Where Parallax Heading to
EE_davran
Posts: 3
in Propeller 2
Hi P2 Designers,
I want to share my opinion on latest status of the P2 forum. According to my observation, Parallax grew by making micro-controllers easy for engineers and hobbyist with products like basic stamp not by making microcontrollers.
In P1 philosophy, there was a statement saying that in a development of a project, major expense is a software development side.
I think, consumers does not want a super sophisticated product from Parallax, consumers want something easy to use. Company owner Chip also may have forgotten this.
Focusing on the "silicon die", "in-depth technical things" is good for technical success but this does not guarantee commercial success. Looking back to consumer needs is important.
P2 OBEX or documents like "Propeller 2 Education Kit Labs Fundamentals" are things which are as important as the silicon die. This is my feedback to you guys for commercial success before P2 becomes alive. I also should note that, although i am 10.000km far away from you, i m excited for the new product and plan to buy "Propeller 2 ES Evaluation Board" when it is available.
I want to share my opinion on latest status of the P2 forum. According to my observation, Parallax grew by making micro-controllers easy for engineers and hobbyist with products like basic stamp not by making microcontrollers.
In P1 philosophy, there was a statement saying that in a development of a project, major expense is a software development side.
I think, consumers does not want a super sophisticated product from Parallax, consumers want something easy to use. Company owner Chip also may have forgotten this.
Focusing on the "silicon die", "in-depth technical things" is good for technical success but this does not guarantee commercial success. Looking back to consumer needs is important.
P2 OBEX or documents like "Propeller 2 Education Kit Labs Fundamentals" are things which are as important as the silicon die. This is my feedback to you guys for commercial success before P2 becomes alive. I also should note that, although i am 10.000km far away from you, i m excited for the new product and plan to buy "Propeller 2 ES Evaluation Board" when it is available.
Comments
Having said that; I can't get others interested in the Prop because the first impression that one gets when visiting the site is Toy & Hobby.
Perhaps there is some truth to that, or more correctly, kids and education, certainly not engineers and commerce. I went to the Parallax site and took a "screenshot" of my impression, and that is what it was.
Parallax did setup a website many years ago just for the commercial engineering side, but that was a bit premature. Maybe it's not a bad idea?
IIRC, it was where all the app notes were but apart from that, a bit short on content.
I thought it was a great idea but it seemed to just blink into existence, not be actively updated or maintained then *poof* gone. Peter, I agree very much with your observation.
Ken seems to have done an amazing job keeping the parallax brand on-track with it's core educational robotics business with Chip doubling down with a 'build it and they will come' mentality.
It's unclear to me what the P2 offers to the core education market that the P1 doesn't already offer.
Without doubt the P2 is a serious personal achievement for Chip and I for one am looking forward to use it in a commercial project which outgrew the P1.
I use PASM and since anyone choosing a P2 for a commercial project almost certainly cares mostly about getting the maximum performance out of the P2's unique features I wonder if any of the high level language efforts will ever get any traction.
Maybe parallax should sponsor a PASM backend blockly interpreter for the absolute beginners and then just move them directly into PASM and then stop there. At least there would be clarity. Sometimes it's knowing what NOT to build or so said Steve Jobs.
I can remember being personally disappointed with PBASIC and such back when I was in school.
Who cares about simple and easy to use when the end result is underwhelming.
I had barely enough memory in my Stamp 1 to even get my math fair project done.
I wasn't wowed... Moved over to messing with DirectX 3.0 and the then new idea of 3D graphics accelerators on a windows PC.
Who knows what kids nowadays see: Arduino is that thing you can use to blink an LED or set colors on light strip? yaawn.
Kids are smart.
The spin language is easy enough to take an existing program as reference and add to/change it.
And when the disappointment from friends or oneself kicks in, you can jump into the compiled languages or assembly.
At the clock speeds we're talking about with P2, it can get exciting again.
So right now it's a matter of building the foundation for excitement.
I do agree documentation is lacking. No app notes.
I was hoping the same continuous iteration as on the Verilog would have been happening with the P2 documentation manual.
There are a bunch of suggested edits never touched. People get discouraged when even simple edits collect and stagnate. It's like... why bother.
Another thing I think I see we're collectively getting bogged down on is the smartpin signal path latency.
Without diagrams from the designer himself, our mental models of what we expect to happen are woefully incomplete to what we actually observe.
Being an engineer, at work my bosses never visit any of the vendor webpages I use, so who cares about website appearance? Sterile white background pages, headshots of suits, fake excellence awards, a map of all the continents with dots on it, and models smiling at each other all reaching inward at some nightmare conference table? Not necessary.
The vast majority of professional views of Parallax should be on the digikey, mouser, etc. supplier pages. If Parallax marketing doesn't botch the company description there, everything will be fine and the purchasing bureaucrats won't alert the head honchos.
I can tell you that if you toss a P1 board into a high school STEM classroom and throw-down a challenge with a bit of hands-on guidance, kids will turn into hungry little monsters and will shock the HECK out of you. Its a beautiful thing to behold. Some wont get it, but those that do... man. Now give the top 20% of these kids a P2 and you better invest in some serious safety glasses and extra fire extinguishers. Kids are amazing adopters of technology.
I do agree that with regards to the P2 doc, things seem to be lagging a bit, but I am also sensitive to the fact that this impression is enhanced because we’ve all been a fly on the wall during the development of this product. We are getting a view into places we normally would not. The P2 hasnt been officially released yet, so the grumbling about “no manuals” or “wonky tools” is premature if we use a “normal” product release as our reference point.
So I’ll withhold judgment until we see what happens, but given everything that I’ve seen about the P2 thus far (even with the occasional warts and snakes) it still makes me scream “I WANT!”, and I think the vast majority of the lurkers on here would agree.
