As a minimum it provides a big digital IO expander. Or multiple servo PWM driver. That's already a good start.
Have you heard of pi-blaster (derived from ServoBlaster)? It seems pretty capable and uses the DMA engine so the ARM isn't heavily burdened. I haven't used it so perhaps there's a serious downside. Perhaps other forms of sensor management would be compelling. I'm guessing that hundreds of millions of "sensor hubs" are being sold into the SmartPhone market. One trivial example (trivial for the prop ;-) would be managing multiple Pings.
Are you going to be the first to explain that to Chip re: the P2 video?
In our office VGA monitors out number HDMI capable screens ten to one. Those VGA monitors are only two years old.
Analog only VGA? Or do they have DVI too? HDMI to DVI cables are affordable. I have a free (to me) Dell LCD with DVI connected to my Pi. I picked up a cable at the local Fry's Electronics and they didn't gouge me. It's worked fine with Raspbian and RISC OS.
As far as I know the P2 can only generate video for VGA and old analog TV. That is David's "ancient technology". I too have been increasingly wondering why this is so.
As far as I know the P2 can only generate video for VGA and old analog TV. That is David's "ancient technology". I too have been increasingly wondering why this is so.
Obviously, "ancient" was an overstatement. However, I just went over to a neighbor's house to help him setup his new 55" Samsung LCD TV and it didn't have a VGA input. It did have composite and component video though so I guess it could still be connected to a P1 or P2. I just wonder how common VGA will be in a few years. I used to argue with Andre' LaMothe about this. He said that you could get any number of really good used VGA CRT monitors for next to nothing these days. I guess that's true but do you really want to build a business model on surplus parts? In any case, I think a P1 plate for the RaspberryPi would be useful even without a VGA output. It seems to me that the biggest advantage of the Propeller is the ability to interface to hardware using either standard (SPI or I2C) or custom interfaces easily. It is valuable even without taking into account its video capability.
Component video will be around for a long time yet. With HDTV, it's a great connection option. Old NTSC / PAL component scan rates still work just fine and the format provides for all the latest signals.
Got a 4K display? I know a friend who score one cheap this year. Component video is the only thing he can find to drive it at acceptable frame rates. HDMI isn't yet updated to that screen at more than a few HZ.
VGA monitors will be around for some time yet. When I read "need to move", I think adding capability is fine. I don't think removing it makes any real sense, unless cost is a serious driving factor. Doing this maximizes investments that are out there. An analog P2 will quite frankly drive anything people want to connect to it. On that basis alone, it's a nice chip.
HDMI requires lots of ugly licensing terms, conditions, etc... This is why display port got out there. And it's going to stay too, for those same reasons. I like HDMI, but I really don't like the legal along for the ride. Right now, we can do that with a small chip and that legal ends up being their problem, not ours.
I think VGA is in more trouble than component video is, but not for some time yet. Way too many VGA devices out there for there to be a worry just yet.
You are all talking like people who know what they can do with a Propeller, but I think the significant majority of RPi users are not going to be in that position. VGA? No good unless there are linux drivers that makes it all just work. I/O expander? Again, not much good without drivers.
But let's suppose we did have the drivers. Then what? Most of those RPi owners aren't going to know about the Propeller any more than they did before. Yes, some Propellers have been sold, but from the RPi community's point of view, they've bought a VGA board, an I/O expander board, etc. Don't get me wrong, that's a good thing. Propeller sales is Propeller sales. I just don't think it's Propeller exposure.
I think the exposure part comes in when you can use the Propeller in the way that distinguishes it from other CPUs. You need to be able to make it easy for the typical RPi user to change the code running in the cogs, and there needs to be a compelling reason for the RPi user to do it. In other words, the goal is for the RPi user to say "this ABC has a propeller, so I can easily change it to be an ABCD".
My idea about VGA on the Prop for the Pi does not require Linux drivers. All I want there is a VT100 terminal emulation running on the Prop that connects to the Pi UART for use as a console port. That software exists for the Prop already so an out of the box solution as you propose is quite doable.
Same applies for the I/O expander. It would not take much for Parallax to produce the Prop software required, No Linux drivers required. Just let people control it from their apps via the Pi UART.
