Looking forward to P2
potatohead
Posts: 10,261
Here's why:
Just for grins, I bought two of the TI Stellaris Launch Pad packages. Now that I've built up some skill, I thought it good to explore another CPU. Still do, but...
Good grief! What a slog!
When I saw the Propeller years ago, and this was just after release time and before we had the great docs we do now, I jumped on a Demo Board. When the thing arrived, it didn't take 20 minutes to get setup to write some code. Stunningly easy. Still is.
Nothing else really compares. Downloads and installation easily took a few hours. Did it all on a fairly clean machine too. My old XP Thinkpad, which has done Propeller development up through just a few months ago when a Windows 7 machine replaced it, rarely has issues with software, because it's not used for much besides this kind of thing.
Then there are the sheer number of things one needs to know just to do ANYTHING.
Say what you want about SPIN + PASM. They are lean, mean, fast, robust, powerful, fun, and very easy. One of these days, I'll have to run Arduino in the same way for that data point.
Let's put the whole thing at 3 hours to setup to blink the LED. Demoboard? 10 minutes. Seriously. Maybe half an hour, if you count futzing around with the graphics_demo and loading up a few things just to see it go.
Now I don't have the TI booster pack, which includes some graphical display and can make better use of the library / driver code sitting in the chip ROM. That exists to make the most of the 256K flash, 32K RAM on-board. Both do serial though.
***Interestingly, TI simply uses their chip twice on their board. One is the debug / USB programming interface, the other is the target. That's notable, particularly at $4.99. Total loss, I know, but still notable that they kind of self-host and use that other chip to manage software license locking, etc...
I didn't actually do serial of any kind on the Prop for a long time, and that was due to the built in video display, which I found more to my liking. Once I did, I had very little trouble. Many other users, including some I'm helping get started, go right for it because it's easy to just use the terminal, or better PUTTY. I read through the TI getting started with serial example, and... it's thick. Won't be the quick, few line conversation on the serial port we are used to.
Most interestingly, I found that TI package sports a bunch of peripherals. These exceed what a P1 can do, particularly ADC / USB, but there is a lot of complexity about them as well. To really get 'em rocking, it's back to interrupt land and all that goes along with that. So there are 64 interrupts, levels, modes, and speedups / optimizations, like look-head register push consolidation onto the stack.... Cool beans, if you have to manage all of that.
I'm thinking of porting a Propeller video driver to it. Just because.
A quick survey of the required elements is stunning! For the sake of sane discussion, forget the P1 video generator features. Just bit bang both of them... On the P1, and soon P2, one can compartmentalize that whole thing into a COG and build out from there. On this thing, there is a lot more to do...
Now this isn't a "ARM / TI sucks" post. I honestly am not going there. No need to.
My point is the P1 isn't quite where it needs to be to address the more sophisticated or simply larger projects without a lot of skill being involved, and some external hardware here and there. P2 will be though.
Let's say a simple "Launch Pad" type starter kit is $30 - $50. In terms of time and barrier to entry, P2 has very good potential. TI bumped their kit to $12 - $18, I think and it needs extras to do graphics and such, though somebody could author a bit bang library and use simple circuits like we will end up doing on P2.
For all the discussion about things being minimal, friendly, yet capable, etc... Parallax is doing the right things. What I just experienced is the "pro" kit and all that goes along with that. Frankly, it's not lean, and it's not easy. On P2, we will have both options available, and that is a very exciting prospect!
Discussion? Please do not enter "it sucks" discussions. Really. Thanks.
Just for grins, I bought two of the TI Stellaris Launch Pad packages. Now that I've built up some skill, I thought it good to explore another CPU. Still do, but...
Good grief! What a slog!
When I saw the Propeller years ago, and this was just after release time and before we had the great docs we do now, I jumped on a Demo Board. When the thing arrived, it didn't take 20 minutes to get setup to write some code. Stunningly easy. Still is.
Nothing else really compares. Downloads and installation easily took a few hours. Did it all on a fairly clean machine too. My old XP Thinkpad, which has done Propeller development up through just a few months ago when a Windows 7 machine replaced it, rarely has issues with software, because it's not used for much besides this kind of thing.
Then there are the sheer number of things one needs to know just to do ANYTHING.
