Scope on ARM (Apple M1)

[robinette]

A decade later, after Scope abandoned the development of the platform in OsX, it seems to me a great opportunity to enter the market through the big door if development was thought to run on Apple ARM.
I don't know what you think about it.
I know SonicCore is a small company, but there may be a turning point if you think about broadening your vision. It depends on it whether it continues to be a smaller and smaller company (to the point of disappearing) or, on the contrary, whether it finds a market niche like the one it had before OsX.
These are times when large studios are scarce and small professional home studios proliferate. SonicCore has a chance to meet the public again.
Also, it's not just about ARM with Apple macOS, it's also about ARM with Windows. Everything points that way. There is less and less and we should think about doing homework.

A greeting.

[dante]

How would Scope on Arm require any less R&D investment / effort than Scope on OsX ? And S|C is already as small as it can get without disappearing... so cant see either of those outcomes happening in the next 5 years.

[nebelfuerst]

I think it was no good decision of apple to choose M1 with respect to the "full option PC" market.
Many friends were pissed when apple left PPC and chose Intel. The only advantage was to run dualboot with windows and you could build a hackintosh.
Changing the CPU family once more disqualifies apple to be used for projects with a long roadmap.

[Bud Weiser]

May be many didn´t recognize ...
But,- the Apple M1 thing is a return from CISC to RISC processor.

Looking back,- the Motorola procs Apple used for their PPC computers were RISC.
That were the days I liked Apple and Mac OS 9.x.

Well, Apple obviousely had not much options than Intel when leaving Motorola in the need of higher CPU speed in the past.

I don´t doubt this new switch to a different processor will make some trouble for a limited time,- but I also believe Apple can use a lot of code running previousely on RISC procs and improve for ARM (RISC).

I can imagine we might see applications like (Emagic) Sounddiver again,- maybe new look and optimized/adapted for today´s MIDI hardware brands but also supporting the old gear´s sysex.
Who knows ?

And, don´t forget old SCOPE (up to v4.5) for Mac and up to OS 9.2.2 was also coded for RISC.
This code base is still existing.

Somwhere here, I´ve read S|C will come up w/ a new application running on Mac in about 1 year ...
I´ve also read it won´t be SCOPE ...
But I believe it happens because Apple switches back to RISC / M1.



Bud

[dawman]

I’d love the ESX Ssmpler on the PC.

[nebelfuerst]

RISC or CISC doesn't mean the instructionset is compatible. Intel is also RISC, as it decodes the instructions into microcode-operations, which are run through RISC-pipelines. RISC is the result of bad compilers, which don't use the full instructionset of a CISC CPU and was invented by Cray. Most compilers are still very bad, so RISC is still the way to go. Putting RISC instructions directly into memory ( like PPC did) requires more memory bandwidth and larger caches.
Apple uses CPU-emulation to run Code of former CPUs. This is always a challenge for compatiblility and timing and often not for the driver level. For me, Apple just doen't focus in PCtype computers any more but tablets, phones.
Apple bought some companies in the audio business and dumped their products without selling a replacement. Emagic is just one company to name. I'm no M$ fanboy, but apple is no alternative any more, today.

[valis]

Agreed on the micro-op side of things. The reason for this is this is how the pipeline actually works, when instructions are fed in and 'decompiled' the scheduler can make appropriate choices about what resources get used. OS Awareness of resources <> EFI mapped awareness <> PCU/IME internal CPU awareness (and resulting pipeline schedule) all work together to 'balance' a load. This is the current state of affairs, and so the load of determining how to best represent a given high-level language block of code gets arrayed across an ecosystem that has basically evolved beyond RISC vs CISC. Intel & AMD have ported tech they either absorbed or developed under other products (Itanium) into the x86 architecture to the point where the lines aren't just blurred, they don't really exist now imo.

As for Apple, I'm not really a fanboy and even think that the linux market has a lot ot offer now (Bitwig & Renoise would be my preference over Harrison & etc but that's more about my workflow). However...

On the video side, Final Cut Pro X was a bit 'kiddie' to pro users when it debuted, but the TV market largely dried up while their 'export to platfom Y' approach has a pretty broad userbase now. Sure, Adobe Premiere stuck with the 'tried and true' workflow from Avid & Final Cut Pro 7 & before, and a lot of traditional outlets made the transition. Who was right?

Logic Pro X works great on a Mac, well enough that I'm still using aging Mac Pros that run it. That iPad that sits here, can do just about all my 10 year old laptop can do and then some, when it comes to the compute power of a session and routing audio & video around. iRig Pro Duo takes care of interfacing elsewhere, as can the iConnect interfaces.

