A world that appreciates dumbed down C running on a dumpy little AVR will surely come to appreciate the power and elegance of the P2, eventually. Nothing will hasten this more than a bunch of killer demos that 'duino Smile can't touch. Pretty soon the Arduino community will will be coming here to add an "accelerator board" to their hardware.
The way you reach a market segment is not to treat them like garbage or idiots, okay?
The fact is, the Arduino market is composed of non-technical artist types. The kind that don't do FFT audio filters in assembly for fun, or compose a video driver as a mental exercise like the folks here can do. They merely want to accomplish a given task and don't care about the supposed elegance of a slab of silicon that some guy composed on a workstation or the virtues of Spin and Pasm.
This difference has always ate at the Prop community and a lot of hate has been aimed at people who did nothing more than buy a board, that people here didn't approve of.
But the Arduino community is moving past the Prop community with the introduction of the Maple, Due, Pic32 based boards along now with the Beagleboneblack which is cheaper than a BS2 and can do HDMI video and can do real time I/O.
The way you reach a market segment is not to treat them like garbage or idiots, okay?
I wasn't treating the users as idiots. Just trying to put into perspective the hardware and software they use.
BTW, I don't see a big part of the Arduino community cottoning up to the BBB because its 'knowability' factor is approximately zero.
Edit: Just to make my position clear...I am delighted that the Arduino community exists! I'm quite familiar with the sense of satisfaction that comes from making something that works and learning a bundle in the process. Compared to traditional tools, It is delightful how the extraneous was filtered out and thrown away, along with the cryptic nomenclature.
The more Arduino users there are, the more P2 users there will eventually be.
That is what will be on my Parallella board when it arrives.
rjo__
I have a Raspberry Pi... really lame.
That's a bit harsh. In what way has the Raspi disappointed you? As far as I can tell they work as advertised and are very cheap.
...IMHO waste of time to hook a P2 to it. The worlds greatest micro-controller hooked up to the world's cheapest computer? Why?
But that's exactly what I want to do. Here are some "whys":
1) As you say it's very cheap.
2) Size, Combined Prop board and ARM board in the space of a cigarette packet.
3) Low power for remote or mobile use. Robot brains for example.
4) Out of the box I have: Ethernet, USB, superb video, audio, piles of storage space. WIFI is only a cheap dongle away.
5) I get all these services that come with Linux: file system, networking, http server, security, TLS/SSL, ssh etc etc etc
6) Easy to program, at both the Prop and ARM end.
7) Can run all the Prop development tools on the device itself.
8) Huge user base and hence lot's of support for sorting out those niggly problems that inevitably arise.
Pi and Prop look like a great combo to me. Of course there are more Pi like boards turning up on the market all the time. Like the recent BBB.
It's interesting that the ARMs are sprouting "cogs". Those little cores dedicated to doing real-time I/O intensive stuff. They are going to put pressure on the Prop II.
C... .. not so good for normals.
Strangely there seems to be hundreds of thousands of "normals" using C++ on their Arduino's quite happily. I think the reason they get on so well is that no one ever told them they were using C++ so they never flinched:)
Perhaps I should reconsider my statement that Arduino plus Prop is pointless. My thinking at the time was something like:
a) The Prop can do all that an Arduino can.
b) So why use the Arduino? Just use the Prop.
c) A companion processor for the Prop should provide the stuff that is hard/impossible on the Prop.
d) Conclusion: Prop + ARM is the way to go.
However. I can see that perhaps there are things that a Prop could offer Arduino users, even if they never actually program the Prop itself.
What about a VGA or TV terminal display for you Arduino project for example.
This would be brilliant. Instead of fighting the Arduino horde get out in front.
