Page 6 of 7
Re: QL/E - The QL Environment (QLE)
Posted: Thu May 04, 2023 8:46 pm
by QLvsJAGUAR
QL/E on Q68 in High Colour modes with automatic positioned QDT icons.
Edit: Ah, now I see, the Y-axis is the problem. Added to the to-do list of v3.23.
Gesendet von iPhone mit Tapatalk
Re: QL/E - The QL Environment (QLE)
Posted: Thu May 04, 2023 10:56 pm
by Derek_Stewart
Hi Urs,
Another problem I have come across, is that, I have move the position of the QDT Icons and save the QDT Desktop.
On bootup of QL/E the saved Desktop file is not loaded. I thought QDT date stamps the Desktop congiguration file and then loads the latest Desktop file.
Re: QL/E - The QL Environment (QLE)
Posted: Fri Dec 29, 2023 6:29 pm
by QLvsJAGUAR
Dear QLers,
for me it was a year with not that many hours QL screen time. Nonetheless I finally managed to spend two days straight to complete another QL/E release.
QL/E v3.23 (Codename "Midnight Sun", 2023 edition) is out now!
Details about it, Download & More as usual on the sinclairql.net website, link see message footer.
I wish you all a Happy and Healthy New Year 2024! May we have a chance to celebrate "QL is 40" in one way or another!
Cheers,
Urs
Re: QL/E - The QL Environment (QLE)
Posted: Fri Dec 29, 2023 6:31 pm
by QLvsJAGUAR
QL/E v3.23 final tests completed!
Gesendet von iPhone mit Tapatalk
Re: QL/E - The QL Environment (QLE)
Posted: Fri Dec 29, 2023 8:27 pm
by stephen_usher
Just a quick bug report:
Hardware:
- BBQL JS ROM
- Tetroid SGC
- QIMSI
Using QLE.WIN
Boot program:
Line 931: bad name
Looks like DISP_TYPE isn't defined.
Re: QL/E - The QL Environment (QLE)
Posted: Sat Dec 30, 2023 2:29 pm
by stephen_usher
With regards to the QIMSI, it requires the mouse driver to be installed after the Pointer Environment has been initialised.
I can't see in the documentation how you add an automatic LRESPR of a driver at boot time rather than manually opening a SuperBASIC prompt and typing the commands in. Is there a way of doing this rather than hacking the start-up BASIC program?
Re: QL/E - The QL Environment (QLE)
Posted: Tue Jan 02, 2024 8:07 pm
by QLvsJAGUAR
stephen_usher wrote: ↑Fri Dec 29, 2023 8:27 pm
Just a quick bug report:
Thank you! I like feedback.
stephen_usher wrote: ↑Fri Dec 29, 2023 8:27 pm
Hardware:
- BBQL JS ROM
- Tetroid SGC
- QIMSI
Using QLE.WIN
Boot program:
Line 931: bad name
Looks like DISP_TYPE isn't defined.
FYI: In QL/E, on non-SMSQ system, the DISP_TYPE machine code command is generated by the Q2S*_bin toolkit.
Q2S was added to QL/E v3.21 (in 2021). The way DISP_TYPE is called was changed in QL/E v3.22 (in 2022). Since then the boot program does:
This works well on SMSQ (native, without Q2S) and on Minerva (with Q2S). But as you noticed, makes a problem on QDOS' JS (also with Q2S).
I can offer the following hot-fix. Change line 931 in boot to
OK for you, for now?
On the long run, I need to check with Per regarding Q2S.
Best regards,
Urs
Re: QL/E - The QL Environment (QLE)
Posted: Tue Jan 02, 2024 8:16 pm
by QLvsJAGUAR
stephen_usher wrote: ↑Sat Dec 30, 2023 2:29 pm
With regards to the QIMSI, it requires the mouse driver to be installed after the Pointer Environment has been initialised.
I can't see in the documentation how you add an automatic LRESPR of a driver at boot time rather than manually opening a SuperBASIC prompt and typing the commands in. Is there a way of doing this rather than hacking the start-up BASIC program?
Yes. Have a look for "user exit (PROCedure QLE_user_exit1) to LOAD a custom set of software (using LRespr and EXexute)..." in section "Good to know/Hints" in the QLE_manual*txt.
Works for you?
Re: QL/E - The QL Environment (QLE)
Posted: Tue Jan 02, 2024 9:10 pm
by pjw
QLvsJAGUAR wrote: ↑Tue Jan 02, 2024 8:07 pm
<>
I can offer the following hot-fix. Change line 931 in boot to
<>
Qdos doesnt do SELect ON very well. It only accepts float variables or functions. Minerva is a little better as it also accepts integer% variables and functions, but also doesnt like m/c functions like DISP_TYPE (or PI, for that matter).
SMSQ/E does all except string types
Re: QL/E - The QL Environment (QLE)
Posted: Wed Jan 03, 2024 10:38 am
by pjw
QLvsJAGUAR wrote: ↑Tue Jan 02, 2024 8:07 pm
stephen_usher wrote: ↑Fri Dec 29, 2023 8:27 pm
Just a quick bug report:
Thank you! I like feedback.
stephen_usher wrote: ↑Fri Dec 29, 2023 8:27 pm
Hardware:
- BBQL JS ROM
- Tetroid SGC
- QIMSI
Using QLE.WIN
Boot program:
Line 931: bad name
Looks like DISP_TYPE isn't defined.
FYI: In QL/E, on non-SMSQ system, the DISP_TYPE machine code command is generated by the Q2S*_bin toolkit.
Q2S was added to QL/E v3.21 (in 2021). The way DISP_TYPE is called was changed in QL/E v3.22 (in 2022). Since then the boot program does:
This works well on SMSQ (native, without Q2S) and on Minerva (with Q2S). But as you noticed, makes a problem on QDOS' JS (also with Q2S).
I can offer the following hot-fix. Change line 931 in boot to
OK for you, for now?
On the long run, I need to check with Per regarding Q2S.
Best regards,
Urs
I was listening to the news while responding. I see now that I explained
this very badly!
Under Qdos SELect only works with floating point variables or floating
point SuperBASIC functions. So SELect ON DISP_TYPE - or any other machine
code function - will not work. The line returns Bad name.
Minerva appears to accept a machine code function in that the parser
doesn't throw a Bad name error, but in fact
it doesn't work! Try this
rather dumb example:
Code: Select all
10 SELect ON PI
20 = 0 TO 4: PRINT 'ok'
30 = REMAINDER : PRINT 'Huh?'
40 END SELect
50 :
(I cant think of another common m/c function that doesn't require a
parameter. In Qdos/Minerva you cant even enter a line like SELect ON
PEEK(adr))
Minerva does, however, accept integer% variables - but
not SuperBASIC
integer functions! Minerva even accepts a string parameter, which neither
Qdos nor SMSQ/E accept. See the Minerva manual for details..
The above example works perfectly in SBASIC (SMSQ/E), as does SELect ON
PEEK(adr).
The compilers, Qlib and Turbo have their own foibles, but that's another
story.
In other words, DISP_TYPE is not the bad guy here! For maximum
compatibility stick to floating point variables.
