Seems to me that 90% of anything anyone tries fails. People start businesses, big companies develop products, all for nought.
Placing a bet on a horse that is not even born yet is rather a risky proposition. If you see what I mean.
As for the "community" thing. I have a GG board and it's great. I just wish there was a Parallax board that many others felt compelled to clone. Like the Arduino is an Arduino no matter who makes it. This sounds like "Oh no they are stealing our design and profits" but I believe such networking effects establish a standard and grow the market for everybody.
I've pretty much ignored the P2 until recently. There's no business plan even to be contemplated around a non-existent product, and I didn't want to waste any time grokking a moving target. Besides, there's still plenty of novelty to exploit in the P1.
I've only awakened to P2 issues lately because its lengthy and expensive dev cycle seems to have become a drag on other aspects of Parallax's business. And because Parallax is my biggest customer, that was cause for some concern. However, because the P1 still offers a lot of promise for as-yet untapped apps, I'm not ready to diversify away to other platforms. I mean, there isn't a micro I'd rather design around and program than the Propeller, even if the "A" and "RP" platforms seem to be running away with the market.
Fortunately, the house is paid for, and the '82 Benz still runs. Life is good!
As far as I understand the P2 cannot move forward until the next shuttle run which is scheduled for April or so.
So all the changes that have been going on recently do not impinge on that critical path.
I'm not convinced this is entirely accurate. There is much to do -- testing, verification, documentation, software -- before a Propeller 2 can be marketed to the wider public, and also to ensure that the next run is more successful than the last.
In any case, it may not be for the best to simply view this as a recess of sorts.
I've only awakened to P2 issues lately because its lengthy and expensive dev cycle seems to have become a drag on other aspects of Parallax's business.... the P1 still offers a lot of promise for as-yet untapped apps,... I mean, there isn't a micro I'd rather design around and program than the Propeller, even if the "A" and "RP" platforms seem to be running away with the market.
I put a bunch of work into porting PropGCC to P2 well over a year ago and have worked on it off-and-on since. However, I've more or less decided now to wait until I'm fairly confident that the instruction set has settled down before I go back and redo the work I had already completed before all of the recent changes. It's been fun being in on the process though. I'm not complaining about having spent time on stuff that has since changed. It's all part of being in early on the hardware design process .
An increase in Quickstart price, has forced us to quit selling the Pocket Mini Computer add-on product, but there is still enough interest in the project to build a single board alternative for this year.
... Not that long ago we had a wonderful community based Propeller board which everyone loved. You took a product that was slated to be an evaluation board and make it direct competition with your own community.
Price and products are a double edge sword, Parallax have to compete with other vendors $12 Eval Boards, so entry price is very important.
That said, a price increase on a product you have planned around, is never a good thing, especially given others are lowering their prices.
If Quickstart has moved significantly in price, there does seem to be an opening for a minimal entry system, and smallest PCB size usually helps lower prices.
To me, the CP2105 is a smarter choice than FT23xx, as it can power the P1, and give two separate PGM and Communicate pathways.
Only so much rain can fall in one's life before they finally get smart and find a dry place.
A lot of effort went into the P8x32a - the design is still great, it's just too small and costly. Oh to have a simple upgrade to the B version with more memory and pins ... 3 years ago.
In a way Parallax warned that the Propeller-Platform would not be entirely viable by not initially producing boards from Jon's work (JonnyMac Spin-Zone in N&V May 2009).
Personally, I like the Propeller-Platform much better than Quickstart which is just one black eye after another that still pays boxing dividends - the Rev.B board is a much better design, but certain decisions on the Rev.A still cause me grief beyond the death of the Propeller-Platform.
Just a few assembly changes make the Propeller-Platform better (stacking headers for example). In my opinion, the board needs a revival. Martin's DNA board is really nice. Name another Propeller based board that has so many plug-ins besides the ASC+.
A P2 platform board is an interesting idea. Chip's vertical back-plane P2 board will be useful, but a flat platform board will find more utility IMHO. Of course, we need some chips first.