Yes! While sticking a Propeller shield onto an Arduino might be like "the tail wagging the dog," you must remember that marketing (especially marketing to mass markets) often has more to do with culture than it does with technology. And the Arduino is a cultural phenomenon. While there are much better-tasting, healthier and even cheaper foods available in society, there's a reason the Big Mac is so successful - availability, familiarity, etc. I think the mere fact that artists adopted the Arduino gave the Arduino a "cool factor" that no other device had had until then, so for the first time in human history it became hip to work with these controllers, and that is all it took to get the ball rolling. As has already been mentioned, now engineers are adopting the Arduino for the simple reason that they want to jump on the bandwagon, but instead of merely using what is now available and suffering with its limitations, they are expanding the capabilities of the Arduino and running with it. The Arduino, in effect, then becomes merely a sort of facade, a kind of name-brand storefront for a completely different circus that is being assembled in the back lot.
If done right, an Arduino could well be the sheep's wool the Propwolf needs to wear.
It's interesting that the ARMs are sprouting "cogs". Those little cores dedicated to doing real-time I/O intensive stuff. They are going to put pressure on the Prop II.
Yes indeed. It's very competitive out there, and getting more so by the day.
Heater, almost thanks for the link to FPGA/ARM it's broken.
Is that the Zynq which is an ARM A7, in the class with the RPi, and still a bit pricey? What I'm looking for (which doesn't exist) is an ARM M0 with a small FPGA. Closest might be Actel who have had soft ARM M1s for a long time, and I think they have a hard macro version as of a couple years ago but Actel always seems to be a few years behind Actel/Xilinx in FPGA capability.
Regarding multi-core CPUs, it is actually not new, I was working at Echelon when they introduced a 4 core CPU in the late 80s. In a lot of ways similar to the concept of the prop, no interrupts, tightly coupled IOs, shared memory. At the time we were trying to patent it, and found lots of prior art.
They have dual Cortex A9 processors. I'm not familiar with them but I guess I might be more so when the Parallella board turns up.
No idea about price but they can't be so expensive. That Adaptivia board is only $100 and includes the funky Parallella chip and all the normal ARM board peripherals.
instead Propeller ASC - call it Arduino MultiCore, provide the libs - and there you are
The "provide the libs" part is 90% of the effort. I've gone into debt to provide the hardware in the ASC. I have yet to make a single cent on these. I have offered to provide all the free hardware to anyone who was willing to tackle the "libs" part, because the software is over my head, but was met with the sound of crickets. It makes me very hesitant to spend even more time, and max out another credit card, going forward with a P2 version.
It's of my opinion, worthless as it may seem, that the Propeller community is more interested in impressing each other than in working together to achieve a common goal. I have to include myself and Parallax themselves in that assessment.
It's always hard to guess at gate equivalents between FPGA vendors, but my guess is it would fit maybe even 2 COGs and you could run Linux on it too, faster than a RPi. I actually have one sitting here, but they are a bit pricey at over $1K
.
It's of my opinion, worthless as it may seem, that the Propeller community is more interested in impressing each other than in working together to achieve a common goal. I have to include myself and Parallax themselves in that assessment.
Guilty. I confess to having common goal issues. It's a resource thing and I believe a community case of ADD. The PropGCC effort is one of the best community (small group) projects that have been spawned. most of the other creations are (sometimes Herculean) single person efforts to create something. It is a strange phenomenon for the overall culture of the community and the company.
There have been fits and starts at libraries to help integrate arduino and propeller...apparently, interests change the the embers die out.
There's pellerduino, an Arduino IDE extension that lets you program a Propeller from an Arduino IDE. It's gone inactive for the past 7 months or so but it did perform basic functions - a working proof of concept.
There's your great PropASC (which I own a few of) but a library would be a nice thing to make it easier to transition form Arduino to Propeller since you have provided the hardware platform to do so.
Currently, Parallax has resource constraints so it becomes a community project to do much of this. Get us to focus on something for long enough and it can get done!
It's of my opinion, worthless as it may seem, that the Propeller community is more interested in impressing each other than in working together to achieve a common goal.
How do you pick a common goal? Just looking over this thread alone, there are numerous disparate opinions of what would benefit the community the most. Meanwhile, there are all sorts of cooperative enterprises in the works; gcc features prominently among them.
