Using distinct working names, that don't relate to the product name, is a typical method of reference a particular working incarnation. Eg: Teddy Bear, Black Bear, Brown Bear, Wheat Beer, ...
And there is a simple method to make numbers unique: invent a new one every time.
There aren't enough integers to label all the iterations of the Propeller design.
We could move to fractions or rational numbers I guess.
However some of the Propeller designs have been irrational. Then there are the complex Propeller designs to enumerate.
Luckily many of the suggestions made here are purely imaginary.
There aren't enough integers to label all the iterations of the Propeller design.
Yes, we could call the iterations like P3+i5, complex numbers. And then we could go and use quaternions, like this: http://www.wackerart.de/quaternionen/Quaternion12.jpg which gives a good imagination of how we are turning round and round. In the future we could use the cordic of the next prop to document the process of decision in real-time.
By the way: if we do not have enough integers to enumerate, we could use prims, as the Greek proved, there is an infinite number of those!
Regarding the naming convention debate, even for Parallax, I think that being this a niche product (do not misunderstand me please, there is some niche marketr where the prop can play as leader imho) it is not needed that it also haves a niche documentation... and this is more important than a name. In todays world, unfortunately, sometimes the dress counts more than the body it dresses.
If for example (even the current prop) will adopt a more standard (classic) datasheet like, ti, microchip and other players it will be better.
If you take eg a pic, 8052 based or other datasheet you see in the summary 1 usb, 3 uarts ,1 i2c, 1 spi, 3 timers ....
While everything is true than, by reading along the datasheet/manual, you discover that eg you lose a timer because it's needed for baudrate generation for the second uart you are using, and so on. All this figures are available all at the same time.
In the same way the current prop can states:
- 32 bit architecture
- 8 cores
- 32KB rom
- 48KB ram (16 dedicated, 32 shared)
- 32 IO
- 16 counters
- 32 pwm
- 16 tx/rx only uart
- 7 full (rx/tx/rts/cts) uart
- 7 spi
- 7 i2c
Of course this can't happen all at the same time, but where is the difference compared to other mcu?
At least the prop will have some official hardware (traditional/classic) figures that can be listed on various online mcu selection guides and thus be listed in digikey, mouser, rs-components and other ic distribution sites searches.
Regarding the naming convention debate, even for Parallax, I think that being this a niche product (do not misunderstand me please, there is some niche marketr where the prop can play as leader imho) it is not needed that it also haves a niche documentation... and this is more important than a name. In todays world, unfortunately, sometimes the dress counts more than the body it dresses.
If for example (even the current prop) will adopt a more standard (classic) datasheet like, ti, microchip and other players it will be better.
If you take eg a pic, 8052 based or other datasheet you see in the summary 1 usb, 3 uarts ,1 i2c, 1 spi, 3 timers ....
While everything is true than, by reading along the datasheet/manual, you discover that eg you lose a timer because it's needed for baudrate generation for the second uart you are using, and so on. All this figures are available all at the same time.
In the same way the current prop can states:
- 32 bit architecture
- 8 cores
- 32KB rom
- 48KB ram (16 dedicated, 32 shared)
- 32 IO
- 16 counters
- 32 pwm
- 16 tx/rx only uart
- 7 full (rx/tx/rts/cts) uart
- 7 spi
- 7 i2c
Of course this can't happen all at the same time, but where is the difference compared to other mcu?
At least the prop will have some official hardware (traditional/classic) figures that can be listed on various online mcu selection guides and thus be listed in digikey, mouser, rs-components and other ic distribution sites searches.
Totally agreed. A number of us have said this many times. An "*" with footnote that all peripherals are implemented by software using standard objects and are easily made more intelligent by extending the software.
Before you can solve a problem you need to know what the problem is...
So which market segment is the P16X64A meant to address?
The same market as the P8X32A - except this time around we will make the P16X64A twice as complicated in the hope that it will attract twice as much interest!
So which market segment is the P16X64A meant to address?
You are not the first to ask. No one knows.
There are rumours of educators who are looking forward to new improved Propeller systems. That makes a lot of sense to me because if you are introducing people to programming, including assembler, and micro-controllers and so forth the Propeller is about the easiest way to get started I have ever seen. Including those stupid Arduino things.
There are rumours that there are lot's of hobbyists out there who like the Propeller:) Again mostly because of it's amazing ease of use. They don't need massive performance most of the time just lot's of pins and RAM and interrupt/operating system free programming (i.e. many cores).
There was talk of industrial users. No idea if there are actually any or projected to be any. The folding of the Parallax Semiconductor initiative seems to suggest that is not a serious concern any more.
