DeskTopTracker

I saw in the ADFFS download a 32bit version of DeskTopTracker module.

I think the App gave up working a long time ago – was it SArm?

It now seems to work ok with vrpc-dl r0 4.02.

So I had a go at 32bitting the player – the front end seems to work – vrpc-dl – ro5.25 – but the output of the two modules dtt & Qtm – sound is a bit strange. (similar effect with RPCEmu ro5.25)

Just tried it on the Iyonix ro5.23 – sound seems to work ok there.

Just points to a zero pain -in the Dtt module – at &7c0 ldr r6,[r4],#4 r4=0.

dtt v2.05
qtm v1.45c

Do me a favour. Create a thread on the SoundTracker development forum along with the file you’re testing with, so I can reproduce the issue(s).

Just tried to join (NetSurf) – but refused :-(

You have provided an invalid anser to the question.

Acorn quiz!

I see that there is a later version of ‘DesktopTracker’ module – I seems to have v1.12 (01 May 1994) – is this available?

Thanks

Just tried to join (NetSurf) – but refused :-(

Well…I did download the NetSurf source onto my RISC OS computer to have a stab at doing some development work and clearly the ADFFS forums need something because they don’t render very well and insist on downloading the zip files as “file/php”.

So there might be a little tiny microscopic pin-prick of light at the end of a very long tunnel, which doesn’t help you much at the moment but I am in the mood for spreading positivity (even though I blitzed my machine trying to update to 5.24 and haven’t started development yet…)

You have provided an invalid anser to the question.
Acorn quiz!

Where was that? What was the question?

I can’t imagine there’d be too many people on the non RISC OS side that would know Acorn trivia…

In the end I used RO ‘Firefox’ – which worked – but I still needed FFox to log on to the site.

RO ‘Firefox’ sources (Bit old) are available I believe – just another option for you Glen :-)

Just tried to join (NetSurf) – but refused :-(

You’ll need to use a PC as most RISCOS browsers haven’t kept up with browser standards. It might work with Otter which is possibly the most up to date browser.

I see that there is a later version of ‘DesktopTracker’ module

The most recent version I’m aware of is 1.15 which came bundled with Global Effect, I started converting it to 32bit but put it on the back burner as I modified DTT instead.

P. Naulls Firefox 2008 – seems to do the trick – ten year old prog – where have all the years gone :-(

Is DesktopTracker module v1.15 available to download?
Thanks

Ref http://www.tribbeck.com/software/

A copy of DesktopTracker 26bit including mod v1.15 (23 Sep 1994)

Also a 32bit copy of !Sonar – sound editor.

Mmm – I wonder which version of the OS – did DTT work?

What is going on in that module v1.15!!

Since Jon has RedSq – I gave my version of DTT a try
RO5.25 – and it worked :-)

So it works vrpc-dl ro4.02 but not ro5.25 – sound curupted.

RedSq is SArm vrpc-DL is plain Arm.
Why work ok in one OS – but not the other?

Why work ok in one OS – but not the other?

Load ADFFS in the background and see if the sound issue is resolved.

Puzzle

V2.60 (21-Dec-16) doesn’t seem to make any difference.

If if quit !DeskTracker the sound continues – with sound still wrong.

There is a small app which plays Qtm files – that seems to produce incorrect sound output.

RedSq ro 5.25 (23 Apr 2018) works ok.
vrpc-dl 4.02 works ok
vrpc-dl 5.25 (23 Apr 2018) poor music.
Iyonix 5.23 seems to work ok.

Use the same /HostFS/!Boot.

While the music is playing poorly – vrpc-dl ro 5.25 – if I rotate a Dir around the centre of the Desktop – or indeed- a StrongEd txt window – the music sounds correct.

When I stop – back to corrupted music :-(

[edit 1]
Updated vrpc-dl to ro5.25 (09 Jul 2018)

Still the same effect.

[edit 2]
!Zap on ro5.25 (09 Jul 2018)
When a file loaded as text – is changed to ‘Code’ – Zap gives an error and crashes.

back to 5.25 (23 Apr 2018) for the moment.

*rmkilling the ‘Portable’ module – seems to stop the Breakup with the music with vrpc-dl ro5.25.

Loading the module again – problem reappears.

So will have to look though the ‘Portable’ code.

Is there a way of testing the module Portable’s SWI’s?

ie calling them from BASIC – and knowing what the response should be.