I think it is great that individuals are working on different thing. The broader the project base, the better: One thing we all seem to agree on is that it's not good for a chip to be pigeon-holed.
The "provide the libs" part is 90% of the effort. I've gone into debt to provide the hardware in the ASC. I have yet to make a single cent on these. I have offered to provide all the free hardware to anyone who was willing to tackle the "libs" part, because the software is over my head, but was met with the sound of crickets.
It seems to me like what we are fretting over is the utter commoditization of computing, which threatens to melt everything into one big pot, with perhaps a few strata layers (C++, Unix, and protocol standards), ultimately leaving only MIPS- and megabytes-per-dollar as product differentiators. If one doesn't meet the qualifying criteria, they're not even going to wind up part of the melt, but rejected early as dross. This is probably where 90% of new, inexpensive embedded computing is headed, but as computing know-how expands in the world, so should the need for the odd, little bits of technology, which maybe most will never even hear of.
just give me a cool board that aint too blamed expensive to purchase so I can easily rationalize buying a small pile of them-- heck even do something like a couple of that have been described here alread. and make use of what has already been provided languages wise -- let me program it in whatever language suits me (forth, c/c++, basic, forth, etc.) -- ideally I would like to program it without needing a PC of any kind lashed to the thing and it would know how to auto load whatever badly formed program I should put into it to chew on. but if I prefer a PC to bang code on, then let me do that if I want to. Give me a nice pile of pin header areas (they don't need to be stuffed) that are laid out in a sensible arrangement. give me some supply pins (Vin, 5, 3.3, etc)....
and most of all, realize that I don't want to dominate the planet....I just want to have some fun and do something useful....imho some here fret too much....
It looks like he didn't get too far with the API. I did something similar, porting the basic API to SPIN with my Arduino Light object.
Maybe he was looking for help with the API and heard Martin's crickets.
While that spin object is no doubt great, the idea that someone has most of the pro-pellerino infrastructre ready for hookup is compelling for anyone who cares.
Maybe he was looking for help with the API and heard Martin's crickets.
While that spin object is no doubt great, the idea that someone has most of the pro-pellerino infrastructre ready for hookup is compelling for anyone who cares.
No arguments from me. I could probably flesh out the rest of the API in a few hours, that's how long it took to read the original Arduino API docs and re-write it in SPIN.
1) Be able to use Arduino shields with a Propeller micro - Propeller ASC, thank you, Martin - but we need libraries. And a P2 version will require funding....Kickstarter??
2) Plug a Propeller into a PC and program in using the Arduino IDE (convert the masses) - Pellerduino - but it needs a champion and a leader.
3) Plug a Propeller into a PC and program in SimpleIDE using Spin or PropGCC with API glue libraries - started but the development needs to be continued
4) Plug a Propeller board into a RasPi or BBB or favorite single board of your choice - all talk so far, needs hardware and the glue software (like Pi Alamode)
5) Sit down with a nice P2 board and program in whatever language we want and have fun (P2 BOE?, P2 Protoboard? P2 Quickstart?) - need a defined pinout standard for the platform so folks can make shields/capes/add-on boards/etc. that will fit more than one offering. Of course, there will be the P2 nano, P2 micro, P2 Mega, etc.......
6) Sit down at my P2 retro computer and play my favorite game or program a (relatively) simple system in a language I love like I did in the 80's
7) Work with a knowable system because things are getting too darn complex - thank you Chip, for the P1 and P2
8) Rapid prototype with a flexible chip that can be many things depending on what I need - thank you again Chip!!
9) Dominate the micro-controller world!!!
I think I've missed some but you can see there is a lot of hardware and software work to bring all these possibilities to market in a polished, usable form. All these needs project leaders and champions (and some need funding for the hardware) and teams to work on them and concentrate on delivery of a final solution.
So where do we take this as a community?
What should we expect Parallax to do to develop and market P2 products (keep their resources in mind and realize that some micro companies have groups dedicated to industries that dwarf the ENTIRE Parallax payroll)? What can we do as a community? What niches can be developed and exploited? (Face it, industry domination is NOT an option).