I do agree with your main point though. Luckily Simple IDE, propgcc, and openspin all work easily on the Pi so users can program their Propeller "plate" anyhow they like directly from the Pi. Same as they do now for attached Arduino hardware. Of course it would be down to Parallax to make all this easily available and raise awareness of how the Prop can help Pi users with their projects.
My idea about VGA on the Prop for the Pi does not require Linux drivers. All I want there is a VT100 terminal emulation running on the Prop that connects to the Pi UART for use as a console port. That software exists for the Prop already so an out of the box solution as you propose is quite doable.
Same applies for the I/O expander. It would not take much for Parallax to produce the Prop software required, No Linux drivers required. Just let people control it from their apps via the Pi UART.
I do agree with your main point though. Luckily Simple IDE, propgcc, and openspin all work easily on the Pi so users can program their Propeller "plate" anyhow they like directly from the Pi. Same as they do now for attached Arduino hardware. Of course it would be down to Parallax to make all this easily available and raise awareness of how the Prop can help Pi users with their projects.
Is the UART really the best way to talk between Linux and the Propeller? Would it be faster than a SPI connection for example?
VGA, etc... work driverless. In the case of it being a console, it just needs to get the text. The Prop does the rest. If it's other things, then it just needs the data.
Expanding the GPIO seems a smart use case to me as well. Since the GPIO pins are already available, what is really needed is some sample code to demonstrate how to talk to the Prop. If that sample code did a few useful things, like display some text, set I/O pin states, streamed data in and out, etc... that's it. The Pi users could take it from there just like they would any other thing they hang off the GPIO pins.
Same goes for the UART.
The fact that the Pi is running Linux really doesn't mean anything. If they are programming in something and that something gets at the GPIO pins, they have what they need to work with the Propeller.
On the Prop side, just package up stuff we've got right now and write that other little bit of sample code.
Ahh, never mind. I see Heater got there in plenty of time.
Just out of curiosity, does anyone know what percentage of Parallax P1 "commercial" customers make use of the video features of P1? The only commercial customer that I know of is Hoverfly and they don't seem to use video. Are there others that do? It seems like the ability of P1 to generate video easily is one of the things that distinguish it from other MCUs. It is obviously heavily used by hobbyists but I'm wondering if it is also a big reason that commercial customers choose Propeller.
Is the UART really the best way to talk between Linux and the Propeller? Would it be faster than a SPI connection for example?
I don't know about best but it is certainly the simplest to get people started with. I could use a Propeller VGA VT100 terminal right now so serial is a must.
SPI gets a lot more complex at the Pi end, I'm sure it requires a Linux kernel driver.
SPI is there as an option for anyone who wants to jumper the Pi GPIO to a Prop pin so all we need is the software at both ends.
I too have wondered about use of video on the Prop. With the recent emphasis of Parallax on commercial customers I might have expected video to have been dropped on the P2, I don't think Chip would ever consider that though.
I love the video features on the p1...it is the best feature of the chip...but, I am only hobbyist...and the video is very low res in a high def video world. However, with the p2 high res video capabilities I am certain many many commercial customers will be very interested in that feature and it could be a huge selling point. So, I am not sure it is fair to compare the commercial usage of the p1 video capabilities to the p2 video capabilities. Plus...in the world of electronics one of the biggest growth has to be in video driver displays...look around...we now have refrigerators with video displays..:)
I love the video features on the p1...it is the best feature of the chip...but, I am only hobbyist...and the video is very low res in a high def video world. However, with the p2 high res video capabilities I am certain many many commercial customers will be very interested in that feature and it could be a huge selling point. So, I am not sure it is fair to compare the commercial usage of the p1 video capabilities to the p2 video capabilities. Plus...in the world of electronics one of the biggest growth has to be in video driver displays...look around...we now have refrigerators with video displays..:)
Actually, I was talking about P1 because we were discussing making a RaspberryPi plate using P1 and whether it would be useful to have VGA on it. You're right that video on P2 will be a whole new story and it will be interesting to see how it is used.
P2 video is, with perhaps the only real issue being the need to add an HDMI encoder. If P2 video doesn't see wider adoption, then I don't think there will be broad video adoption.
There are some nice things about simple video P1 style, and one of them is a very lean API and speed is the other. I don't know whether or not P1 video + a Pi makes sense for things besides a console, but it might. Not hard to offer, if the P1 is offered. On P1, it doesn't do much, but then again, it doesn't take much either. Just a nice to have option, and I remember Chip talking about how very little it actually required on P1.