Just points to a zero pain -in the Dtt module – at &7c0 ldr r6,[r4],#4 r4=0

Without seeing your code or the tune in question, I’m guessing you’ve not called DeskTracker_SetMemory with the correct parameters. R0 and R1 should both point to the tune you’re attempting to play.

swi DeskTracker_SetMemory – does’t seem to be used by !DeskTracker.
Email sent to Jon.

Is there any extra info on ‘swi Portable_Idle’ other than what is on the Rool site?

Has StrongHlp OS ‘file’ been updated to enclude ‘Idle’?

Thanks

Just had a go with ‘Wimp Zarch’ (vrpc-dl ro5.25 – in a Window-windows mode) – able to read Mips.

When started up – ‘Wimp Zarch’ goes like the clappers – 370Mips.

When it stops to take a breather – it drops down to about 40Mips – and when it continues slow (auto mode).

When I move another window – away it goes again.

With ‘DeskTracker’ – when it starts playing a tune – it drops to about 5Mips.
(no wonder the sound drops off)

Sounds like VRPC is being over-eager with how long Portable_Idle puts the system to sleep for. A problem to take up with the VRPC developers.

Mmm – something about a squadron of ‘pigs’ comes to mind :-(

Running ’Rick’s ‘C’ article on ‘USBeep’ in !TextEase – with it being read out – the Mips range between 60Mips and about 150Mips – drops to about 2Mips when doing nothing.

So the system works.

In the ‘Portable_Idle’ there seems to be a loop calling a address in the podule area – until a certain value is returned.

What decides when vrpc-dl slows down?
Is this the Tracker mod job?

[edit 1]
Just tried !DigitalCD – that works with the same tracker – Mips go up/down as required.
(uses TimPlayer)

What decides when vrpc-dl slows down?
Is this the Tracker mod job?

It’s mainly the Wimp – it contains some code which counts the number of idle poll events per second, and then drops CPU speed + starts using Portable_Idle if the number of events reaches a certain threshold.

swi DeskTracker_SetMemory – doesn’t seem to be used by !DeskTracker.

If DeskTracker is C, DeskTracker_SetMemory will be called via _kernel_swi. You’ll need to search for MOV R0,#&41 followed by two ADD’s to get the correct SWI number (&44B41)

The ZeroPain backtrace should show you where it’s being called from.

There are two swi’s DeskTracker_SetMemory one uses mvn r0,#0 -1 onto the stack. (Read)

The other is referred to from the Iyonix zero pain file.

Still trying to work out why the vrpc-dl ‘Portable’ module is having trouble with the DeskTracker/QTMTracker modules?

I wonder what RO is – when swi Podule_Idle is called?

With RO4.02 when you go to the command line – the Mips speed up – with RO5.25 they slow down.

DigitalCD seems to work ok with the same file.

Louder volume output too.

Still with swi Podule_Idle – what should R0 equal when called – value seems different between ro4.02 and ro5.25

The swi podule_idle uses it-something like this.

    LDR     R10,[R12,#4] ;r10=2
    TEQ     R10,#0
    MOVeq   R0,#0
    msreq   cpsr_f,#&10000000
    BVS     error_exit
    LDR     R11,[R12,#8] ;get podule address
    ;what should r0 equal?
    STRB    R0,[R11,#8] ;in podule space
.loop%
    LDRB    R0,[R11,#8] ;in podule space
    TEQ     R0,#0
    BNE     loop%%
    LDMFD   R13!,{R9-R11,R14}
    mov     pc,r14

• * * * * * * *

*soundgain 7 = max volume?

It is a pity the comand ‘*soundgain’ doesn’t give its present value.

The other is referred to from the Iyonix zero pain file.

And is it passing R0=R1?

I’ve emailed Jason Tribbeck to see if the source can be released, hopefully that will help resolve the issue and allow us to 32bit the original DeskTracker Module as well.

Still with swi Podule_Idle – what should R0 equal when called – value seems different between ro4.02 and ro5.25

It shouldn’t matter. I expect that VRPC only uses the write to the podule as a trigger for putting the emulator to sleep – the actual value being written isn’t likely to matter.

*soundgain 7 = max volume?

0 = max volume, 7 = higher than max :-)

https://www.riscosopen.org/wiki/documentation/show/*SoundGain