear on an eZ80 miniPC,
running a CP/M superset.

Q: Does TML2 belong in the Modula-2 FAQ ?

It is a pascal/m2 hybrid, but mostly 'm2 tolerant', or 'M2 leaning' :
( there are examples at http://home.swipnet.se/~w-68269/tml2/tml2.htm )

PROCEDURE tty_read(): char;
BEGIN
while in(ttycmd) = #0 loop
end loop;
return in(ttydat);
END tty_read;

also has ELSIF, and the begin..end pairs are gone, and has RETURN
in functions.

mark...@cix.compulink.co.uk

unread,
Oct 4, 2002, 5:08:20 PM10/4/02
to
In article <ankmhv$ev0n4$1...@ID-161168.news.dfncis.de>, pa...@attryde.com
(Paul Attryde) wrote:

> Unlikely to happen, or at least not anytime soon :(

Thanks for filling us in. They'd better not rely overmuch on the
handheld marked since that's moving towards ARM, and for new x86
development who wants to use a compiler where they have no guarantee
additional copies will be available (legally) next week?

-- Mark Morgan Lloyd
-- mark...@cix.co.uk
--
-- [Opinions above are the author's, not those of his employers or
colleagues]

mark...@cix.compulink.co.uk

unread,
Oct 5, 2002, 6:29:55 PM10/5/02
to
In article <3D9DF5...@designtools.co.nz>,
jim.gr...@designtools.co.nz (Jim Granville) wrote:

> I would use more carefully chosen semantics :)
> 'Open source' triggers flags in the corporate types, as they worry
> about copyright etc. 'Abandonware' give them no brownie points.

Frankly, I think we can write off TopSpeed as open source- I'm told by
somebody that knows that it took a week of training before somebody was
able to compile the compilers and IDE.

I suggest that we throw our weight behind Gaius and the GNU
implementation.

Paul Attryde

unread,
Oct 7, 2002, 3:56:53 PM10/7/02
to

"Jim Granville" <jim.gr...@designtools.co.nz> wrote in message

news:3D9DF5...@designtools.co.nz...


> Paul Attryde wrote:
> >
> > Unlikely to happen, or at least not anytime soon :(
> >
> > Bob Brooker from Softvelocity (the company that currently has the
> > rights to the TS compiler) can sometimes be found on the Wedneday
> > Clarion IRC chat.
> >
> > In an IRC chat at the end of August (apologies for so taking so
long)
> > I asked him what the chances were of getting some sort of
open-source
> > / freeware / abandoware release. Apparently they're still selling
the
> > JPI compilers, mostly to handheld/embedded developers, so they're
not
> > likely to do anything soon.
>
> Thanks for asking..
>
> I would use more carefully chosen semantics :)
> 'Open source' triggers flags in the corporate types, as they worry
> about copyright etc. 'Abandonware' give them no brownie points.

Well, you can certainly try asking Bob using different terminology,
although I don't think it's going to do you any good. Not that it
bothers me personally (I haven't written any M2 code in about 10 years
now) but I would have liked to have seen something come of it, rather
than see it drift off into nothing.

As Mark says, I think you might as well write them off and concentrate
on something else

Paul

Marco van de Voort

unread,
Oct 8, 2002, 12:46:13 PM10/8/02
to
In article <annp53$5h4$1...@thorium.cix.co.uk>, mark...@cix.compulink.co.uk wrote:
> In article <3D9DF5...@designtools.co.nz>,
> jim.gr...@designtools.co.nz (Jim Granville) wrote:
>
>> I would use more carefully chosen semantics :)
>> 'Open source' triggers flags in the corporate types, as they worry
>> about copyright etc. 'Abandonware' give them no brownie points.
>
> Frankly, I think we can write off TopSpeed as open source- I'm told by
> somebody that knows that it took a week of training before somebody was
> able to compile the compilers and IDE.

It is exactly the same thing with Borland.

These companies don't seem to understand that they hindering their own old
customers with this kind of defensive behaviour, and are sending a message
"lifetime of this development tool, and the applications will be over when
we don't officially support it anymore".