Only so much rain can fall in one's life before they finally get smart and find a dry place.
A lot of effort went into the P8x32a - the design is still great, it's just too small and costly. Oh to have a simple upgrade to the B version with more memory and pins ... 3 years ago.
In a way Parallax warned that the Propeller-Platform would not be entirely viable by not initially producing boards from Jon's work (JonnyMac Spin-Zone in N&V May 2009).
Personally, I like the Propeller-Platform much better than Quickstart which is just one black eye after another that still pays boxing dividends - the Rev.B board is a much better design, but certain decisions on the Rev.A still cause me grief beyond the death of the Propeller-Platform.
Just a few assembly changes make the Propeller-Platform better (stacking headers for example). In my opinion, the board needs a revival. Martin's DNA board is really nice. Name another Propeller based board that has so many plug-ins besides the ASC+.
A P2 platform board is an interesting idea. Chip's vertical back-plane P2 board will be useful, but a flat platform board will find more utility IMHO. Of course, we need some chips first.
It would be kind of nice if there was a standard Propeller platform. I was encouraged when Parallax announced support for the Propeller Platform format a while back but that seems to have fallen by the wayside. As Steve said, the DNA board is very nice as is the ASC+. There is no shortage of candidates for a standard Propeller board but none have the critical mass to achieve that status. Seems like we're spread too thin sometimes. There are also lots of Arduino boards but as far as I know the shields work on all of them. That's a huge advantage.
There is still no propeller 2.
The Propeller 1 and its boards are simply too expensive compared to for example Raspberry Pi, which (1) costs $35 (2) I can buy it in my country without any problems.
Powerful FPGA boards like DE0-nano are available and cheap for educational purposes.
I am doing a PhD dissertation now. This is about signal processing - intelligent algorithms for old audio recording restoration. I planned to use P2 for it, but now it is impossible, as there is no P2. I switched to RPi+FPGA. My university paid $250 for DE2-115 I am using now - they have 50% discount for education. Then I learned I can make a lot of processors (nanoprocessors?) designed to do all I want in an FPGA chip using 1% of its resources. I don't need a Propeller board anymore for doing all this job. Then I discovered the Pi can be programmed in bare metal from scratch and then it can do all stuff hard to write in Verilog.
So, what I want from the Propeller? It's simple: (1) availability (2) availability (3) availability (4) availability... (n) availability
... and some lower prices and educational discounts.
And, of course, P2. And "a P2 Pi" - a cheap, wide available board with a P2, RAM and all connectivity like SD, USB, GPIO, VGA/HDMI etc.
A Pi and a Propeller are not comparable. They are totally different animals. All thought it does seem that Propeller boards are often expensive. As has been said here many times a Pi and a Prop complement each other and a Prop "plate" board for the Pi would be a great idea.
FPGA boardas are not really comparable either. Getting into Verilog/VHDL is a lot harder than writing C or Spin, the dev tools are huge and complicated the chips themselves are harder to work with if you want to build your own thing, The last time I looked GPGA boards were not cheap. However, yes, a PII should be able to tackle a lot of jobs people reach for an FPGA for.
"Nanoprocessors", for audio analysis, interesting, I would love to hear more about their architecture.
What's the idea with programming a Pi from scratch? I would have thought that throws away all the goodness of Linux and leaves you with a lot of work to do.
I'm all for a cheap cheap P2 board. However I'd like it to be a minimal as possible. Enough on board for the thing to operate and break out all the I/O pins. Even better if it is a Rasperry Pi format "plate". It should be usable stand alone of course. Some of those 2 million Pi users would love to look at the P2 that way, And that is a lot of potential customers.
It would be kind of nice if there was a standard Propeller platform. I was encouraged when Parallax announced support for the Propeller Platform format a while back but that seems to have fallen by the wayside. As Steve said, the DNA board is very nice as is the ASC+. There is no shortage of candidates for a standard Propeller board but none have the critical mass to achieve that status. Seems like we're spread too thin sometimes. There are also lots of Arduino boards but as far as I know the shields work on all of them. That's a huge advantage.
