Perhaps the P2, as it is different from the Propeller, could be renamed the Plex, as in a complex structure. Not that it is complex from a software viewpoint, rather a behind the scenes hardware organisation and computing foundation.
I seem to recall the original name for data structures was plex, but that was many, many moons ago; way back in the early days.
What comes after "Propeller" is "Turbine". The T16X32.
"Turbine" might be more compelling than "egg-beater", in reference to the cog/hub memory interface. Egg-beater sounds like old people in pajamas.
High speed generally, so what does that do for Spin? High Spin or Hi-Spin, TSpin (as in tsar, tsunami), Vortex, Whirl, Swirl, Twirl or perhaps even Tesla after his turbine and electrical engineering? You may even settle for 100K, after its speed, not as a financial offer! The world awaits.
Cyclone - as in Pratt & Whitney Cyclone radial engine. It had a troubled and prolonged development but went on to have a long illustrious career. Sounds like a fitting moniker.
Plus a radial engine is much like a Propeller. The only drawback it is 18 cylinders...Chip, can you squeeze in a couple more COGs?? (For marketing? )
All I can say is, I'm burnt out on this whole idea.
Many good comments above, however whicker really seems to have hit one of the nails on the head. It seems about impossible to resist the urge to tinker and improve an issue rather than simply fix it.
This is like Groundhog day, over and over again.
And it continuing to occur such that some would like to end it by reverting back to a P1+. Yes, Chip did say early on that such a thing would probably take no more than a week to do, however I read that as him coming off of a major P2Hot time sink/stressor, and not something he ever would have said in less stressful times.
There are so many new things in this new P2, that I think if so many here are confused, how the heck is anyone else new ever going to even remotely get up to speed? There is so much documentation that is going to be required, I don't think Parallax is going to be able to get it adequately documented by the time silicon starts rolling out of the fabs.
A project manager would be putting pressure on Chip to lock it down, freeze features or improvements, and make what you have work as your said it would, describe exactly what the Smartpins are, what they are to do, and get that done in the next 4 weeks.
So there you have it.
Whatever isn't working, get it fixed, and only fix it. No futzing around trying to meet some posters "Hey Chip, could you".
Smart pins are fine to have a 1st Gen, just get them to do what you originally wanted them to do, and thats it.
4 weeks seems like it should be enough time to do that, or at least it seemed like Chip had some pretty firm ideas on what he wanted of them.
So set a date already.
I suggest shooting for 12/15 to wrap up the main, feature complete design.
That way you and your family get some real quality time over the holidays, and more importantly, you can come back fresh in the new year to start getting some broadly scoped documentation done, so that testers can wrap their heads around what you did, and why you did it, and hopefully start proofing the actual design.
Or, keep the crazy train rolling along into the next fiscal quarter.
@koehler: It's getting done. Not too much left at this stage. You read it earlier, when I asked about smartpins.
PH: We aren't far from SmartPins now, are we Chip?
CG: No. I want to get the streamer docs done tomorrow, then the colorspace converter, then check into what the pixel mixer would need. Then, smartpins.
We are about to move onto those smart pins, and leave this part behind, sans the odd fix for bugs we may find along the way, and or a tweak to maximize something based on code we are writing.
Everybody flipped (well, some of us did) when Chip chose to drop the color circuit and pix into this design. Thing is, those are well designed and proven circuits, needing only integration. It was smart to keep this one lean while all the tricky stuff, streamer, fifo, hubex, etc... got done. Now that they are, and we have test suites running against them, it's not tough to add this stuff. And nobody asked for it either, just to nip that one in the bud early. Chip wants it in there. He gets to do that. I don't blame him. Bet you he worked pretty hard on those circuits.
In addition, we are zeroing in on a few sweet spot cases, making sure features do meaningful stuff. Peter has brought up a couple of those, as have others. What does "lock down" actually mean in this context? Does that mean we ship it with an errata sheet detailing all the stuff we locked down? lol This is novel design work. Doesn't work that way.
Here's the spec: It's done when it kicks Smile. Seriously. No joke.
Most everything is doing that now, and it's not all that tough to write code on. I've been at a version of Potatotext for P2 tonight, and here's the thing! One doesn't have to jump right into the deep water at all. Mostly, treat it like a P1, and as you get good at stuff, use it, or others can add / modify, whatever. I'm keeping it pretty simple myself, looking to exercise the streamer and such. Others are jamming hard on other parts of it, and together it's not looking bad at all.