Scope Machine <> Win10 Xeon rig running VCV & visual software <> WIn10 i7 rig running Acid Pro, Bidule & Visual software <> Mac Pro running Logic Pro X & Live 9 <> Macbook Pro running Live 10 (and Logic runs fine there)

What isn't obvious there is that those machines are from 2001, 2009, 2016, 2009, 2012 respectively (the iPad is current 10.2 model). I don't really see what would change with M1 aside from waiting for 3rd party plugins to be ported. The only thing lost is the Hackintosh ability, and while that's truly a loss (no joke) for many it's not really a show stopper.

[dante]

I’d love the ESX Ssmpler on the PC.

U mean EXS24 ? Whats so good about it ?

RISC or CISC doesn't mean the instructionset is compatible.

I thought RISC - meaning Reduced Instruction Set - just meant less instructions - so still can have processors with different instruction sets - all of them being 'Reduced'.

[dawman]

EXS24.
Been around forever, reminds me of old hardware, and it’s a cool name.

[dante]

A cool name - maybe that’s why I bought Zebra 2

[valis]

I’d love the ESX Ssmpler on the PC.

U mean EXS24 ? Whats so good about it ?

RISC or CISC doesn't mean the instructionset is compatible.

I thought RISC - meaning Reduced Instruction Set - just meant less instructions - so still can have processors with different instruction sets - all of them being 'Reduced'.

Thought maybe he meant ESX-1. EXS-24 finally got a facelift in the most recent Logic updates btw, along with an ableton live like UI mode for triggering loops.

As for RISC, it does mean Reduced Instruction Set Computing and Cisc means Complex Instruction Set Computing. That's easy to look up on wikipedia of course, but the overarching narrative of a battle in the market would require a bit of reading to follow. This was between 'intelligent compilers' and higher CPU speeds on the RISC side (due to the simplicity of design) as well as higher yields, versus players who originally lauded the 'everyman' angle (Apple/Commodore/etc) and the 'our growth will come via the everyhman model (Intel). Ie, make things as easy as possible for users and devs.

Back when we were at the 200nm gate limit and before that, CISC was seen to 'waste' a lot of silicon on instructions that many never used, or may only be used occasionally. These same instructions could be emulated by the compiler, by creating machine code that replicated the same output from the simpler RISC architecture, but this meant that not only compiler times went up dramatically, but also that compilers had to be very well written and the bulk of the engineering work was focused on the software side: compilers, software stacks and software devs stayed 'experts' in white lab coats and executives on this side espoused things like "there is no reason anyone needs a computer in their home, ever!" I'm paraphrasing here a bit, but it was clear the RISC players wanted to keep computing among hte esoteric arts.

The same "Complex silicon" argument was put forth originally for 'math' functions, which were left out of earlier Intel CPU's meant for the masses. This meant a lot of people eschewed X86 for patforms where math functions were integral--like Apple which also had a very advanced math API available and ready to go for scientific & engineering purposes in that era. On the intel side we're talking about the x87 FPU which was originally an addon card with a limited number of API functions that grew through the 286 & 386 era (when it was a second socketed CPU), until the FPU was finally integrated in the 386/486 era (SX & DX if you recall). This sped up mandelbrots quite a bit when I was a kid, but didn't impact Chuck Yeager Flight Simulator, Wordstar 2000 or QEMM desktop manager one whit.

Anyway the basic idea is that logic gates are capable of a few low level functions (pointing & adding being the core functionalities) and therefore all higher functionality can essentially be coded in software rather than via adding additional logic to the silicon map (hardware). The end result was higher core speeds, and simpler pipelines that could be easily optimized on the compiler side. The latter part was somewhat a failed promise, as the continued speed in the pace of computing development eventually meant that most of these players wound up fighting amongst themselves for domination (OSF-1) and their tools remained tethered to the markets they had already created (microcomputing--for thin clients deployed in a building or a section of a building, and mainframes etc).

nebelfuerst is 100% correct when he says modern designs are a hybrid of sorts, but that was explained above as well.

[dante]

Thanks Valis thats one of the best explanations Ive seen so far & w/- historical context.

[robinette]

How would Scope on Arm require any less R&D investment / effort than Scope on OsX ?

Who has said such a thing? Creamware (and SonicCore) did not develop for OsX and ... from those waters, these sludge (you want to see it or not).
The idea of resuming the platform for Apple's ARMs is none other than resuming a market that for now they have lost and that is inevitably leading to their disappearance.

Open Scope was a great idea in this regard. Opening the platform to potential developers would have meant multiplying hands to make this happen.

I know it is not easy at all, but there are companies as small as SonicCore developing great software in much less time than it took SonicCore to "change the color of the Routing Window and call it Scope 7".

It was simply a reflection in the form of an idea. I know very well that nothing is going to be done and that the platform will end up dying of sheer neglect.

A greeting.

[robinette]

May be many didn´t recognize ...
But,- the Apple M1 thing is a return from CISC to RISC processor.