I agree with Phil that there is still a lot to do with the P1, and with David regarding boards. I have used the P1 for years, but have many more to go before even coming close to using most of its potential. I too would love to see a standard platform. To the list of great boards I would add Wulfdens (Brian Riley) PPTH (Propeller Platform Thru Hole). It comes in kit form so students can develop soldering skills while building their own, and they have a great sense of ownership when their blinky LED or "Hello World" program runs. Priceless!
A Pi and a Propeller are not comparable. They are totally different animals. All thought it does seem that Propeller boards are often expensive. As has been said here many times a Pi and a Prop complement each other and a Prop "plate" board for the Pi would be a great idea.
FPGA boardas are not really comparable either. Getting into Verilog/VHDL is a lot harder than writing C or Spin, the dev tools are huge and complicated the chips themselves are harder to work with if you want to build your own thing, The last time I looked GPGA boards were not cheap. However, yes, a PII should be able to tackle a lot of jobs people reach for an FPGA for.
"Nanoprocessors", for audio analysis, interesting, I would love to hear more about their architecture.
What's the idea with programming a Pi from scratch? I would have thought that throws away all the goodness of Linux and leaves you with a lot of work to do.
I'm all for a cheap cheap P2 board. However I'd like it to be a minimal as possible. Enough on board for the thing to operate and break out all the I/O pins. Even better if it is a Rasperry Pi format "plate". It should be usable stand alone of course. Some of those 2 million Pi users would love to look at the P2 that way, And that is a lot of potential customers.
While I agree that the RaspberryPi (mine is coming today!) and an FPGA are not directly comparable with the P1 or P2 when those chips are used to their strengths, there seem to be many people trying to turn the Propeller into a general purpose computing platform and the RaspberryPi anyway is far better at that. However, I also agree that the combination of a RaspberryPi and a P1/P2 would be wonderful!
However, I also agree that the combination of a RaspberryPi and a P1/P2 would be wonderful!
Parallax would also like to see a P1 Plate for the Pi. At present, our Product Plan for 2014 has our first Pi effort. Jeff Martin has been working to streamline the use of the Pi as a programming tool for Propellers, so it's a tool-oriented example rather than a piece of hardware. I think he's been in contact with Heater along the way, too. A video should be forthcoming within a week at the most. The next phase would be what you're all talking about - a hardware adapter providing access to the Propeller. However, we don't have the internal resources to do this design right now. If the community could make it happen we'd be supportive and possibly look at the manufacturing of the hardware.
Regarding OBC's comments, we were selling our QuickStart at less than our cost so it only made good business sense to us to try to cover our costs. I'd like to think that we were helpful in this regard for a long period of time.
Parallax would also like to see a P1 Plate for the Pi. At present, our Product Plan for 2014 has our first Pi effort. Jeff Martin has been working to streamline the use of the Pi as a programming tool for Propellers, so it's a tool-oriented example rather than a piece of hardware. I think he's been in contact with Heater along the way, too. A video should be forthcoming within a week at the most. The next phase would be what you're all talking about - a hardware adapter providing access to the Propeller. However, we don't have the internal resources to do this design right now. If the community could make it happen we'd be supportive and possibly look at the manufacturing of the hardware.
Regarding OBC's comments, we were selling our QuickStart at less than our cost so it only made good business sense to us to try to cover our costs. I'd like to think that we were helpful in this regard for a long period of time.
Ken Gracey
Hi Ken,
It's great to hear that you have plans in place to do a RaspberryPi plate for the P1! I'd love to get involved on the software side if you need any help there. Let me know!
...there seem to be many people trying to turn the Propeller into a general purpose computing platform...
To be fair, I think this point of view has always existed for the Propeller platform. After all, how many existing Propeller products are meant to be used in conjunction with a genera-purpose computing platform and how many Propeller products are meant to be used as a stand-alone platform?
