Digital Symphony Installer

I knew it was a remote chance but I managed to image my Digital Symphony installation floppy. Unsurprisingly it crashes RPCEmu.

Does anyone know how it’s protection worked?

Has Oregan been approached about their old RO products?

I had a brief online conversation with the Author(Gil Damoiseaux) of Digital Symphony just last week on YouTube. He says he still has his sources (if the media hasn’t got corrupted over time).
He doesn’t do ARM coding anymore, so he wouldn’t be converting the apps to be 32-bit compliant. I have no idea if he could be convinced to release his source code into the wild.

The conversation was → https://www.youtube.com/watch?v=6wFsDKNSMQ4

I don’t believe there is a current way to get Digital Symphony to run on a modern version of RISC OS. However, QTM is available and is fully 32-bit compliant and has been updated by the original author to include some new features. Available → http://www.phlamethrower.co.uk/riscos/qtm.php

I know it’s heresy round these parts, but have you tried it under RO 3.71?

I know it’s heresy round these parts, but have you tried it under RO 3.71?

Can I wear my heretic badge with pride? I ask as the first thing that occurred to me was whether Nathan ran it under 3.7, 4.02, or 5.2x

It might be that the installer has a pretty splash screen that only runs on 26-bit mode.

If I were working through this I’d start with something early on the RO versions and the RPCEmu in non-SA emulation.

I don’t believe there is a current way to get Digital Symphony to run on a modern version of RISC OS

Never, say never. It might take a lot of reverse engineering and not actually be worth the effort except in pride terms but it is probably possible.
To quote a work colleague “given enough time, money and manpower almost anything is possible”

I was running under 3.7 and tried it under ARM610 but no difference.

@Alan Excellent news!. I’m always a fan of getting hold of the sources mainly so they don’t disappear on corrupt media. Any chance you get Gill to do a bit of digging and get the sources over to you/Jon (JASPP) or myself?

@nathan task complete. You can update your gannt chart :)
I asked Gil over on the YouTube comments section (link above).

I am almost 100% sure I had Digital Symphony working on 3.7. You had to switch off the cache before running it (assuming you’re using a StrongARM), and perhaps at a lower screen resolution as well. My memory is a bit hazy unfortunately.

@alan Top marks mate.

I don’t know how DigSym installed but the same HD in a different RPC refused to load as it detected it was the wrong machine. RPCEmu wouldn’t install it, presumably it does something weird with the installer floppy.

@alan Top marks mate.

I don’t know how DigSym installed but the same HD in a different RPC refused to load as it detected it was the wrong machine. RPCEmu wouldn’t install it, presumably it does something weird with the installer floppy.

Floppy Disc Copy Protection

I do recall that DSym had a floppy disc protection system where it stored your computers Machine ID number on the disc so it could only be used on that machine. The secret was to make a copy of the disc before using it. No idea if there are any unregistered DSym discs still around. I did receive a copy from someone when I fell foul of this myself, but have since lost my copy.

Clever stuff. Let’s hope Gil retrieves the sources. Then it’s just a case of contacting Oregan.

I knew it was a remote chance but I managed to image my Digital Symphony installation floppy. Unsurprisingly it crashes RPCEmu.

Can you send us a copy of the disc image? I can imagine it not working, crashing RPCEmu however is a bug that needs to be fixed.

I do recall that DSym had a floppy disc protection system where it stored your computers Machine ID number on the disc so it could only be used on that machine. The secret was to make a copy of the disc before using it.

Alternatively, if you know someone with a working copy on a machine with a known ID then search for that ID with a sector editing tool.

Normal state with those protected discs was to store 00’s at the location before use.
Other mechanisms would store the number of licenses – I think Beebug were fond of that one.

Alternatively, if you know someone with a working copy on a machine with a known ID then search for that ID with a sector editing tool.

Or just fake the machine ID on the RISC OS side. Just two words in Zero Page IIRC.

Interesting how nearly all companies using the machine ID ignored Acorn’s advice on what to do if there is a mismatch…

Any chance you get Gill to do a bit of digging and get the sources over to you/Jon (JASPP) or myself?

Source for the latest version of the player would certainly be handy as its the only mainstream tracker I’ve yet to 32bit.

I did attempt to build a new version from a decompilation, with mixed results and in the end TimPlayer added support for it, so I stopped work.

If memory serves me correct, it’s one of those players that had a dozen different versions, none of which are fully backward compatible and was frequently embedded into games. All the Paradise Games games do this and decrypt/insert it during loading, I know I had a real challenge getting some features to work under ADFFS, particularly one of the Paradise Games titles, which relied on a feature of Digital Symphony which was poorly coded and frequently crashed due to reentrancy.

Wizards Apprentice might also use it, which took the best part of a year to get working due to its channel handler being so heavily optimised for ARM2. On ARM3+ it thrashes the cache due to its large unrolled loops and reliance on self-modifying code, not to mention corruption from reentrancy.

DigitalSymphony might also be the tracker that flushes the cache to make itself StrongARM compatible. It’s been a few years since I last touched it, so memory is a bit hazy!

Long story short, I would like to 32bit it, bundle with ADFFS and replace the embedded players in games that use it.

@Jon, I’m relieved to hear you’re so positive about 32-bitting it, as I realised I had told Gil that JASPP would want the sources – without actually consulting with you in advance.

BTW, thanks for all the great work you’ve done on preserving so many Acorn 32-bit games and utilities. It really is fantastic.

BTW, thanks for all the great work you’ve done on preserving so many Acorn 32-bit games and utilities . It really is fantastic.

Fixed that. :)

I’d like to just say that, as a total non-games person¹, I think it’s wonderful that there is so much activity and it seems to be building all the time.