It seems to me like what we are fretting over is the utter commoditization of computing, which threatens to melt everything into one big pot, with perhaps a few strata layers (C++, Unix, and protocol standards), ultimately leaving only MIPS- and megabytes-per-dollar as product differentiators. If one doesn't meet the qualifying criteria, they're not even going to wind up part of the melt, but rejected early as dross. This is probably where 90% of new, inexpensive embedded computing is headed, but as computing know-how expands in the world, so should the need for the odd, little bits of technology, which maybe most will never even hear of.
This is something I've said for a long time, after surviving (somewhat) an acquisition that resulted in absorption. The lesson I took away from that experience is an understanding of how smaller, higher priced, companies and products compete in a world where everyone else has reduced the marketplace to commodities where margins are measured in small percentages.
Here are the secrets I learned:
Define your product, what it can do and what it can't do, stick to your guns.
Don't alter your product to chase fads, someone else always has a better plan to race you to the bottom.
Offer superior customer service that goes beyond anything the competitors offer.
Enable the customer, give them the tools to help themselves and make them feel empowered.
Ensure the product works.
Here's what you can expect if you follow those rules:
You can charge higher prices for a product that seemingly doesn't match up competitively with competitors.
You will have a smaller user base than your competitor.
Your product longevity will be longer than your competitor.
Your customers will be fiercely loyal and stay with you even when evidence would suggest overwhelmingly they should switch to a faster, cheaper product.
You will have much, much, higher margins than your competitors.
You will be less nimble.
Here's why it isn't a bad place to be:
If your treat your customers well and provide better service than the competitor, they will want to stay with you because you make it less painful to be a customer. People don't like to change, they like familiarity, which gives them comfort. While you will have less customers overall, each customer will be spending more than a similar customer at a competitor. You will make up much of the perceived loss by simply having fewer, more profitable, customers. Having fewer customers also means less infrastructure to support those customers. Having to support 20,000 customers is easier than 200,000 customers. When your loyalty goes up, the long term support costs go down, so the longer you retain a customer, the less they cost to support. There is initial acquisition costs associated with each new customer, turnover causes these costs to rise and overall efficiency to decline. When you have a turnover of 2% versus 25%, you spend a lot less time supporting existing customers and you can spend less money overall acquiring customers if you don't have to make up for turnover.
These lessons were learned working in an Internet company that sold services, but they apply well to product companies because product companies can't survive on being product companies solely, they have to offer services, and those services are the frosting that keeps the customers returning and happy.
The "provide the libs" part is 90% of the effort. I've gone into debt to provide the hardware in the ASC. I have yet to make a single cent on these. I have offered to provide all the free hardware to anyone who was willing to tackle the "libs" part, because the software is over my head, but was met with the sound of crickets. It makes me very hesitant to spend even more time, and max out another credit card, going forward with a P2 version.
It's of my opinion, worthless as it may seem, that the Propeller community is more interested in impressing each other than in working together to achieve a common goal. I have to include myself and Parallax themselves in that assessment.
Talk is cheap.
Martin, sorry to hear this. unfortunately I can not help here since C is not my language (yet?).
I think part of the problem is that there are to sides to the medal.
1: your Propeller ASC can be viewed as a Propeller dev board in the Aduino form factor and thus can use all the existing shields to it's benefit. Users will be Propeller developers using the propeller dev tools. This should offer quite some advantages, so I wonder, why you say it is not as popular as you expected it. Maybe because the propeller users are technically advanced and build their own HW shields and are not the typical Arduino users.
2. the other side is what I was talking about. having an Arduino board powered instead of the 8-bit AVR or the 32-bit ARM by a Prop1 or later a Prop2.
While the HW might be the same, this is rather different. It is based on the Arduino IDE with a lot of libraries that abstract the HW in the Arduino way. And as you say, this is quite a big effort. And the question is why should a prop developer or the prop community do such a thing. They are technically interested at another level than the usual Ardiono user, so they do not need it for their work.