As processors go, they either need to do a few functions really well, or a lot of functions generally well. And I think it's obvious which category the P2 is falling into. With all of the new features that have been recently added to P2, it's no wonder that people are seeing it as a general-purpose computing platform. In fact, I'd even go so far as to suggest that the current P2 design encourages general-purpose computing.
To be fair, I think this point of view has always existed for the Propeller platform. After all, how many existing Propeller products are meant to be used in conjunction with a genera-purpose computing platform and how many Propeller products are meant to be used as a stand-alone platform?
As processors go, they either need to do a few functions really well, or a lot of functions generally well. And I think it's obvious which category the P2 is falling into. With all of the new features that have been recently added to P2, it's no wonder that people are seeing it as a general-purpose computing platform. In fact, I'd even go so far as to suggest that the current P2 design encourages general-purpose computing.
With hub execution the P2 is certainly a better candidate for a general purpose machine than the P1 was. We'll have to see what people do with it. It should be interesting!
Rumour has it that at least one forum member has designed a Pi plate for production recently. I can't name names but he is well known and I hope he has been in touch with you.
I have not had any contact from Jeff. Be glad to help if I can. I do have a SimpleIDE for the Pi and a loader adapted from the propgcc loader that works
using the UART on the Pi's GPIO.
Seairth, David,
My idea of a general purpose computer now a days is anything grunty enough to run Linux, At least a command line version of Linux. That seems fair enough as that encompasses some very small and cheap machines today.
The Prop and Prop II are not in any way general purpose computers. They may be up to calling "general purpose" in 1970/80 standards but that is nothing today.
The Prop and Prop II are not in any way general purpose computers. They may be up to calling "general purpose" in 1970/80 standards but that is nothing today.
But surely P2 will be able to emulate a ZPU with a simulated MMU and boot Linux! :-)
What if you were to break the P2 into two products.
Propeller:
Basically the same as the current Propeller 2, except:
4 COGs (which would also mean faster hub access)
More memory (possibly even revisit dedicated DDR2 bus)
Comprehensive serial hardware
Emphasis on multi-tasking general computing (with convenient I/O)
Turbine:
16 COGs
Much less hub RAM (possibly none?)
No multitasking
No hub execution mode
Wider inter-COG data path
Comprehensive serial hardware
Instruction set is (nearly) a subset of Propeller instruction set
Emphasis on single-minded I/O processing
Why go with two products? Because they meet two different needs. Propeller can be squarely targeted as a general-purpose computing platform ("do it all with one chip"), while the Turbine can be squarely targeted as an augmentation for all of the low-power computing platforms (RPi, Arduino, etc) that need significantly more I/O (and processing power to handle the I/O). And, of course, there's no reason that the Propeller (1 or 2) couldn't make use of the Turbine as well.
I realize that this would be a big undertaking (and cost), but I think it would give the Propeller product line a much clearer road map.
*] Emphasis on multi-tasking general computing (with convenient I/O)
If you really want general purpose computing and modern operating systems like LInux you'll have to add an MMU and probably a multi-level cache. At that point you're better off adding an ARM core as a hub client with its own interface to memory. Why try to reinvent the wheel? There are proven solutions for general purpose computing and I think it's unlikely that Parallax will be able to improve on them. Better to let the COG do what it's best at and leave general purpose computing to general purpose computers. The only exception to that is emulators of vintage machines. Those can be fun to build and I think P2 will be a good platform for that kind of thing. Of course, that is unlikely to generate enough sales to make the P2 a success. Still, it's an interesting application and a challenging project for whoever decides to do the emulators.
Just to be sure that I didn't miscommunicate, I should clarify what we are doing and what we are not doing.
We have only one Pi project in the works right now. It's a video tutorial and software to make it easy to use the Pi to program the Propeller 1 with SimpleIDE - a portable development tool. The purpose of this effort was to get familiar with the hardware and formalize something our customers have been doing.