I'll never make the mistake of buying something from such a company again.



> I suggest that we throw our weight behind Gaius and the GNU
> implementation.

There will problems again, when you have to reverse engineer each and every
library again. Standard libraries will never be 100% substituable (I'm quite
afraid for mthreaded apps and module Window)

Gaius Mulley

unread,
Oct 8, 2002, 7:48:57 PM10/8/02
to

> There will problems again, when you have to reverse engineer each
> and every library again. Standard libraries will never be 100%
> substituable (I'm quite afraid for mthreaded apps and module Window)

yes unlikely to be a set of exactly the same - but this is more a
reflection on the shift in OS (from DOS -> GNU/Linux or DOS -> NT).
Of course the impact of this all depends on the applications.
Multi-threaded apps shouldn't fair too badly as long as they used a
standard executive rather than using IOTRANSFER, and as we've
discussed before TRANSFER, NEWPROCESS can be implemented easily in
pthreads.

An equivalent (ish) IOTRANSFER probably can be written with pthreads
- I hope so as I have a coursework on this :-) Applications which used
an extra layer of abstraction via some sort of executive will benefit
even more..

Gaius

Marco van de Voort

unread,
Oct 9, 2002, 3:49:27 AM10/9/02
to
In article <877kgs7...@isode.i-did-not-set--mail-host-address--so-shoot-me>, Gaius Mulley wrote:
>
>> There will problems again, when you have to reverse engineer each
>> and every library again. Standard libraries will never be 100%
>> substituable (I'm quite afraid for mthreaded apps and module Window)
>
> yes unlikely to be a set of exactly the same - but this is more a
> reflection on the shift in OS (from DOS -> GNU/Linux or DOS -> NT).
> Of course the impact of this all depends on the applications.
> Multi-threaded apps shouldn't fair too badly as long as they used a
> standard executive rather than using IOTRANSFER, and as we've
> discussed before TRANSFER, NEWPROCESS can be implemented easily in
> pthreads.

I was more thinking about
- matching the TS rtl 100% (we did this for BP/Delphi, not fun) Not only
in interface, but also good enough in implementation to be a substitute.
- keeping that RTL itself mthreaded, while retaining that same behaviour.

mark...@cix.compulink.co.uk

unread,
Oct 9, 2002, 2:43:56 PM10/9/02
to
In article <slrnaq62ul....@toad.stack.nl>, mar...@toad.stack.nl
(Marco van de Voort) wrote:

> afraid for mthreaded apps and module Window)

As I have commented before, I'm still hoping to find something that will
do the job of this using termcap/curses. I know that you say there's
something similar with FP but have noted the comments about it being
rewritten to avoid copyright code- can you confirm it is still
text-oriented or has it migrated to X?

Marco van de Voort

unread,
Oct 9, 2002, 5:29:01 PM10/9/02
to
In article <ao1tdc$pc8$1...@thorium.cix.co.uk>, mark...@cix.compulink.co.uk wrote:
> In article <slrnaq62ul....@toad.stack.nl>, mar...@toad.stack.nl
> (Marco van de Voort) wrote:
>
>> afraid for mthreaded apps and module Window)
>
> As I have commented before, I'm still hoping to find something that will
> do the job of this using termcap/curses. I know that you say there's
> something similar with FP but have noted the comments about it being
> rewritten to avoid copyright code- can you confirm it is still
> text-oriented or has it migrated to X?

It is still textorientated. There are two versions, one based on Borlands
code, and one own (that is text orientated, though one could slap a graphics
driver underneath).

Most of the omissions and other instabilities in the own one have been fixed
except some rather fundamental one. The coordinate origins differ (off by
one problem).

This is not really bad for new apps, but is a killer for full TV compability.

mark...@cix.compulink.co.uk

unread,
Oct 9, 2002, 6:48:37 PM10/9/02
to
Thanks :-)

In article <slrnaq97st...@toad.stack.nl>, mar...@toad.stack.nl

(Marco van de Voort) wrote:

> can you confirm it is still text-oriented or has it migrated to X?

> It is still textorientated.

-- Mark Morgan Lloyd

0 new messages