P2 is a bigger video investment obviously.
After thinking about it, the other use for P1 video would be P1 diagnostic / status, etc... and it could be very useful while they are programmed in tandem. It might be largely ignored, but for the console case though. But then again, I really don't have a good handle on what the vast majority of users actually do with their Pi.
I know a lot of people are using them as emulation platforms too. The composite video output is seeing some use in that area. Too bad they didn't provide for component on the board at least... It's old school composite, or HDMI.
The student I'm mentoring got a Pi, and it took him forever to get a handle on the GPIO pins and PYTHON. He really liked PYTHON, but he didn't like the mess involving the pins. It was that which brought him to a Propeller. And he's worked up to a spiffy robot, some of which has appeared here. The Prop and SPIN + PASM got him to a place where he's right at the basics, learning about signals, etc... and that is the "Chip sweet spot" as intended.
BTW: I did IR learning with him, because I never went down that road. It was a great experience. We had a couple of goofy Propeller PASM things to get past, but the overall process was as lean as any. He has good mastery now, and that's just one of many things a Prop could do for a Pi, BTW Maybe that is something Pi users would appreciate. Package up an I/O + sensor problem in the Prop domain and just stream the results to the Pi where higher level things respond. IMHO, that's a likely winner.
The Pi was going to be the center of that stuff, but it's too thick, not lean enough to stay out of the way of the real basic learning that needs to happen. I thought this notable. Still do.
I use a Prop for the very same reason I will still use old computers. Lean. You can get right to the problem, clear most everything else away and just solve it, learn it, do it and that's a powerful thing.
**And that no OS dynamic is one reason I'm pretty stoked about P2, which should be attached to a Pi from near day one. Somebody needs to be planning that board as we speak here. I think it's a can't miss kind of thing.
Perhaps this dynamic has played out in ways that limit how many Pi users actually do that kind of thing with their Pi. Perhaps a much larger fraction of two million would be interested. Nobody is going to know, unless they get out there and ask, which is rational and recommended, or offer the product, which may well be the cheaper option overall. Just do it reasonably and quickly and see where it goes.
I'm sure only a small percentage of them are being used for hardware related projects.
I think so too. And I suspect that a good many end up just collecting dust. Also there's education, which, of course, was ostensibly RPi's purpose in life anyway.
BUT a small percentage of two million is still a big number!
Yes! And even if Parallax doesn't turn a large profit from the plate (?) itself, it will at least earn Propeller some notice.
My Pi has been gathering dust since a month after acquiring it.
I was actually hoping for a powerful microcontroller with hi-res graphics, which it sort of is but I didn't want it to be reliant on linux to be able to use usb and the graphics etc. I felt sure that someone would put together a simple instant boot up operating system (like the beeb), with an assembler and maybe a high level language; (even 'C' which I am beginning to accept as inevitable, as I start to get good results on the Arduino).
I really must get around to trying RiscOs - my bad!
Thats why I got the P1, and monitor with anticipation this forum and dream of the P2 fulfilling my dreams. (especially if Bean ports his Basic over).
Dave
Just out of curiosity, does anyone know what percentage of Parallax P1 "commercial" customers make use of the video features of P1? The only commercial customer that I know of is Hoverfly and they don't seem to use video. Are there others that do? It seems like the ability of P1 to generate video easily is one of the things that distinguish it from other MCUs. It is obviously heavily used by hobbyists but I'm wondering if it is also a big reason that commercial customers choose Propeller.
None of my products are 'production', but there are several dozen huge welding positioners that I've upgraded to use a prop1 for HMI and kinematics. I use the various VGA drivers exclusively.
There are also two power plants that display various stats about coal and ash silos (Wirelessly gathered, XBEE) on a vga monitor. These are blatantly constructed on protoboards! The protoboard is installed into the back of the monitor and even runs from the monitor's power supply.
This is about 15% of my applications.
BTW..
I am currently planning a project with the R-PI that will use the propeller for timing and I/O. I plan to boot the prop directly from the R-PI with 12c as needed, using a gpio for reset, so retaining the ability to load and reload prop firmware as needed.
Comments
We need to move into the modern age rather than assuming people have ancient technology.