And to make a profit from the HW you would have to invest a lot in marketing and sell at a pricepoint that might not be attractive to the arduino buyers.
So I see only one party that would really benefit from driving this: PARALLAX
Using it as a marketing tool for the Prop1 and with this opening the field for - as Localroger wrote here: Instead of fighting the Arduino horde get out in front
with the new Arduino Multi Core
Arguing about MIPS and MHz has long been a pet peeve of mine. I've even tangled with Leon about it a few times. It's like shopping for a car: Car A will go from 0 to 100 in 4.6 sec, whereas Car B takes 10.2. So you purchase Car A for your family, just to find out that it doesn't come with doors. You have to climb through the windows. And you have to sit on nylon webbing.
Making an engineering workstation with a BBB vs a P2 isn't the same thing at all...at least not for what I have in mind. I've got a BBB, and it's a wonderfully slick little device. I view it as an appliance, just like I do my Win7 laptop. But when I have to count precisely 4 billion clock ticks and compare them against the 1pps output of a Trimble Resolution T, while at the same time providing a display, filters, and a PLL function, and crank this code out in a leisurely 30 minutes (which is what I was doing this morning), nothing but the Prop will do. My experience has been that the combination of determinism, code re-usability, speed, and productivity are unparalleled on the Propeller.
P2 promises to increase such capabilities many-fold with dramatic speed, math, and analog enhancements. Meanwhile, doesn't the world already have enough web browsers and 'desktop-metaphor' operating systems? I'm absolutely ecstatic at the expectation of obtaining the performance I seek without the attendant intractable complexity. (n.b. I didn't say "baggage")
Comments
Hah! Ray, that's an interesting idea.
There is on Propeller beanie on the chip, and there is new artwork. More soon.
The fact is, the Arduino market is composed of non-technical artist types. The kind that don't do FFT audio filters in assembly for fun, or compose a video driver as a mental exercise like the folks here can do. They merely want to accomplish a given task and don't care about the supposed elegance of a slab of silicon that some guy composed on a workstation or the virtues of Spin and Pasm.
This difference has always ate at the Prop community and a lot of hate has been aimed at people who did nothing more than buy a board, that people here didn't approve of.
But the Arduino community is moving past the Prop community with the introduction of the Maple, Due, Pic32 based boards along now with the Beagleboneblack which is cheaper than a BS2 and can do HDMI video and can do real time I/O.
instead Propeller ASC - call it Arduino MultiCore, provide the libs - and there you are
This would be brilliant. Instead of fighting the Arduino horde get out in front.
I wasn't treating the users as idiots. Just trying to put into perspective the hardware and software they use.
BTW, I don't see a big part of the Arduino community cottoning up to the BBB because its 'knowability' factor is approximately zero.
Edit: Just to make my position clear...I am delighted that the Arduino community exists! I'm quite familiar with the sense of satisfaction that comes from making something that works and learning a bundle in the process. Compared to traditional tools, It is delightful how the extraneous was filtered out and thrown away, along with the cryptic nomenclature.
The more Arduino users there are, the more P2 users there will eventually be.
http://www.xilinx.com/products/silicon-devices/soc/zynq-7000/index.htm
That is what will be on my Parallella board when it arrives.
rjo__ That's a bit harsh. In what way has the Raspi disappointed you? As far as I can tell they work as advertised and are very cheap. But that's exactly what I want to do. Here are some "whys":
1) As you say it's very cheap.
2) Size, Combined Prop board and ARM board in the space of a cigarette packet.
3) Low power for remote or mobile use. Robot brains for example.
4) Out of the box I have: Ethernet, USB, superb video, audio, piles of storage space. WIFI is only a cheap dongle away.
5) I get all these services that come with Linux: file system, networking, http server, security, TLS/SSL, ssh etc etc etc
6) Easy to program, at both the Prop and ARM end.
7) Can run all the Prop development tools on the device itself.