Looking back,- the Motorola procs Apple used for their PPC computers were RISC.
That were the days I liked Apple and Mac OS 9.x.

Well, Apple obviousely had not much options than Intel when leaving Motorola in the need of higher CPU speed in the past.

I don´t doubt this new switch to a different processor will make some trouble for a limited time,- but I also believe Apple can use a lot of code running previousely on RISC procs and improve for ARM (RISC).

I can imagine we might see applications like (Emagic) Sounddiver again,- maybe new look and optimized/adapted for today´s MIDI hardware brands but also supporting the old gear´s sysex.
Who knows ?

And, don´t forget old SCOPE (up to v4.5) for Mac and up to OS 9.2.2 was also coded for RISC.
This code base is still existing.

Somwhere here, I´ve read S|C will come up w/ a new application running on Mac in about 1 year ...
I´ve also read it won´t be SCOPE ...
But I believe it happens because Apple switches back to RISC / M1.



Bud

More or less that's where the shots go (that was the idea that hovered in my mind).

A greeting.

[dante]

The idea of resuming the platform for Apple's ARMs is none other than resuming a market that for now they have lost and that is inevitably leading to their disappearance.

Is the current Apple ARM the same architecture / codeset that Scope for Mac used to run on all those years ago ? Its the 'resuming' bit I don't get.

[robinette]

Is the current Apple ARM the same architecture / codeset that Scope for Mac used to run on all those years ago ?

But let's see ... where am I saying that? Please try to understand what I have written and do not assume what I have not said.
What I am saying is that it is a good time for SonicCore to resume supporting MacOS, since the architecture is new and everything that is developed for it now will serve as a foundation for a few years (maybe decades).
How hard it is to understand?.

[valis]

No need to be spurious, I didn't catch that meaning from your posts (meaning separated from the previous PowerPC ISA) either. At least not in the way you just stated it, which seems clearer.

The only real upshot I can see for Sonic|core to port Scope to M1 would be that there's a lot less QA to do than on the PC side. The downside? There's a lot of work to be done regardless

[dante]

Is the current Apple ARM the same architecture / codeset that Scope for Mac used to run on all those years ago ?

But let's see ... where am I saying that? Please try to understand what I have written and do not assume what I have not said.
What I am saying is that it is a good time for SonicCore to resume supporting MacOS, since the architecture is new and everything that is developed for it now will serve as a foundation for a few years (maybe decades).
How hard it is to understand?.

What’s hard to understand is this : Given Scope for Mac OSX was non viable (from an S|C standpoint) what is it about Mac ARM that makes it (more) viable ?

[robinette]

What’s hard to understand is this : Given Scope for Mac OSX was non viable (from an S|C standpoint) what is it about Mac ARM that makes it (more) viable ?

Viable is everything that makes you grow. As a company, if you don't invest in growing, you just have to wait to die.
Apple switched to OsX at the time that Creamware was sinking (bad decisions), and from Creamware it was decided to jump ship before grabbing the bull by the horns.
SonicCore has tried to build the company with only Xite. They invested everything in Xite-1 whose market launch has been lame. Failure to support OsX was to lose more than half of potential users (more than half of professionals and amateurs in the audio world work with OsX).
If we are not able to understand this, and we are not able to see that the Scope platform would have a decent place in the audio world if it had support in MacOS, we are totally blind.
Apple just made the biggest move in recent history. Much more important than the move to OsX with Intel. Now you have full control of your software and hardware to take your platform beyond what you have been able to achieve in the past. He has done very well (the data is there) and it is only the beginning. They continue to work on improving their CPUs and further optimizing their software. It's a winning bet, and if a development company with SonicCore is moderately smart, it would study the feasibility of getting on that winning train.
Don't you see it viable? Is it more viable to continue hitting the blind as they have been doing up to now? Is "Changing the color to the Routin Window and calling it Scope 7" more viable? Would you like the Scope platform to continue growing or do you prefer to see it die dying?
Anyway...

[garyb]

producing another version of Scope at this time would mean no more SonicCore. fertilizer makes a plant grow, but too much fertilizer will kill it.
who will produce it?
who will pay?

i assure you that SC would LOVE to jump on this.

argue and debate all you like.

what is lame is that the music production industry has become about computers ONLY, buying them, updating them, new pictures of old things, and the loss of real products and skills for automated machines that make cookie-cutter products that are mostly indistinguishable from everything else. the lemming-like push for new products that do not make better music, but have interesting specs is truly lame. the industry is lame.

again, SC would LOVE to jump on this. SC would LOVE to have made an OSX version.

who will do this feasibility study? some "team"? who pays this "team"? is there a "marketing staff"? FUCK CORPORATE OVERLORDS!(the last part is just my opinion). development company! how rich!

Would you like the Scope platform to continue growing or do you prefer to see it die dying?

which person here has any say in this, (besides spending their money)?