In our office VGA monitors out number HDMI capable screens ten to one. Those VGA monitors are only two years old.
Have you heard of pi-blaster (derived from ServoBlaster)? It seems pretty capable and uses the DMA engine so the ARM isn't heavily burdened. I haven't used it so perhaps there's a serious downside. Perhaps other forms of sensor management would be compelling. I'm guessing that hundreds of millions of "sensor hubs" are being sold into the SmartPhone market. One trivial example (trivial for the prop ;-) would be managing multiple Pings.
Analog only VGA? Or do they have DVI too? HDMI to DVI cables are affordable. I have a free (to me) Dell LCD with DVI connected to my Pi. I picked up a cable at the local Fry's Electronics and they didn't gouge me. It's worked fine with Raspbian and RISC OS.
Got a 4K display? I know a friend who score one cheap this year. Component video is the only thing he can find to drive it at acceptable frame rates. HDMI isn't yet updated to that screen at more than a few HZ.
VGA monitors will be around for some time yet. When I read "need to move", I think adding capability is fine. I don't think removing it makes any real sense, unless cost is a serious driving factor. Doing this maximizes investments that are out there. An analog P2 will quite frankly drive anything people want to connect to it. On that basis alone, it's a nice chip.
HDMI requires lots of ugly licensing terms, conditions, etc... This is why display port got out there. And it's going to stay too, for those same reasons. I like HDMI, but I really don't like the legal along for the ride. Right now, we can do that with a small chip and that legal ends up being their problem, not ours.
I think VGA is in more trouble than component video is, but not for some time yet. Way too many VGA devices out there for there to be a worry just yet.
But let's suppose we did have the drivers. Then what? Most of those RPi owners aren't going to know about the Propeller any more than they did before. Yes, some Propellers have been sold, but from the RPi community's point of view, they've bought a VGA board, an I/O expander board, etc. Don't get me wrong, that's a good thing. Propeller sales is Propeller sales. I just don't think it's Propeller exposure.
I think the exposure part comes in when you can use the Propeller in the way that distinguishes it from other CPUs. You need to be able to make it easy for the typical RPi user to change the code running in the cogs, and there needs to be a compelling reason for the RPi user to do it. In other words, the goal is for the RPi user to say "this ABC has a propeller, so I can easily change it to be an ABCD".
My idea about VGA on the Prop for the Pi does not require Linux drivers. All I want there is a VT100 terminal emulation running on the Prop that connects to the Pi UART for use as a console port. That software exists for the Prop already so an out of the box solution as you propose is quite doable.
Same applies for the I/O expander. It would not take much for Parallax to produce the Prop software required, No Linux drivers required. Just let people control it from their apps via the Pi UART.
I do agree with your main point though. Luckily Simple IDE, propgcc, and openspin all work easily on the Pi so users can program their Propeller "plate" anyhow they like directly from the Pi. Same as they do now for attached Arduino hardware. Of course it would be down to Parallax to make all this easily available and raise awareness of how the Prop can help Pi users with their projects.
Expanding the GPIO seems a smart use case to me as well. Since the GPIO pins are already available, what is really needed is some sample code to demonstrate how to talk to the Prop. If that sample code did a few useful things, like display some text, set I/O pin states, streamed data in and out, etc... that's it. The Pi users could take it from there just like they would any other thing they hang off the GPIO pins.
Same goes for the UART.
The fact that the Pi is running Linux really doesn't mean anything. If they are programming in something and that something gets at the GPIO pins, they have what they need to work with the Propeller.
On the Prop side, just package up stuff we've got right now and write that other little bit of sample code.
Ahh, never mind. I see Heater got there in plenty of time.
SPI gets a lot more complex at the Pi end, I'm sure it requires a Linux kernel driver.
SPI is there as an option for anyone who wants to jumper the Pi GPIO to a Prop pin so all we need is the software at both ends.
I too have wondered about use of video on the Prop. With the recent emphasis of Parallax on commercial customers I might have expected video to have been dropped on the P2, I don't think Chip would ever consider that though.
I love the video features on the p1...it is the best feature of the chip...but, I am only hobbyist...and the video is very low res in a high def video world. However, with the p2 high res video capabilities I am certain many many commercial customers will be very interested in that feature and it could be a huge selling point. So, I am not sure it is fair to compare the commercial usage of the p1 video capabilities to the p2 video capabilities. Plus...in the world of electronics one of the biggest growth has to be in video driver displays...look around...we now have refrigerators with video displays..:)
P2 video is, with perhaps the only real issue being the need to add an HDMI encoder. If P2 video doesn't see wider adoption, then I don't think there will be broad video adoption.
