68060 based QL system?

Nagging hardware related question? Post here!

Do you want a 68060 based QL system?

No, better improve speed of FPGA based 68K
12
43%
Original Q60 with video updates
5
18%
New 68060 system design
11
39%
 
Total votes: 28

User avatar
Dave
SandySuperQDave
Posts: 2765
Joined: Sat Jan 22, 2011 6:52 am
Location: Austin, TX
Contact:

Re: 68060 based QL system?

Post by Dave »

Peter wrote:Actually I am quite uncertain, because fact remains that emulation will be faster, no matter how optimized the design.
There's a lot of places where emulation can't really go. Embedded functions like intelligent thermostats or car electronics suit an FPGA better than a QL emulated on a PC - which would be much harder to embed transparently and cost effectively. Emulation fails from the user experience POV and the hardware implementation POV.

Fact is, we're pretty much dependent on your Q68 architecture to go a lot of places. A lot of places that are current world relevant. There are applications that suit the Q68 ICs (but not PCB) that open up a LOT of doors.

But yeah, that's not gonna happen. You won't allow it.

It's hard to get excited about future hardware when we can't even explore all the possibilities with the current hardware. How can you know which way to develop it if most of the options or possibilities have already been ruled out?


User avatar
tofro
Font of All Knowledge
Posts: 2685
Joined: Sun Feb 13, 2011 10:53 pm
Location: SW Germany

Re: 68060 based QL system?

Post by tofro »

Dave,

have you touched an embedded system the last 10 years? Agreed the Q68 would be nice for embedded applications - For QL tinkerers.

But there's wonders to find out there, for about $5 each, that had 30 years of time to develop, with free development environments we can only dream of in the QL world. I'm afraid a 68000-based system wouldn't stand the slightest chance against these systems.

Tobias


ʎɐqǝ ɯoɹɟ ǝq oʇ ƃuᴉoƃ ʇou sᴉ pɹɐoqʎǝʞ ʇxǝu ʎɯ 'ɹɐǝp ɥO
User avatar
Dave
SandySuperQDave
Posts: 2765
Joined: Sat Jan 22, 2011 6:52 am
Location: Austin, TX
Contact:

Re: 68060 based QL system?

Post by Dave »

tofro wrote:Dave,

have you touched an embedded system the last 10 years? Agreed the Q68 would be nice for embedded applications - For QL tinkerers.

But there's wonders to find out there, for about $5 each, that had 30 years of time to develop, with free development environments we can only dream of in the QL world. I'm afraid a 68000-based system wouldn't stand the slightest chance against these systems.

Tobias
A small part of my embedded work is with Pi Zero, Pi3B etc. I know their capabilities. I have a Pi3 webserver. I also know if you mount a PiZero in a car in Texas, with the CPU not being auto rated the thermals are impossible. There's automotive rated FPGAs and RAM of the type used in the Q68. For flexible data display and logging purposes, SMSQ on reasonable yet simple hardware beats Pi/Linux or arduino hands down. It definitely beats 68K assembly on no OS, which is what is being used now.

That's without even mentioning vibration.

If anything, those more powerful systems also cause too much overhead: long booting times for Linux (vehicle could move several hundred meters before a Pi even boots let alone launches software, checks timestamp of last logfile, etc etc...) and more complex coding. The Q68 is just right for this job, would give simpler, more lean hardware that would fit the task better, is quick enough, a better development environment, etc.

If it dropped ethernet and added a 2nd video output of the 2nd screen area, for 2 LCD support, that would be a small customization making it "more" suited. The extension bus would support a few GPIOs via a CPLD for controlling LED brightness, light level sensing, a few input buttons.... Provision for an onboard EEPROM for storing secured logged data. It cannot be stored on removable or undetectably alterable storage, by federal regulation. On return to base, logged data is transferred to a central system by wifi for long term storage.

The extended purpose of the next version of this product, currently designed around a 68HC000 and a CPLD for IO, is too "fiddly" (their words) and they're open to a bespoke and novel solution if it simplifies their development process. Their big stumbling block is not having an OS that's close to instant on. It's a black box and parameter display type product and has to be logging within 5 seconds of power-good. With a little coding, I think the Q68 system could get from power-good to running a loaded image in under 2s. The two developers at this company are both in their 50s and cut their teeth on Commodore and Tandy using BASIC. SBASIC would be a dream to them.

That's just one real world example.

I would say more, but NDA...

If you think modern products couldn't be built around 68ks, my boat's chartplotter has multiple CPUs including a 68020 which appears to load/retire tiles of map into a cache memory... The GPS, route plotting etc and video generation are done by an older ARM 250. It was built in 2009-2017.


User avatar
Pr0f
QL Wafer Drive
Posts: 1298
Joined: Thu Oct 12, 2017 9:54 am

Re: 68060 based QL system?

Post by Pr0f »

Windows (CE or NT embedded) in ROM - boots very quickly :P

But actually a good point - RTOS type applications are still out there - just look at all the coffee machines, and goodness only knows what's in a lot of 'smart' white goods these days.

Your set top boxes have mostly gone to Linux or a derivative.


User avatar
vezhlys
ROM Dongle
Posts: 27
Joined: Sun Sep 03, 2017 8:45 am

Re: 68060 based QL system?

Post by vezhlys »

Dave, I am a bit confused that you are considering Q68 or even QL software for some serious "commercial" projects when the world moved on so far these days both in software and hardware (I believe it should be more than enough options for embedded development, even non linux ones, something like contiki os or so). Isn't Peter's projects are more a hobby ones with the goal to be QL compatible and provide some "modern" hardware options for QL tinkerers, plus they have extremely limited manufacturing volume. What you are saying is partially true, but it sounds just like fantasies what would have been if everything would have been different... On another hand if you really see it feasible to use SMSQ/E to achieve such goals and there is kind of weird requirement to go with obsolete 68k and you are serious about it (and there are real business ideas behind that), it sounds like quite a separate design project which needs the same "commercial" resources business should be willing to provide. However, currently it sounds more like either deliberate provocation or just blind fantasies. What is the purpose of even discussing that in a serious manner? Don't want to add negativity but I really don't understand your goals.