The main market segment that the P16X64A is aimed at is the same one the P8X32A was aimed at.
If I recall correctly, the first was primarily Chip's toy, something Chip wanted to build, because he wanted to bring the fun back into programming. to which I have to say, he did a damn fine job of it
To me it was and still is the best micro controller to program for, I even enjoy programming on it more than I enjoy my work. ( and yes, I enjoy my work )
This time, because Chip opened the design process to us all, it's grown into something we all want!
There are rumours of educators who are looking forward to new improved Propeller systems. That makes a lot of sense to me because if you are introducing people to programming, including assembler, and micro-controllers and so forth the Propeller is about the easiest way to get started I have ever seen. Including those stupid Arduino things.
There are rumours that there are lot's of hobbyists out there who like the Propeller:) Again mostly because of it's amazing ease of use. They don't need massive performance most of the time just lot's of pins and RAM and interrupt/operating system free programming (i.e. many cores).
There was talk of industrial users. No idea if there are actually any or projected to be any. The folding of the Parallax Semiconductor initiative seems to suggest that is not a serious concern any more.
I don't know, just saying...
LOL, I have a post from early this AM that I didn't post because it was perhaps a little too brutal.
In essence, after thinking about it, I may now be siding with Heater et al, and just say forget hub sharing and get it out the door.
My change of mind is based upon several things.
1. Parallax has tried to push s/w peripherals for 8 years now, with minimal success outside their existing market/s. Note, I am assuming their markets are just what Ken recently posted, educational, hobbyist, small design/prototype/tinkerers, etc.
2. They've been repeatedly pushed to put their money where their mouth is (so to speak), by providing basic s/w peripheral objects that do in fact meet or beat standard h/w periphperals. IIRC, nothing has happened on that front. Even the Gold Standard object idea from several years ago died.
3. As mentioned, the idea of going more mainstream with Parallax-Semi seems to have died. One can argue that if there were any actual revenue coming in from that, of any appreciable size, keeping that website up would have been a no-brainer.
4. Somewhere along the line, I and others in the forum seem to have gotten the idea that the Prop is/should be expanding and improving to meet the needs of the common uC engineer. Thats a natural and logical extension of any endeavor. In this case, I think it possible we all have it wrong, and Parallax is quite fine staying with the markets they have now, and the add-on sales that are pulled through their online store.
Yes, they have experimented with testing the waters with Parallax-Semi to see if they can expand, however that appears to have been shelved.
Considering all of this, which is partly obersvation, partly suppositition, I'm no longer convinced arguing for hubsharing is worth any effort.
Others who really need it of course can/will argue for it.
However, anyone who is arguing for it on the basis that this will ultimately improve P2 sales, adoption, general market appeal, is I think misguided.
If after 8 years Parallax can't be bothered to put a dozen objects up on the Prop page to show and compare against industry standard h/w, then that pretty much tells me they are not focused on the general market whatsoever.
So rather than being somewhat agnst ridden as to why Parallax hasn't done x, y, z, looking at it from this perspective makes a whole lot of sense in seeing what they do, and don't do maybe, from their perspective.
Unique company with unique products. HubExec/sharing would be significant improvements all around, however if it doesn't meet time/cost constraints, its lack of inclusion is not going to hurt them at all, as they aren't interested in greatly expanding their market.
There are rumours of educators who are looking forward to new improved Propeller systems. That makes a lot of sense to me because if you are introducing people to programming, including assembler, and micro-controllers and so forth the Propeller is about the easiest way to get started I have ever seen. Including those stupid Arduino things.
Parallax seems to sell a lot of product to educational users. Why is this? Is it because the Propeller is a great MCU for learning about programming? Maybe, although it might just be that Parallax is great at putting together exceptional hardware packages aimed at education and excellent teaching materials to go with them. If that's the case then it may not really matter that much what processor is running that high-level code. In fact, they're currently finding that many educators are reluctant to use a non-standard language and are asking for C. Not sure why they don't want a more modern language but C is what they seem to want. Also, is the Propeller really best for teaching assembly language programming? There are certain fairly standard things you can't do with the Propeller without resorting to self-modifying code which isn't a technique used by any other modern processor. It is a technique that has been largely abandoned for decades. So it could be that Parallax's success in the educational market is more due to their excellent educational materials rather than the Propeller's suitability as the execution platform. It could even be that the Propeller is holding back the success of their efforts in education because they have to constantly justify its odd architecture. Now, I like the Propeller and I'd like to see the new P2 succeed. I just think it's useful to try to figure out where its strengths really lie.
Now, I like the Propeller and I'd like to see the new P2 succeed. I just think it's useful to try to figure out where its strengths really lie.