There are some nice things about simple video P1 style, and one of them is a very lean API and speed is the other. I don't know whether or not P1 video + a Pi makes sense for things besides a console, but it might. Not hard to offer, if the P1 is offered. On P1, it doesn't do much, but then again, it doesn't take much either. Just a nice to have option, and I remember Chip talking about how very little it actually required on P1.
P2 is a bigger video investment obviously.
After thinking about it, the other use for P1 video would be P1 diagnostic / status, etc... and it could be very useful while they are programmed in tandem. It might be largely ignored, but for the console case though. But then again, I really don't have a good handle on what the vast majority of users actually do with their Pi.
It's a mystery.
I get the feeling a vast majority of them end up getting used as media players with XBMC.
I'm sure only a small percentage of them are being used for hardware related projects.
BUT a small percentage of two million is still a big number!
I know a lot of people are using them as emulation platforms too. The composite video output is seeing some use in that area. Too bad they didn't provide for component on the board at least... It's old school composite, or HDMI.
The student I'm mentoring got a Pi, and it took him forever to get a handle on the GPIO pins and PYTHON. He really liked PYTHON, but he didn't like the mess involving the pins. It was that which brought him to a Propeller. And he's worked up to a spiffy robot, some of which has appeared here. The Prop and SPIN + PASM got him to a place where he's right at the basics, learning about signals, etc... and that is the "Chip sweet spot" as intended.
BTW: I did IR learning with him, because I never went down that road. It was a great experience. We had a couple of goofy Propeller PASM things to get past, but the overall process was as lean as any. He has good mastery now, and that's just one of many things a Prop could do for a Pi, BTW Maybe that is something Pi users would appreciate. Package up an I/O + sensor problem in the Prop domain and just stream the results to the Pi where higher level things respond. IMHO, that's a likely winner.
The Pi was going to be the center of that stuff, but it's too thick, not lean enough to stay out of the way of the real basic learning that needs to happen. I thought this notable. Still do.
I use a Prop for the very same reason I will still use old computers. Lean. You can get right to the problem, clear most everything else away and just solve it, learn it, do it and that's a powerful thing.
**And that no OS dynamic is one reason I'm pretty stoked about P2, which should be attached to a Pi from near day one. Somebody needs to be planning that board as we speak here. I think it's a can't miss kind of thing.
Perhaps this dynamic has played out in ways that limit how many Pi users actually do that kind of thing with their Pi. Perhaps a much larger fraction of two million would be interested. Nobody is going to know, unless they get out there and ask, which is rational and recommended, or offer the product, which may well be the cheaper option overall. Just do it reasonably and quickly and see where it goes.
Yes! And even if Parallax doesn't turn a large profit from the plate (?) itself, it will at least earn Propeller some notice.
I was actually hoping for a powerful microcontroller with hi-res graphics, which it sort of is but I didn't want it to be reliant on linux to be able to use usb and the graphics etc. I felt sure that someone would put together a simple instant boot up operating system (like the beeb), with an assembler and maybe a high level language; (even 'C' which I am beginning to accept as inevitable, as I start to get good results on the Arduino).
I really must get around to trying RiscOs - my bad!
Thats why I got the P1, and monitor with anticipation this forum and dream of the P2 fulfilling my dreams. (especially if Bean ports his Basic over).
Dave
None of my products are 'production', but there are several dozen huge welding positioners that I've upgraded to use a prop1 for HMI and kinematics. I use the various VGA drivers exclusively.
There are also two power plants that display various stats about coal and ash silos (Wirelessly gathered, XBEE) on a vga monitor. These are blatantly constructed on protoboards! The protoboard is installed into the back of the monitor and even runs from the monitor's power supply.
This is about 15% of my applications.
BTW..
I am currently planning a project with the R-PI that will use the propeller for timing and I/O. I plan to boot the prop directly from the R-PI with 12c as needed, using a gpio for reset, so retaining the ability to load and reload prop firmware as needed.
2 millions+ units sold means at least some reverse engineers having some of them