Also, of note is that, unless Dave is zapping most before I see them, the incidence of new accounts for genuine people new to the platform seems to outweigh the spam accounts

¹ So “non-games” that back in my youth I built a TV-games unit and then handed it over to my sisters to play with.
The circuit build and the programming work (including using an EPROM programmer at the local college my father worked at) was the interesting bit for me.

in my youth I built a TV-games unit

In my late 20s and early 30s I did several stints working on contract for MicroVitec, mostly designing PCBs for monitors (common in schools, attached to BBC Model Bs) and arcade games machines…

(Steve designed the circuits and got them working on breadboards, I turned them into manufacturable products.)

I’d like to just say that, as a total non-games person

Me too, and I’ll +1 what Steve said.

When I was younger, I found hacking games to be more interesting than the games themselves; I suck at games, so usually got bored of them pretty quickly, but dealing with “protection” was a lot more interesting.

Special kudos to whoever it was that loaded a module that hijacked the file ops to unscramble data loaded off of floppy on the fly.
Extra special kudos to whoever it was that presented a regular 800K floppy disc, with just an !RunImage visible, and all of the game data residing on the second half of the disc that was actually formatted differently (so disc copying/verifying would get so far and then give up with disc errors).
And a big ya-boo-sucks to that game whose “protection” (in scare quotes) was sticking hard spaces (Alt-160) into the filenames. At the time, I was using a custom VDU font for better clarity, and it put a little dot into hard spaces. So, yeah, the game was blown in however many milliseconds it took my eyes to register that ¹.

I think it’s wonderful that there is so much activity

It certainly is.

Which reminds me… haven’t heard from AmCog for a while. Is something brewing? ;-)

¹ Cue you Steve (and not any other Steve) pointing out that the vision-brain-thought interface is actually pathetically slow, which is why people get paid megabucks to drive cars like nutters, and those who try driving like that in normal cars usually end up in the hospital or the morgue. It may take around 150-200ms for visual stimulus to reach the brain, but then it has to be thought about; meaning a typical fast reaction time is in the order of 300ms ². Or put more simply, a third of a second. You can double or triple that if something needs to happen, the final speed depending on what – it’s a lot easier to press a button or wobble a joystick than it is to slam on brakes.

² At school we had a test where we had to catch a ruler being dropped by somebody else. It could happen at any moment.
I had an idea that instantly made everybody’s results much worse. I said that the teacher should construct a little box with a hole in the bottom that the ruler falls out of.
Turns out that due to reaction time, one could greatly improve their catch time if they stopped watching the ruler and started watching the hand, to notice when it begins to move before the ruler has started to drop. Remove that visual stimulous and you need to depend entirely upon the movement of the ruler. The dropped-on-floor rates ³ increased dramatically with the hand hidden.

³ Including me…

in my youth I built a TV-games unit

Youth in that being early/mid-teens.
Hardened character by then – I was 8 when an Antex soldering iron shorted out through me to the metal sink top.

I was 8 when an Antex soldering iron shorted out through me to the metal sink top.

8~)

Perhaps a little more precocious than me then…

http://clive.semmens.org.uk/Recounts/ThreeLessons.html

Sorry to derail the thread a little – addressing Clive here.
Do you have any remaining artefacts of your time at Microvitec?

I’ve managed to hunt down and scan one of the service manuals for the early Cub monitors – the 653 service manual is out there (in HTML form) – but I haven’t seen the Cub 3000 supplement in any form.

I’m curious what was done to it as the A3000’s video output range is 0.7V p-p into 75R while the Cub wants 4V in analog mode. Mostly I’m wondering if it’s possible to add that onto a 653.

Rick — if memory serves, it was 4th Dimension who had the “Protection” module which overrode the file vectors. From memory the workaround was to write a program to go through every file on the floppy, loading the module, loading the file, killing the module, saving the file to RAMFS, then going onto the next file. Something like that anyway.

Do you have any remaining artefacts of your time at Microvitec?

Sorry, probably nothing useful. I’ve got a couple of photocopies of PCB artworks – copper, resist and ident sheets – and I might have a circuit diagram, I’m not sure. It’s all at our old house (our son’s house now) not where I am. If things like that are any use to you, I could get him to dig them out and send me photos.

Sorry, probably nothing useful. I’ve got a couple of photocopies of PCB artworks – copper, resist and ident sheets – and I might have a circuit diagram, I’m not sure. It’s all at our old house (our son’s house now) not where I am. If things like that are any use to you, I could get him to dig them out and send me photos.

That would be very useful in fact — the PCB images in some of the service manuals are quite badly reproduced.

If you don’t have the facilities to scan them and can bear to let them out of your site — feel free to drop me an email (philpem@gmail.com). I’d be happy to scan them and make them available.

Cue you Steve (and not any other Steve) pointing out that the vision-brain-thought interface is actually pathetically slow

That brings back memories of an item in one of the books on my shelves¹. As memory serves the title was Pictorial Knowledge and it showed a giant laid on the earth and plotted relative times to send a message. The “speed of thought” is actually best described as dog slow when compared with wireless or electrical signals.

¹ When I was little (primary schools) the insulation on most of one wall in my room was essentially books. The collection, text and fiction grew.
When I got my own house the insulation on two walls of the back bedroom and part of the living room was books.
In this house most of them are in bedroom 4, except the ones in bedroom 1, bedroom 2, bedroom 3, the bookcases on the landing and the bookcase in the living room. Plus the ones that are loose, but I have a plan, and some shelf brackets.
I think I may need to build that two storey extension…

If you don’t have the facilities to scan them

I do here, but I don’t think my son does – but he does have a very good camera. I’ve also got an actual unpopulated PCB there – and possibly a scan of it here. I’ll see if I can dig it out.