User avatar
Peter
QL Wafer Drive
Posts: 1953
Joined: Sat Jan 22, 2011 8:47 am

Re: 68060 based QL system?

Post by Peter »

vezhlys wrote:Isn't Peter's projects are more a hobby ones with the goal to be QL compatible and provide some "modern" hardware options for QL tinkerers, plus they have extremely limited manufacturing volume.
Yes, QL-SD and Q68 are strictly a hobby projects for me. I provided all work and time totally for free, not even recovering any expenses.


User avatar
Dave
SandySuperQDave
Posts: 2765
Joined: Sat Jan 22, 2011 6:52 am
Location: Austin, TX
Contact:

Re: 68060 based QL system?

Post by Dave »

Fantasies?

Ok.

Now I'm pissed off.

In a black box situation where a controller simply does some simple action like monitor inputs, display data, log that data securely then at a defined end point upload that data to a modern archival system, the working of the innards are never exposed. They could be anything. It is not essential to a washing machine, dishwasher etc. what the process is. My washing machine, a Maytag, uses a 1990s atmega controller. My tumble dryer uses a 1930s style mechanical controller. There is no advantage in using anything more capable.

To say that using SMSQ or Q68-level hardware in this type of application but for a higher level task (storing a dozen channels of data in 8-bit resolution once per second, displaying the data in green if within certain limits or in red if outside and making an annunciation if anything is red) and operating either four or six stepper motors to show gauges.... Currently done by an MC912DG128ACPVE which costs $45 per piece.


Q68-style hardware is almost ideally suited to this application above many others both too slim and too advanced. It's a static product that is embedded and does not receive updates.

Your lack of imagination of the doors Q68-style implementations open up for us does not mean I am living in a fantasy world. I could apply this to this problem more quickly and simply than the current family of controllers and a CPLD can do it, and with most of the development path being moved to SBASIC. Which is what my clients want: a hardware platform that is lightweight, quick booting, requires no licensing fees/costs to use the OS, with a couple of programmatically advanced featured.

It could be SMSQ on Q68 or RISC OS on ARM (their preferred route) or etc, and you could recognize that in this environment, Q68/SMSQ is not ill-suited.

A full 68060 would be.

And that was the point of my line of reasoning - to show that 68060 developments aren't going anywhere new, but Q68 development opens up a lot of new possibilities, and a lot of work opportunities for QLers.

It isn't fantasy. I'm getting paid. :P


User avatar
vezhlys
ROM Dongle
Posts: 27
Joined: Sun Sep 03, 2017 8:45 am

Re: 68060 based QL system?

Post by vezhlys »

Dave wrote:Fantasies?

Ok.

Now I'm pissed off.
Sorry, I didn't mean to piss you off.
Dave wrote: It could be SMSQ on Q68 or RISC OS on ARM (their preferred route) or etc, and you could recognize that in this environment, Q68/SMSQ is not ill-suited.

A full 68060 would be.

And that was the point of my line of reasoning - to show that 68060 developments aren't going anywhere new, but Q68 development opens up a lot of new possibilities, and a lot of work opportunities for QLers.

It isn't fantasy. I'm getting paid. :P
Yes, I agree with that such designs have potential (Raspberry Pi or andruino popularity showed that). You would like to have this design freely available for reuse? My point was that the original purpose of these boards wasn't to be a commercial product for embedded purposes but rather QL compatible board and, even though you are right about potential possibilities, somehow it is not a primary goal and not a desire of QL user (OK, it is just my humble opinion). Are your clients ready to pay for such side developments (new PCBs, drivers, accommodating SMSQe, etc)? This is primary reason why it sounds a bit unrealistic to me. Those improvements would have been targeted to QL users first who just want to use real hardware for personal reasons mostly (like ability to connect to modern monitors). Let's say 68060 based system is more interesting to me personally because it provides opportunity to run modern system like Linux as well or maybe even port some BSD to it. Then it can be used both for QL purposes and some other modern activities where I don't need to much CPU power (plus some geekiness feeling to run it on non conventional hardware). But again it would have been more hobby activities from my side than business what you are targeting for.


User avatar
Peter
QL Wafer Drive
Posts: 1953
Joined: Sat Jan 22, 2011 8:47 am

Re: 68060 based QL system?

Post by Peter »

Dave wrote:Your lack of imagination of the doors Q68-style implementations open up for us does not mean I am living in a fantasy world.
If you wish to discuss this, why not start a new topic like "New Q68-style implementations" or "New applications for the Q68 architecture"?


User avatar
Dave
SandySuperQDave
Posts: 2765
Joined: Sat Jan 22, 2011 6:52 am
Location: Austin, TX
Contact:

Re: 68060 based QL system?

Post by Dave »

Because you opened the door. If it doesn't interest you that this door is opened, then it doesn't matter what threads we start - unless someone reproduces all your many years of work (which isn't going to happen) then the matter is moot.

You asked an interesting question. I have conflicted thoughts about it. Two of the answers benefit the community but in different ways.

Nobody would expect you to support these extended, out of band uses of your hardware. However, people aren't going to use the hardware because it's closed, only available as complete units with mark-ups that mean only 5% or 10% of the market is even being tapped.... It's an interesting curio. If the programmed Q68 FPGAs were made available for, say, €25, then your IP would be protected AND the tech could be put to other uses.


Post Reply