8) Huge user base and hence lot's of support for sorting out those niggly problems that inevitably arise.
Pi and Prop look like a great combo to me. Of course there are more Pi like boards turning up on the market all the time. Like the recent BBB.
It's interesting that the ARMs are sprouting "cogs". Those little cores dedicated to doing real-time I/O intensive stuff. They are going to put pressure on the Prop II. Strangely there seems to be hundreds of thousands of "normals" using C++ on their Arduino's quite happily. I think the reason they get on so well is that no one ever told them they were using C++ so they never flinched:)
Perhaps I should reconsider my statement that Arduino plus Prop is pointless. My thinking at the time was something like:
a) The Prop can do all that an Arduino can.
b) So why use the Arduino? Just use the Prop.
c) A companion processor for the Prop should provide the stuff that is hard/impossible on the Prop.
d) Conclusion: Prop + ARM is the way to go.
However. I can see that perhaps there are things that a Prop could offer Arduino users, even if they never actually program the Prop itself.
What about a VGA or TV terminal display for you Arduino project for example.
Yes! While sticking a Propeller shield onto an Arduino might be like "the tail wagging the dog," you must remember that marketing (especially marketing to mass markets) often has more to do with culture than it does with technology. And the Arduino is a cultural phenomenon. While there are much better-tasting, healthier and even cheaper foods available in society, there's a reason the Big Mac is so successful - availability, familiarity, etc. I think the mere fact that artists adopted the Arduino gave the Arduino a "cool factor" that no other device had had until then, so for the first time in human history it became hip to work with these controllers, and that is all it took to get the ball rolling. As has already been mentioned, now engineers are adopting the Arduino for the simple reason that they want to jump on the bandwagon, but instead of merely using what is now available and suffering with its limitations, they are expanding the capabilities of the Arduino and running with it. The Arduino, in effect, then becomes merely a sort of facade, a kind of name-brand storefront for a completely different circus that is being assembled in the back lot.
If done right, an Arduino could well be the sheep's wool the Propwolf needs to wear.
Is that the Zynq which is an ARM A7, in the class with the RPi, and still a bit pricey? What I'm looking for (which doesn't exist) is an ARM M0 with a small FPGA. Closest might be Actel who have had soft ARM M1s for a long time, and I think they have a hard macro version as of a couple years ago but Actel always seems to be a few years behind Actel/Xilinx in FPGA capability.
Regarding multi-core CPUs, it is actually not new, I was working at Echelon when they introduced a 4 core CPU in the late 80s. In a lot of ways similar to the concept of the prop, no interrupts, tightly coupled IOs, shared memory. At the time we were trying to patent it, and found lots of prior art.
Sorry, I fixed it. Here it is again anyway: http://www.xilinx.com/products/silicon-devices/soc/zynq-7000/index.htm
They have dual Cortex A9 processors. I'm not familiar with them but I guess I might be more so when the Parallella board turns up.
No idea about price but they can't be so expensive. That Adaptivia board is only $100 and includes the funky Parallella chip and all the normal ARM board peripherals.
The "provide the libs" part is 90% of the effort. I've gone into debt to provide the hardware in the ASC. I have yet to make a single cent on these. I have offered to provide all the free hardware to anyone who was willing to tackle the "libs" part, because the software is over my head, but was met with the sound of crickets. It makes me very hesitant to spend even more time, and max out another credit card, going forward with a P2 version.
It's of my opinion, worthless as it may seem, that the Propeller community is more interested in impressing each other than in working together to achieve a common goal. I have to include myself and Parallax themselves in that assessment.
Talk is cheap.
Guilty. I confess to having common goal issues. It's a resource thing and I believe a community case of ADD. The PropGCC effort is one of the best community (small group) projects that have been spawned. most of the other creations are (sometimes Herculean) single person efforts to create something. It is a strange phenomenon for the overall culture of the community and the company.
There have been fits and starts at libraries to help integrate arduino and propeller...apparently, interests change the the embers die out.