(It's just gotta have a simple 8x8 character driver for testing. )
Want to help? Write some PASM. Also no joke. You want proofing? That's precisely how it's gonna get done. The more the merrier.
I'm quite tired of the naysayers. This one is looking good, the right people are involved, there is a partner for the difficult layout / integration tasks, etc... And all the lessons learned on hot are being applied here on a simpler design.
A few of us have written some nice code, using pretty advanced features. Others, like me for example, have glanced through that, mooched a piece or two, or saw how it's used and did just that. It's just not as hard as people are wanting it to be. I remember jumping on P1 and what that was like. People are gonna be able to use this. Besides, it's not all gonna be PASM anyway. We will get C soon, and SPIN soon after that, and both will offer true in-line PASM on top of that. Know what that means? Dead simple, easy, examples and just the perfect thing for people wanting to learn a feature or try a few instructions. Frankly, there is very little that is better for learning and doing things in assembly language.
And this one is gonna reward people who do that, and it's gonna perform just fine for those who don't want to do that too.
So, when we introduce the thing, we can do it in nice layers and with sample code, same as happened with P1. All we got with that was a Propeller Guts document, and the rest was code samples as docs. It was just fine. Over time, great tools happened.
On this one, we know stuff we didn't, and those better tools will be earlier and better too.
All we need to do is put some "Hey, look! It's magic!" sample / demo programs out there, and people will jump on board same as they did the last one, and they will do that, because this one is fun and powerful like the last one too.
Over time, docs, and other goodies will happen same as they did for the last one. It's getting done. Relax.
I presume that Parallax does not have a market-driven product manager that can speak Development, Customer, and Executive? Such a person would greatly benefit all parties.
I think I could have the ENTIRE Parallax P2 development/product management team over to my house for dinner without having to extend the dinner table...we might need one of the extra chairs. Which would be fun because Chip and Ken are very interesting and Daniel appears to be a good guy too! (let see, add my wife and daughter...yup, we're good for chairs! )
BTW: I think I'll feel a whole lot more comfortable when Parallax brings back the P2 web site... Then, I'll think that they think it's for real again...
Didn't someone recently mention that they were working on a high level PASM assembler? Presumably, that is in at least relatively portable C. I think I saw mention they were using C. I guess that would provide a tool that could be used natively under Linux.
Comments
I think it would have to be "Verilog (tm) Vortex"
Lazy Susan would probably be bad...
Plus a radial engine is much like a Propeller. The only drawback it is 18 cylinders...Chip, can you squeeze in a couple more COGs?? (For marketing? )
Turbines also take a while to 'spin up'...
Can't sleep now, I have a mind full of cows...
Try counting them?
And the rest of that disturbing collection:
https://www.youtube.com/user/cyriak/videos
Not sure what you've started, PH...
Many good comments above, however whicker really seems to have hit one of the nails on the head. It seems about impossible to resist the urge to tinker and improve an issue rather than simply fix it.
This is like Groundhog day, over and over again.
And it continuing to occur such that some would like to end it by reverting back to a P1+. Yes, Chip did say early on that such a thing would probably take no more than a week to do, however I read that as him coming off of a major P2Hot time sink/stressor, and not something he ever would have said in less stressful times.
There are so many new things in this new P2, that I think if so many here are confused, how the heck is anyone else new ever going to even remotely get up to speed? There is so much documentation that is going to be required, I don't think Parallax is going to be able to get it adequately documented by the time silicon starts rolling out of the fabs.
A project manager would be putting pressure on Chip to lock it down, freeze features or improvements, and make what you have work as your said it would, describe exactly what the Smartpins are, what they are to do, and get that done in the next 4 weeks.
So there you have it.
Whatever isn't working, get it fixed, and only fix it. No futzing around trying to meet some posters "Hey Chip, could you".
Smart pins are fine to have a 1st Gen, just get them to do what you originally wanted them to do, and thats it.
4 weeks seems like it should be enough time to do that, or at least it seemed like Chip had some pretty firm ideas on what he wanted of them.
So set a date already.