We don't have a Propeller Pi Plate hardware design in the works. This is primarily because we have limited resources to do the project. If the community drives the effort and we can facilitate and support it, we will. It's a substantial software and hardware effort to do it correctly. This piece of hardware would bring the greatest financial return, but requires more effort than we are able to allocate directly. Basically, putting our key engineers on it would mean we've got to forgo or delay efforts around SimpleIDE 1.0, FPGA hardware for P2, etc. I'd love to see the community create these hardware designs and move Parallax out of the realm of thinking that we need to create every board. Having the community do this would allow us to increase focus on the chips.
If you really want general purpose computing and modern operating systems like LInux you'll have to add an MMU and probably a multi-level cache. At that point you're better off adding an ARM core as a hub client with its own interface to memory. Why try to reinvent the wheel? There are proven solutions for general purpose computing and I think it's unlikely that Parallax will be able to improve on them. Better to let the COG do what it's best at and leave general purpose computing to general purpose computers. The only exception to that is emulators of vintage machines. Those can be fun to build and I think P2 will be a good platform for that kind of thing. Of course, that is unlikely to generate enough sales to make the P2 a success. Still, it's an interesting application and a challenging project for whoever decides to do the emulators.
If that's the case, then why not go the "Turbine" route altogether? Why bother with all of the current "generality" improvements that are being added to the P2? If it can't compete as a general-purpose platform, then why make it look like one?
(side note: I disagree with Heater's view of "general purpose must run linux". I think "general purpose" is more about whether it can be used as an all-in-one solution for a reasonably wide variety of tasks. Running an operating system is not, in my opinion, necessary to still be considered "general purpose".)
If that's the case, then why not go the "Turbine" route altogether? Why bother with all of the current "generality" improvements that are being added to the P2? If it can't compete as a general-purpose platform, then why make it look like one?
(side note: I disagree with Heater's view of "general purpose must run linux". I think "general purpose" is more about whether it can be used as an all-in-one solution for a reasonably wide variety of tasks. Running an operating system is not, in my opinion, necessary to still be considered "general purpose".)
Certainly if you take "running a standard OS" out of the mix, the P2 will be a fine general purpose MCU. Even the P1 can serve in that role if you're willing to connect a SPI or SQI flash chip to gain more program space.
If that's the case, then why not go the "Turbine" route altogether? Why bother with all of the current "generality" improvements that are being added to the P2? If it can't compete as a general-purpose platform, then why make it look like one?
These are excellent questions that have not been asked often enough.
Propeller is a microcontroller - simple, painless, cheap (relatively, hopefully), and highly effective in certain applications. Competing directly with RPi, BBB, et al. would be a fool's errand.
In place of that, however, could be made an interactive programming environment that runs bare-metal on a microcontroller, allowing rapid programming/prototyping without any loss or obscuration of real-time functionality. That is a market niche possibly open to be filled.
To be fair, I think this point of view has always existed for the Propeller platform. After all, how many existing Propeller products are meant to be used in conjunction with a genera-purpose computing platform and how many Propeller products are meant to be used as a stand-alone platform?
I hope I'm not being too pedantic, but I want to emphasize the point that there are many embedded applications which do not require the Propeller to ape a general-purpose computing platform, nor to connect to one.
I hope that you, Chip, your families and everyone here had a great Christmas!
I guess its time for me to publicly announce that I've been developing Propeller based add on boards for the Raspberry Pi
The first board would already have been fully tested had I not been down with the flu for a week... fortunately I am well enough to resume testing now!
My first two Pi products have been announced (two prototyping boards) ... I used them to develop my other Pi plates such as the Prop based plates...
(Heater knew a bit about this, and I appreciate him keeping it under his hat)
Parallax would also like to see a P1 Plate for the Pi. At present, our Product Plan for 2014 has our first Pi effort. Jeff Martin has been working to streamline the use of the Pi as a programming tool for Propellers, so it's a tool-oriented example rather than a piece of hardware. I think he's been in contact with Heater along the way, too. A video should be forthcoming within a week at the most. The next phase would be what you're all talking about - a hardware adapter providing access to the Propeller. However, we don't have the internal resources to do this design right now. If the community could make it happen we'd be supportive and possibly look at the manufacturing of the hardware.