Yes Educational sales are more about support than the silicon.
The other designers will be drawn to smart pins, and I think the Prop needs to focus on what the other uC do not do so well.
Cycle-deterministic execution and design control should be pushed, as others cannot do that.
There are plenty of ATE / T&M area applications, which might have used a FPGA.
The smart pin idea Chip floated sounded a great extension of this, a little unclear if that still is in the Mix ?
If selling this as a Smart Peripheral Controller, with 200MHz clocks, I will also add that USB is both expected, and a benchmark.
ie if anyone pitching a 200MHz/100MOP Smart Peripheral Controller, then has to say, "Oh, it can't quite do USB properly", that rather kills the sales pitch.
On a part over ~$4 these days, USB is pretty much a default. On some, even High Speed USB is now std.
Here is a topical example (chosen as not ARM deliberately)
[B]FTDI FT900 $4.25/10K[/B]
100MHz and 293DMIPS
256kB CODE 64kB DATA
USB 2.0 480Mbits/s
CAN
10/100 Ethernet support with built-in physical layer
10-bit, 1MHz DAC
7-channel multiplexed 10-bit, 1MSample/s ADC.
I2C slave and master,
I2S for external audio,
SPI slave and master,
SD card interface (2.0)
A variety of PWMs for standard and audio support.
Agreed regarding the smart pins, definitely something compromised on almost all other processors
Adding external ftdi or other comms chips is not a big deal and allows some futureproofing. Its interesting me even things like the DE0-Nano with full blown fpga onboard still have an FTDI 245 chip.
The Propeller is a great place to learn assembly language.
PASM does use self-modifying code and assembly language programmers need to know what that is and how it works. PASM is simple too. This means somebody can get to know what assembly language is. Having done that, they can go on to other CPU's and grok them more easily on a basic level. Propellers share these attributes with older, simple, 8 bit CPU's, which are also great ways to learn assembly language.
The Propeller is a great place to learn assembly language.
PASM does use self-modifying code and assembly language programmers need to know what that is and how it works.
I don't see any reason why programmers need to learn about self-modifying code. Can you give me one example of any modern processor that makes use of it other than the Propeller?
Edit: Of course, you *can* write self-modifying code for many if not most processors. I guess what I'm really asking is if there are any other modern processors that require the use of self-modifying code for critical operations like calling subroutines or indexing arrays.
Comments
We could move to fractions or rational numbers I guess.
However some of the Propeller designs have been irrational.
Then there are the complex Propeller designs to enumerate.
Luckily many of the suggestions made here are purely imaginary.
Yes. We could refer to the previous P2 design as "Edsel", or "DeLorean". Not sure what the current design concept would be in the car theme.
So P1/P8X32A for the one we already have, P2/P16X64A for the next iteration, and P3/Pwhatever is appropriate for the one after that.
By the way: if we do not have enough integers to enumerate, we could use prims, as the Greek proved, there is an infinite number of those!
If for example (even the current prop) will adopt a more standard (classic) datasheet like, ti, microchip and other players it will be better.
If you take eg a pic, 8052 based or other datasheet you see in the summary 1 usb, 3 uarts ,1 i2c, 1 spi, 3 timers ....
While everything is true than, by reading along the datasheet/manual, you discover that eg you lose a timer because it's needed for baudrate generation for the second uart you are using, and so on. All this figures are available all at the same time.
In the same way the current prop can states:
- 32 bit architecture
- 8 cores
- 32KB rom
- 48KB ram (16 dedicated, 32 shared)
- 32 IO
- 16 counters
- 32 pwm
- 16 tx/rx only uart
- 7 full (rx/tx/rts/cts) uart
- 7 spi
- 7 i2c
Of course this can't happen all at the same time, but where is the difference compared to other mcu?
At least the prop will have some official hardware (traditional/classic) figures that can be listed on various online mcu selection guides and thus be listed in digikey, mouser, rs-components and other ic distribution sites searches.
Sometimes Ross, you're simply too agreeable!
Only sometimes though ;-)
Hey! I resemble that remark!
One of my favourite sayings is "I could agree with you, but then we'd both be wrong."
Even when your logic is wrong, some of the time you just happen to get the right answer anyway, and so occasionally we are bound to agree!
Ross.
You certainly do ;-), and I am enjoying our little debate, it's definitely keeping me on my toes.
So which market segment is the P16X64A meant to address?
The same market as the P8X32A - except this time around we will make the P16X64A twice as complicated in the hope that it will attract twice as much interest!
Ross.
You are not the first to ask. No one knows.
