The Sinclair QL Forum

RalfR wrote: Thu Jan 09, 2025 1:47 pmDo you invoke Qmon during tests?

No I didn't yet.

RalfR wrote: Thu Jan 09, 2025 1:47 pmIn my opinion, testing only makes sense if someone does it with a Qxx and also programs without toolkits.

I also suspect that toolkits are involved. I wonder if someone with knowledge/interest in QLiberator has a 68040/68060 machine.

Would you be interested in debugging this if you had a Qx0?

Peter wrote: Thu Jan 09, 2025 10:01 am

pjw wrote: Mon Jan 06, 2025 1:06 pmIf we could manage to narrow down where the problem could lie, it might be possible to do something.

Maybe in a first step, we should define a simple BASIC program on which to reproduce it, rather than a large executable like Q-Dock?
Do you own a 68040 or 68060 machine?

Sadly, no. So Im of little use here. Im still trying to understand how Qlib works, but it is a slow business and I dont have much time for looking for a needle in multiple possible haystacks. With time, patience and luck a solution or workaround will be found, Im sure. So to everyone following this thread: Work your corner!

Just in case, heres another one. Its compiled with V3.46, with lines only and no open window, ie "ghost" windows only. The RTM version includes the runtime.

This one has a timer. Compare the timings w/o cache, SBASIC and compiled.
Sorry for flooding you, if you can suggest specific tests I can have a go at producing them..

pjw wrote: Thu Jan 09, 2025 2:16 pm<>I'm still trying to understand how Qlib works, but it is a slow business<>

Unfortunately, there are several places where you can check. It could be a slightly wrong parsed/compiled keyword, it could be the runtimes not correctly using a keyword (or several), the possibilities are not unlimited, but they are varied.

Unfortunately, I don't have a Qxx either.

RalfR wrote: Thu Jan 09, 2025 2:45 pm Unfortunately, I don't have a Qxx either.

I know, but my question was: Would you be interested in debugging this if you had a Qx0?

Peter wrote: Thu Jan 09, 2025 4:44 pm

RalfR wrote: Thu Jan 09, 2025 2:45 pm Unfortunately, I don't have a Qxx either.

I know, but my question was: Would you be interested in debugging this if you had a Qx0?

Yes, of course. I would start with a small program and gradually add more and more S*Basic commands until an error occurs.

However, I have long since lost my knowledge of assembler. So it might make more sense if someone who knows a lot about it does it (if he has time... and you need it), because the cause has to be found.

pjw wrote: Thu Jan 09, 2025 2:40 pm This one has a timer. Compare the timings w/o cache, SBASIC and compiled.
Sorry for flooding you, if you can suggest specific tests I can have a go at producing them..
time.zip
(889 Bytes) Not downloaded yet

Do we even know if the cache handling code in SMSQ/E is working as it
should? Looking through various correspondence and sources, it apears that
copyback has had some issues over the years. The last person who seemed to
know a anything about itwas Thierry. He also wrote the V2.01 update for
SMSQ/E in 2005.

In 2005 any problems with Qlib and copyback caches on Q40/60 must have been
known! Doesnt anyone recall? Does anyone have contact with Thierry?

I dont have the hardware, nor do I have the knowhow. It seems to me that
those who do, have the greater stake in this, and should try to sort it
out. If anyone can point to a concrete Qlib problem, I may try to fix it.
But now my time is up.

Peter,

on your site, there is a file "Cache Management for SMSQ/E" by Mark Swift. But it is not downloadable...."Timeout".

OK, found at Dilwyn's site.

RalfR wrote: Thu Jan 09, 2025 7:04 pm on your site, there is a file "Cache Management for SMSQ/E" by Mark Swift. But it is not downloadable...."Timeout".

It's only a link to Mark's site, which is probably down.