Regarding OBC's comments, we were selling our QuickStart at less than our cost so it only made good business sense to us to try to cover our costs. I'd like to think that we were helpful in this regard for a long period of time.
Rumour has it that at least one forum member has designed a Pi plate for production recently. I can't name names but he is well known and I hope he has been in touch with you.
I have not had any contact from Jeff. Be glad to help if I can. I do have a SimpleIDE for the Pi and a loader adapted from the propgcc loader that works
using the UART on the Pi's GPIO.
Seairth, David,
My idea of a general purpose computer now a days is anything grunty enough to run Linux, At least a command line version of Linux. That seems fair enough as that encompasses some very small and cheap machines today.
The Prop and Prop II are not in any way general purpose computers. They may be up to calling "general purpose" in 1970/80 standards but that is nothing today.
I hope I'm not being too pedantic, but I want to emphasize the point that there are many embedded applications which do not require the Propeller to ape a general-purpose computing platform, nor to connect to one.
I wonder whether it would be going too far to say that *most* do not require that?
I hope that you, Chip, your families and everyone here had a great Christmas!
I guess its time for me to publicly announce that I've been developing Propeller based add on boards for the Raspberry Pi
The first board would already have been fully tested had I not been down with the flu for a week... fortunately I am well enough to resume testing now!
My first two Pi products have been announced (two prototyping boards) ... I used them to develop my other Pi plates such as the Prop based plates...
(Heater knew a bit about this, and I appreciate him keeping it under his hat)
This sounds very exciting. I'm about to dive into the RaspberryPi world and would love to try out one of your new Propeller Pi Plates. Let us know when you have them available for sale.
Comments
Seems to me that 90% of anything anyone tries fails. People start businesses, big companies develop products, all for nought.
Placing a bet on a horse that is not even born yet is rather a risky proposition. If you see what I mean.
As for the "community" thing. I have a GG board and it's great. I just wish there was a Parallax board that many others felt compelled to clone. Like the Arduino is an Arduino no matter who makes it. This sounds like "Oh no they are stealing our design and profits" but I believe such networking effects establish a standard and grow the market for everybody.
I've only awakened to P2 issues lately because its lengthy and expensive dev cycle seems to have become a drag on other aspects of Parallax's business. And because Parallax is my biggest customer, that was cause for some concern. However, because the P1 still offers a lot of promise for as-yet untapped apps, I'm not ready to diversify away to other platforms. I mean, there isn't a micro I'd rather design around and program than the Propeller, even if the "A" and "RP" platforms seem to be running away with the market.
Fortunately, the house is paid for, and the '82 Benz still runs. Life is good!
-Phil
In any case, it may not be for the best to simply view this as a recess of sorts.
That makes sense, but you can back both products of course.
Price and products are a double edge sword, Parallax have to compete with other vendors $12 Eval Boards, so entry price is very important.
That said, a price increase on a product you have planned around, is never a good thing, especially given others are lowering their prices.
If Quickstart has moved significantly in price, there does seem to be an opening for a minimal entry system, and smallest PCB size usually helps lower prices.
To me, the CP2105 is a smarter choice than FT23xx, as it can power the P1, and give two separate PGM and Communicate pathways.
A lot of effort went into the P8x32a - the design is still great, it's just too small and costly. Oh to have a simple upgrade to the B version with more memory and pins ... 3 years ago.
In a way Parallax warned that the Propeller-Platform would not be entirely viable by not initially producing boards from Jon's work (JonnyMac Spin-Zone in N&V May 2009).
Personally, I like the Propeller-Platform much better than Quickstart which is just one black eye after another that still pays boxing dividends - the Rev.B board is a much better design, but certain decisions on the Rev.A still cause me grief beyond the death of the Propeller-Platform.