There's pellerduino, an Arduino IDE extension that lets you program a Propeller from an Arduino IDE. It's gone inactive for the past 7 months or so but it did perform basic functions - a working proof of concept.
There's your great PropASC (which I own a few of) but a library would be a nice thing to make it easier to transition form Arduino to Propeller since you have provided the hardware platform to do so.
Currently, Parallax has resource constraints so it becomes a community project to do much of this. Get us to focus on something for long enough and it can get done!
How do you pick a common goal? Just looking over this thread alone, there are numerous disparate opinions of what would benefit the community the most. Meanwhile, there are all sorts of cooperative enterprises in the works; gcc features prominently among them.
I think it is great that individuals are working on different thing. The broader the project base, the better: One thing we all seem to agree on is that it's not good for a chip to be pigeon-holed.
I downloaded this earlier today. No time to work on it this week though. https://github.com/yishii/Pellerduino
It looks like he didn't get too far with the API. I did something similar, porting the basic API to SPIN with my Arduino Light object.
just give me a cool board that aint too blamed expensive to purchase so I can easily rationalize buying a small pile of them-- heck even do something like a couple of that have been described here alread. and make use of what has already been provided languages wise -- let me program it in whatever language suits me (forth, c/c++, basic, forth, etc.) -- ideally I would like to program it without needing a PC of any kind lashed to the thing and it would know how to auto load whatever badly formed program I should put into it to chew on. but if I prefer a PC to bang code on, then let me do that if I want to. Give me a nice pile of pin header areas (they don't need to be stuffed) that are laid out in a sensible arrangement. give me some supply pins (Vin, 5, 3.3, etc)....
and most of all, realize that I don't want to dominate the planet....I just want to have some fun and do something useful....imho some here fret too much....
Maybe he was looking for help with the API and heard Martin's crickets.
While that spin object is no doubt great, the idea that someone has most of the pro-pellerino infrastructre ready for hookup is compelling for anyone who cares.
No arguments from me. I could probably flesh out the rest of the API in a few hours, that's how long it took to read the original Arduino API docs and re-write it in SPIN.
1) Be able to use Arduino shields with a Propeller micro - Propeller ASC, thank you, Martin - but we need libraries. And a P2 version will require funding....Kickstarter??
2) Plug a Propeller into a PC and program in using the Arduino IDE (convert the masses) - Pellerduino - but it needs a champion and a leader.
3) Plug a Propeller into a PC and program in SimpleIDE using Spin or PropGCC with API glue libraries - started but the development needs to be continued
4) Plug a Propeller board into a RasPi or BBB or favorite single board of your choice - all talk so far, needs hardware and the glue software (like Pi Alamode)
5) Sit down with a nice P2 board and program in whatever language we want and have fun (P2 BOE?, P2 Protoboard? P2 Quickstart?) - need a defined pinout standard for the platform so folks can make shields/capes/add-on boards/etc. that will fit more than one offering. Of course, there will be the P2 nano, P2 micro, P2 Mega, etc.......
6) Sit down at my P2 retro computer and play my favorite game or program a (relatively) simple system in a language I love like I did in the 80's
7) Work with a knowable system because things are getting too darn complex - thank you Chip, for the P1 and P2
8) Rapid prototype with a flexible chip that can be many things depending on what I need - thank you again Chip!!
9) Dominate the micro-controller world!!!
I think I've missed some but you can see there is a lot of hardware and software work to bring all these possibilities to market in a polished, usable form. All these needs project leaders and champions (and some need funding for the hardware) and teams to work on them and concentrate on delivery of a final solution.
So where do we take this as a community?
What should we expect Parallax to do to develop and market P2 products (keep their resources in mind and realize that some micro companies have groups dedicated to industries that dwarf the ENTIRE Parallax payroll)? What can we do as a community? What niches can be developed and exploited? (Face it, industry domination is NOT an option).
Enough rambling from me........
This is something I've said for a long time, after surviving (somewhat) an acquisition that resulted in absorption. The lesson I took away from that experience is an understanding of how smaller, higher priced, companies and products compete in a world where everyone else has reduced the marketplace to commodities where margins are measured in small percentages.