Say what you want about SPIN + PASM. They are lean, mean, fast, robust, powerful, fun, and very easy. One of these days, I'll have to run Arduino in the same way for that data point.
Let's put the whole thing at 3 hours to setup to blink the LED. Demoboard? 10 minutes. Seriously. Maybe half an hour, if you count futzing around with the graphics_demo and loading up a few things just to see it go.
Now I don't have the TI booster pack, which includes some graphical display and can make better use of the library / driver code sitting in the chip ROM. That exists to make the most of the 256K flash, 32K RAM on-board. Both do serial though.
***Interestingly, TI simply uses their chip twice on their board. One is the debug / USB programming interface, the other is the target. That's notable, particularly at $4.99. Total loss, I know, but still notable that they kind of self-host and use that other chip to manage software license locking, etc...
I didn't actually do serial of any kind on the Prop for a long time, and that was due to the built in video display, which I found more to my liking. Once I did, I had very little trouble. Many other users, including some I'm helping get started, go right for it because it's easy to just use the terminal, or better PUTTY. I read through the TI getting started with serial example, and... it's thick. Won't be the quick, few line conversation on the serial port we are used to.
Most interestingly, I found that TI package sports a bunch of peripherals. These exceed what a P1 can do, particularly ADC / USB, but there is a lot of complexity about them as well. To really get 'em rocking, it's back to interrupt land and all that goes along with that. So there are 64 interrupts, levels, modes, and speedups / optimizations, like look-head register push consolidation onto the stack.... Cool beans, if you have to manage all of that.
I'm thinking of porting a Propeller video driver to it. Just because.
A quick survey of the required elements is stunning! For the sake of sane discussion, forget the P1 video generator features. Just bit bang both of them... On the P1, and soon P2, one can compartmentalize that whole thing into a COG and build out from there. On this thing, there is a lot more to do...
Now this isn't a "ARM / TI sucks" post. I honestly am not going there. No need to.
My point is the P1 isn't quite where it needs to be to address the more sophisticated or simply larger projects without a lot of skill being involved, and some external hardware here and there. P2 will be though.
Let's say a simple "Launch Pad" type starter kit is $30 - $50. In terms of time and barrier to entry, P2 has very good potential. TI bumped their kit to $12 - $18, I think and it needs extras to do graphics and such, though somebody could author a bit bang library and use simple circuits like we will end up doing on P2.
For all the discussion about things being minimal, friendly, yet capable, etc... Parallax is doing the right things. What I just experienced is the "pro" kit and all that goes along with that. Frankly, it's not lean, and it's not easy. On P2, we will have both options available, and that is a very exciting prospect!
Discussion? Please do not enter "it sucks" discussions. Really. Thanks.
Comments
I will play with one as soon as Digikey gets stock - basically it is too cheap to ignore, and it has an FPU.
One thing is certain: Spin+PASM+cogs is a killer RAD for microcontroller projects!
These two items alone will open up a world of amazing projects which will be possible.
Jeff
The Prop 2's design, by contrast, has been scrutinized, discussed, kibitzed, and dissected ad nauseum in this forum. Like the faithful awaiting the Second Coming, forumistas have hung on every hint, rumor, sign, omen, and revelation to come forth. The air of expectation, like that for the next Star Wars episode, James Bond thriller, or iPhone, is almost palpable. The inevitabe result of so much anticiaption is that, rather than experiencing open-minded wonderment, some will expect things from the Prop 2 that it cannot deliver. This is only natural and not a reflection upon its ultimate design and specificaitons.
That's why I have chosen, at times but without much consistency, to quit the feeding frenzy. I want the Prop 2 to surpirse me -- to knock my socks off -- like the Prop 1 did. And, while it's tempting to pat ourselves on the backs for whatever real or imagined influence we may have had on its evolution, I still want to believe that it's the same product of a singular, coherent, and remarkable vision that the Prop 1 was.
So just call me when it's ready, Parallax. I'll be here, still probing the wonders and pushing the limits of the Prop 1!
-Phil
"Kibitz", thanks, great new word for my vocabulary.
Patiently waiting for Chipmas...
It will be a very fun chip to use with Forth... if we can only get everyone to clearly define their Forth lexicons.
In truth, whatever language one chooses to use, it will be a giant eye-opener and mind-bender. That's what Parallax does best.