Just a few assembly changes make the Propeller-Platform better (stacking headers for example). In my opinion, the board needs a revival. Martin's DNA board is really nice. Name another Propeller based board that has so many plug-ins besides the ASC+.
A P2 platform board is an interesting idea. Chip's vertical back-plane P2 board will be useful, but a flat platform board will find more utility IMHO. Of course, we need some chips first.
The Propeller 1 and its boards are simply too expensive compared to for example Raspberry Pi, which (1) costs $35 (2) I can buy it in my country without any problems.
Powerful FPGA boards like DE0-nano are available and cheap for educational purposes.
I am doing a PhD dissertation now. This is about signal processing - intelligent algorithms for old audio recording restoration. I planned to use P2 for it, but now it is impossible, as there is no P2. I switched to RPi+FPGA. My university paid $250 for DE2-115 I am using now - they have 50% discount for education. Then I learned I can make a lot of processors (nanoprocessors?) designed to do all I want in an FPGA chip using 1% of its resources. I don't need a Propeller board anymore for doing all this job. Then I discovered the Pi can be programmed in bare metal from scratch and then it can do all stuff hard to write in Verilog.
So, what I want from the Propeller? It's simple: (1) availability (2) availability (3) availability (4) availability... (n) availability
... and some lower prices and educational discounts.
And, of course, P2. And "a P2 Pi" - a cheap, wide available board with a P2, RAM and all connectivity like SD, USB, GPIO, VGA/HDMI etc.
A Pi and a Propeller are not comparable. They are totally different animals. All thought it does seem that Propeller boards are often expensive. As has been said here many times a Pi and a Prop complement each other and a Prop "plate" board for the Pi would be a great idea.
FPGA boardas are not really comparable either. Getting into Verilog/VHDL is a lot harder than writing C or Spin, the dev tools are huge and complicated the chips themselves are harder to work with if you want to build your own thing, The last time I looked GPGA boards were not cheap. However, yes, a PII should be able to tackle a lot of jobs people reach for an FPGA for.
"Nanoprocessors", for audio analysis, interesting, I would love to hear more about their architecture.
What's the idea with programming a Pi from scratch? I would have thought that throws away all the goodness of Linux and leaves you with a lot of work to do.
I'm all for a cheap cheap P2 board. However I'd like it to be a minimal as possible. Enough on board for the thing to operate and break out all the I/O pins. Even better if it is a Rasperry Pi format "plate". It should be usable stand alone of course. Some of those 2 million Pi users would love to look at the P2 that way, And that is a lot of potential customers.
I agree with Phil that there is still a lot to do with the P1, and with David regarding boards. I have used the P1 for years, but have many more to go before even coming close to using most of its potential. I too would love to see a standard platform. To the list of great boards I would add Wulfdens (Brian Riley) PPTH (Propeller Platform Thru Hole). It comes in kit form so students can develop soldering skills while building their own, and they have a great sense of ownership when their blinky LED or "Hello World" program runs. Priceless!
Parallax would also like to see a P1 Plate for the Pi. At present, our Product Plan for 2014 has our first Pi effort. Jeff Martin has been working to streamline the use of the Pi as a programming tool for Propellers, so it's a tool-oriented example rather than a piece of hardware. I think he's been in contact with Heater along the way, too. A video should be forthcoming within a week at the most. The next phase would be what you're all talking about - a hardware adapter providing access to the Propeller. However, we don't have the internal resources to do this design right now. If the community could make it happen we'd be supportive and possibly look at the manufacturing of the hardware.
Regarding OBC's comments, we were selling our QuickStart at less than our cost so it only made good business sense to us to try to cover our costs. I'd like to think that we were helpful in this regard for a long period of time.
Ken Gracey
It's great to hear that you have plans in place to do a RaspberryPi plate for the P1! I'd love to get involved on the software side if you need any help there. Let me know!
Thanks,
David
To be fair, I think this point of view has always existed for the Propeller platform. After all, how many existing Propeller products are meant to be used in conjunction with a genera-purpose computing platform and how many Propeller products are meant to be used as a stand-alone platform?