I suggest shooting for 12/15 to wrap up the main, feature complete design.
That way you and your family get some real quality time over the holidays, and more importantly, you can come back fresh in the new year to start getting some broadly scoped documentation done, so that testers can wrap their heads around what you did, and why you did it, and hopefully start proofing the actual design.
Or, keep the crazy train rolling along into the next fiscal quarter.
Just my $0.02
PH: We aren't far from SmartPins now, are we Chip?
We are about to move onto those smart pins, and leave this part behind, sans the odd fix for bugs we may find along the way, and or a tweak to maximize something based on code we are writing.
Everybody flipped (well, some of us did) when Chip chose to drop the color circuit and pix into this design. Thing is, those are well designed and proven circuits, needing only integration. It was smart to keep this one lean while all the tricky stuff, streamer, fifo, hubex, etc... got done. Now that they are, and we have test suites running against them, it's not tough to add this stuff. And nobody asked for it either, just to nip that one in the bud early. Chip wants it in there. He gets to do that. I don't blame him. Bet you he worked pretty hard on those circuits.
In addition, we are zeroing in on a few sweet spot cases, making sure features do meaningful stuff. Peter has brought up a couple of those, as have others. What does "lock down" actually mean in this context? Does that mean we ship it with an errata sheet detailing all the stuff we locked down? lol This is novel design work. Doesn't work that way.
Here's the spec: It's done when it kicks Smile. Seriously. No joke.
Most everything is doing that now, and it's not all that tough to write code on. I've been at a version of Potatotext for P2 tonight, and here's the thing! One doesn't have to jump right into the deep water at all. Mostly, treat it like a P1, and as you get good at stuff, use it, or others can add / modify, whatever. I'm keeping it pretty simple myself, looking to exercise the streamer and such. Others are jamming hard on other parts of it, and together it's not looking bad at all.
(It's just gotta have a simple 8x8 character driver for testing. )
Want to help? Write some PASM. Also no joke. You want proofing? That's precisely how it's gonna get done. The more the merrier.
I'm quite tired of the naysayers. This one is looking good, the right people are involved, there is a partner for the difficult layout / integration tasks, etc... And all the lessons learned on hot are being applied here on a simpler design.
A few of us have written some nice code, using pretty advanced features. Others, like me for example, have glanced through that, mooched a piece or two, or saw how it's used and did just that. It's just not as hard as people are wanting it to be. I remember jumping on P1 and what that was like. People are gonna be able to use this. Besides, it's not all gonna be PASM anyway. We will get C soon, and SPIN soon after that, and both will offer true in-line PASM on top of that. Know what that means? Dead simple, easy, examples and just the perfect thing for people wanting to learn a feature or try a few instructions. Frankly, there is very little that is better for learning and doing things in assembly language.
And this one is gonna reward people who do that, and it's gonna perform just fine for those who don't want to do that too.
So, when we introduce the thing, we can do it in nice layers and with sample code, same as happened with P1. All we got with that was a Propeller Guts document, and the rest was code samples as docs. It was just fine. Over time, great tools happened.
On this one, we know stuff we didn't, and those better tools will be earlier and better too.
All we need to do is put some "Hey, look! It's magic!" sample / demo programs out there, and people will jump on board same as they did the last one, and they will do that, because this one is fun and powerful like the last one too.
Over time, docs, and other goodies will happen same as they did for the last one. It's getting done. Relax.
I presume that Parallax does not have a market-driven product manager that can speak Development, Customer, and Executive? Such a person would greatly benefit all parties.
I think I could have the ENTIRE Parallax P2 development/product management team over to my house for dinner without having to extend the dinner table...we might need one of the extra chairs. Which would be fun because Chip and Ken are very interesting and Daniel appears to be a good guy too! (let see, add my wife and daughter...yup, we're good for chairs!
Excellent! For sure let us not add big side-view mirrors to our Ferrari because a customer wants to tow with it and we don't want to disappoint!
Here we are going boldly forward with the P2 and you want to boldly step backwards with an emulator for a legacy operating system!
(I know WINE Is Not an Emulator, but still).
P16X512A (then P16X512B etc for later revisions) - 16 cog, 512KB hub.
IMHO P= Parallax
Perhaps the "X" could become say "Y" ??? to define the cpu variant