Here are the secrets I learned:
Define your product, what it can do and what it can't do, stick to your guns.
Don't alter your product to chase fads, someone else always has a better plan to race you to the bottom.
Offer superior customer service that goes beyond anything the competitors offer.
Enable the customer, give them the tools to help themselves and make them feel empowered.
Ensure the product works.
Here's what you can expect if you follow those rules:
You can charge higher prices for a product that seemingly doesn't match up competitively with competitors.
You will have a smaller user base than your competitor.
Your product longevity will be longer than your competitor.
Your customers will be fiercely loyal and stay with you even when evidence would suggest overwhelmingly they should switch to a faster, cheaper product.
You will have much, much, higher margins than your competitors.
You will be less nimble.
Here's why it isn't a bad place to be:
If your treat your customers well and provide better service than the competitor, they will want to stay with you because you make it less painful to be a customer. People don't like to change, they like familiarity, which gives them comfort. While you will have less customers overall, each customer will be spending more than a similar customer at a competitor. You will make up much of the perceived loss by simply having fewer, more profitable, customers. Having fewer customers also means less infrastructure to support those customers. Having to support 20,000 customers is easier than 200,000 customers. When your loyalty goes up, the long term support costs go down, so the longer you retain a customer, the less they cost to support. There is initial acquisition costs associated with each new customer, turnover causes these costs to rise and overall efficiency to decline. When you have a turnover of 2% versus 25%, you spend a lot less time supporting existing customers and you can spend less money overall acquiring customers if you don't have to make up for turnover.
These lessons were learned working in an Internet company that sold services, but they apply well to product companies because product companies can't survive on being product companies solely, they have to offer services, and those services are the frosting that keeps the customers returning and happy.
Very astute observations, and seems to describe what Parallax is all about.
I think part of the problem is that there are to sides to the medal.
1: your Propeller ASC can be viewed as a Propeller dev board in the Aduino form factor and thus can use all the existing shields to it's benefit. Users will be Propeller developers using the propeller dev tools. This should offer quite some advantages, so I wonder, why you say it is not as popular as you expected it. Maybe because the propeller users are technically advanced and build their own HW shields and are not the typical Arduino users.
2. the other side is what I was talking about. having an Arduino board powered instead of the 8-bit AVR or the 32-bit ARM by a Prop1 or later a Prop2.
While the HW might be the same, this is rather different. It is based on the Arduino IDE with a lot of libraries that abstract the HW in the Arduino way. And as you say, this is quite a big effort. And the question is why should a prop developer or the prop community do such a thing. They are technically interested at another level than the usual Ardiono user, so they do not need it for their work.
And to make a profit from the HW you would have to invest a lot in marketing and sell at a pricepoint that might not be attractive to the arduino buyers.
So I see only one party that would really benefit from driving this: PARALLAX
Using it as a marketing tool for the Prop1 and with this opening the field for - as Localroger wrote here: Instead of fighting the Arduino horde get out in front
with the new Arduino Multi Core
Making an engineering workstation with a BBB vs a P2 isn't the same thing at all...at least not for what I have in mind. I've got a BBB, and it's a wonderfully slick little device. I view it as an appliance, just like I do my Win7 laptop. But when I have to count precisely 4 billion clock ticks and compare them against the 1pps output of a Trimble Resolution T, while at the same time providing a display, filters, and a PLL function, and crank this code out in a leisurely 30 minutes (which is what I was doing this morning), nothing but the Prop will do. My experience has been that the combination of determinism, code re-usability, speed, and productivity are unparalleled on the Propeller.
P2 promises to increase such capabilities many-fold with dramatic speed, math, and analog enhancements. Meanwhile, doesn't the world already have enough web browsers and 'desktop-metaphor' operating systems? I'm absolutely ecstatic at the expectation of obtaining the performance I seek without the attendant intractable complexity. (n.b. I didn't say "baggage")