As processors go, they either need to do a few functions really well, or a lot of functions generally well. And I think it's obvious which category the P2 is falling into. With all of the new features that have been recently added to P2, it's no wonder that people are seeing it as a general-purpose computing platform. In fact, I'd even go so far as to suggest that the current P2 design encourages general-purpose computing.
Great news on the Propeller Pi plate front.
Rumour has it that at least one forum member has designed a Pi plate for production recently. I can't name names but he is well known and I hope he has been in touch with you.
I have not had any contact from Jeff. Be glad to help if I can. I do have a SimpleIDE for the Pi and a loader adapted from the propgcc loader that works
using the UART on the Pi's GPIO.
Seairth, David,
My idea of a general purpose computer now a days is anything grunty enough to run Linux, At least a command line version of Linux. That seems fair enough as that encompasses some very small and cheap machines today.
The Prop and Prop II are not in any way general purpose computers. They may be up to calling "general purpose" in 1970/80 standards but that is nothing today.
Bring back ZOG!!
Propeller:
Turbine:
Why go with two products? Because they meet two different needs. Propeller can be squarely targeted as a general-purpose computing platform ("do it all with one chip"), while the Turbine can be squarely targeted as an augmentation for all of the low-power computing platforms (RPi, Arduino, etc) that need significantly more I/O (and processing power to handle the I/O). And, of course, there's no reason that the Propeller (1 or 2) couldn't make use of the Turbine as well.
I realize that this would be a big undertaking (and cost), but I think it would give the Propeller product line a much clearer road map.
Just to be sure that I didn't miscommunicate, I should clarify what we are doing and what we are not doing.
We have only one Pi project in the works right now. It's a video tutorial and software to make it easy to use the Pi to program the Propeller 1 with SimpleIDE - a portable development tool. The purpose of this effort was to get familiar with the hardware and formalize something our customers have been doing.
We don't have a Propeller Pi Plate hardware design in the works. This is primarily because we have limited resources to do the project. If the community drives the effort and we can facilitate and support it, we will. It's a substantial software and hardware effort to do it correctly. This piece of hardware would bring the greatest financial return, but requires more effort than we are able to allocate directly. Basically, putting our key engineers on it would mean we've got to forgo or delay efforts around SimpleIDE 1.0, FPGA hardware for P2, etc. I'd love to see the community create these hardware designs and move Parallax out of the realm of thinking that we need to create every board. Having the community do this would allow us to increase focus on the chips.
If that's the case, then why not go the "Turbine" route altogether? Why bother with all of the current "generality" improvements that are being added to the P2? If it can't compete as a general-purpose platform, then why make it look like one?
(side note: I disagree with Heater's view of "general purpose must run linux". I think "general purpose" is more about whether it can be used as an all-in-one solution for a reasonably wide variety of tasks. Running an operating system is not, in my opinion, necessary to still be considered "general purpose".)
Propeller is a microcontroller - simple, painless, cheap (relatively, hopefully), and highly effective in certain applications. Competing directly with RPi, BBB, et al. would be a fool's errand.
In place of that, however, could be made an interactive programming environment that runs bare-metal on a microcontroller, allowing rapid programming/prototyping without any loss or obscuration of real-time functionality. That is a market niche possibly open to be filled.
Keep it simple.
I hope I'm not being too pedantic, but I want to emphasize the point that there are many embedded applications which do not require the Propeller to ape a general-purpose computing platform, nor to connect to one.
I hope that you, Chip, your families and everyone here had a great Christmas!
I guess its time for me to publicly announce that I've been developing Propeller based add on boards for the Raspberry Pi
The first board would already have been fully tested had I not been down with the flu for a week... fortunately I am well enough to resume testing now!
My first two Pi products have been announced (two prototyping boards) ... I used them to develop my other Pi plates such as the Prop based plates...
(Heater knew a bit about this, and I appreciate him keeping it under his hat)
I replied to Ken above, letting the proverbial cat out of the bag...