There are rumours of educators who are looking forward to new improved Propeller systems. That makes a lot of sense to me because if you are introducing people to programming, including assembler, and micro-controllers and so forth the Propeller is about the easiest way to get started I have ever seen. Including those stupid Arduino things.
There are rumours that there are lot's of hobbyists out there who like the Propeller:) Again mostly because of it's amazing ease of use. They don't need massive performance most of the time just lot's of pins and RAM and interrupt/operating system free programming (i.e. many cores).
There was talk of industrial users. No idea if there are actually any or projected to be any. The folding of the Parallax Semiconductor initiative seems to suggest that is not a serious concern any more.
I don't know, just saying...
Hmmmm, I wonder if that might be the problem here?
If I recall correctly, the first was primarily Chip's toy, something Chip wanted to build, because he wanted to bring the fun back into programming. to which I have to say, he did a damn fine job of it
To me it was and still is the best micro controller to program for, I even enjoy programming on it more than I enjoy my work. ( and yes, I enjoy my work
This time, because Chip opened the design process to us all, it's grown into something we all want!
LOL, I have a post from early this AM that I didn't post because it was perhaps a little too brutal.
In essence, after thinking about it, I may now be siding with Heater et al, and just say forget hub sharing and get it out the door.
My change of mind is based upon several things.
1. Parallax has tried to push s/w peripherals for 8 years now, with minimal success outside their existing market/s. Note, I am assuming their markets are just what Ken recently posted, educational, hobbyist, small design/prototype/tinkerers, etc.
2. They've been repeatedly pushed to put their money where their mouth is (so to speak), by providing basic s/w peripheral objects that do in fact meet or beat standard h/w periphperals. IIRC, nothing has happened on that front. Even the Gold Standard object idea from several years ago died.
3. As mentioned, the idea of going more mainstream with Parallax-Semi seems to have died. One can argue that if there were any actual revenue coming in from that, of any appreciable size, keeping that website up would have been a no-brainer.
4. Somewhere along the line, I and others in the forum seem to have gotten the idea that the Prop is/should be expanding and improving to meet the needs of the common uC engineer. Thats a natural and logical extension of any endeavor. In this case, I think it possible we all have it wrong, and Parallax is quite fine staying with the markets they have now, and the add-on sales that are pulled through their online store.
Yes, they have experimented with testing the waters with Parallax-Semi to see if they can expand, however that appears to have been shelved.
Considering all of this, which is partly obersvation, partly suppositition, I'm no longer convinced arguing for hubsharing is worth any effort.
Others who really need it of course can/will argue for it.
However, anyone who is arguing for it on the basis that this will ultimately improve P2 sales, adoption, general market appeal, is I think misguided.
If after 8 years Parallax can't be bothered to put a dozen objects up on the Prop page to show and compare against industry standard h/w, then that pretty much tells me they are not focused on the general market whatsoever.
So rather than being somewhat agnst ridden as to why Parallax hasn't done x, y, z, looking at it from this perspective makes a whole lot of sense in seeing what they do, and don't do maybe, from their perspective.
Unique company with unique products. HubExec/sharing would be significant improvements all around, however if it doesn't meet time/cost constraints, its lack of inclusion is not going to hurt them at all, as they aren't interested in greatly expanding their market.
Yes Educational sales are more about support than the silicon.
The other designers will be drawn to smart pins, and I think the Prop needs to focus on what the other uC do not do so well.
Cycle-deterministic execution and design control should be pushed, as others cannot do that.
There are plenty of ATE / T&M area applications, which might have used a FPGA.
The smart pin idea Chip floated sounded a great extension of this, a little unclear if that still is in the Mix ?
If selling this as a Smart Peripheral Controller, with 200MHz clocks, I will also add that USB is both expected, and a benchmark.
ie if anyone pitching a 200MHz/100MOP Smart Peripheral Controller, then has to say, "Oh, it can't quite do USB properly", that rather kills the sales pitch.
On a part over ~$4 these days, USB is pretty much a default. On some, even High Speed USB is now std.
Here is a topical example (chosen as not ARM deliberately)
Adding external ftdi or other comms chips is not a big deal and allows some futureproofing. Its interesting me even things like the DE0-Nano with full blown fpga onboard still have an FTDI 245 chip.
PASM does use self-modifying code and assembly language programmers need to know what that is and how it works. PASM is simple too. This means somebody can get to know what assembly language is. Having done that, they can go on to other CPU's and grok them more easily on a basic level. Propellers share these attributes with older, simple, 8 bit CPU's, which are also great ways to learn assembly language.
Edit: Of course, you *can* write self-modifying code for many if not most processors. I guess what I'm really asking is if there are any other modern processors that require the use of self-modifying code for critical operations like calling subroutines or indexing arrays.