“...If you build it, they will come...”, indeed!
Certainly the smart pin documents and examples, need work, and it is important that happens, because the Smart Pin Cells are a key difference point.
The details matter as you push the limits, and it is important to work out what the limits are.
I expect a significant easy deployment of P2, will be as a flexible peripheral, where users want easy access to the Smart Pin Cells, and cannot care less about the P2-Opcodes.
PASM certainly should be "easy to get to", but users are not going to be interested in being told they have to learn PASM, in order to use P2.
Coding these days, is done to avoid Assembler wherever practical, and only drop into ASM when you really have to. More powerful MCUs give that luxury.
For that, it is important compilers offer clean in-line assembler, and can mix HLL and PASM modules, and compilers should give good listings, showing the Assembler generated.
In many cases with HLLs, you can simply re-code things, to get better assembler, and the listings show users where they need to focus.
P2 looks like it is going to mix languages very well, so you can have PASM on one COG and even Python on another and Spin or compiled P2.FreeBASIC etc on a third....
Blockly generates C code, and I am sure getting Blockly working on P2, will be a significant focus of Ken's.
Python is already working on P2, and that uses C pathways, so getting to a functional stage for Blockly may not be too difficult. Someone has mentioned it needs P2 libraries.
Yes this is how I want to use the P2. Maybe something like the Microchip Harmony 'configurator' code generation system could generate templates for common tasks. It's something I am interested in.
One is make things simple. For example, things like Micropython will open up a lot of doors.
I'm sure there will be BASIC for P2 at some point too.
It can also self-host a development environment. I think Micropython is essentially there.
I'm sure Spin2 and some form of BASIC will follow...
But, the main thing I'm interested in is more power and pins...
The P2 will let us do really professional projects. I'm always out of pins and memory in my big P1 projects...
I started out with Basic Stamps many years ago and they taught me a lot.
But, it was hard to do real projects without adding in a lot of extra boards and chips.
P1 changed that to a large degree as you can do neat stuff with little extras.
But, large projects still need the addition of other boards and chips.
Also, you run out of pins and memory quickly.
And, there are many things that you just can't do with P1.
P2 opens a lot of doors. I think this will let me do pretty much anything I want.
So for me, the timing was close to perfect. As soon as I felt trapped with Basic Stamp, P1 came out.
Then, when I wanted more that P1, P2 started coming out.
I guess 15 could be considered “a kid”. I see the Arduino as over-pepped.(No offense to any Ardu-ites here.) I see much more in the P1, and I love the “no SERIN/SEROUT, but there are drivers for it” functionality of Spin1. With P2, the smartphone eliminate much code needs, and thus opening up memory.
Maybe I just need to get with the flow, but I like using pure Spin/I am the guy who is still wanting the p8x64b. This is what I think the P1 with expanded I/O would be called.
? BASIC is already on P2, incuded in ersmith's language suite.
Similar to FreeBASIC, looks good to me...
From my understanding, one of the reasons the P1 was not taken seriously was it's C support. Or I should say, the lack of C support on release. Some people say the the beanie didn't help. What I think really didn't help was it's unique idea of multi-core microcontroller. At the time it was just so different from what most chip manufactures were doing and what developers were searching for. I think times have changed a bit and I hope once the hardware design is LOCKED, the support software is a good beta quality, and the documentation package is done, Parallax does some serious marketing.
?11 years? on and this is STILL early in the game for the P2. Heck, we know it's not even fair to call it a P2 because that chip never got released
Once the silicon is locked (and Ken starts breathing again) I hope to see some real resources begin to develop. Obex2, Appnotes2... Datasheets and Spec sheets.
2019 and still waiting. Glad the P2 isn't TRULY vaporware, I mean I've got a p2es I can work with and wasn't forced to buy a FPGA. I'm sure I would have gotten 0 propeller work done if I had... Still, the chip we have today doesn't have the feature I needed the propeller 2 for back in 2013. CODE PROTECTION. Hopefully the P2 can get some commercial support without this. Still, I think the education market is going to have a field day with this new chip. I'm betting 2020 will be a good year for Parallax indeed.
There has to be a more modern book explaining how to use SPI/I2C peripherals: like touch screens, sound, and cell phone technologies for example...
Thanks for reminding me...
Easy to do
FastSpin really does need a better name, as it is now so much more than fastspin....
Yes, it's actually quite impressive how many languages have a pulse on P2 already.
There are two types of testing, the limit-testing can appear to have endless problems, because that is what it actively looks for
There is also a lot of compiler testing, which does a lot of silicon exercising, that does not push the limits.
When the new silicon arrives, there will be another burst, to recheck the limit tests to see what has improved/not changed/got worse.
Code protection is missing from other solutions too, and during the P2 gestation, MCUs with unique IDs have gotten very cheap, and better, and there are whole new families of ARM with security as a focus.
If code protection/security really matter, someone will drop in a MCU that is designed from the ground up that does that.
A simple secure serial number MCU is under 50c, and ARM parts are in the $2 region.
P25Q32H-SSH-IT cost $0.16 for 100 units, a 4MB (32mbit) soic8 spi flash
128 bit unique ID for each device
but also:
"The device also contains an additional 3*1024-byte security registers with OTP lock (One-Time Programmable),
can be used for purposes such as unique device serialization, system-level Electronic Serial Number (ESN) storage, locked key storage, etc. "
This was plan B but... Things change and the projects I was working on have ended, now everything I do is for personal use. It's nice being the person in charge and being able to pat myself on the back for having something instead of beating myself up for not having EVERYTHING. I'm not sure what the current volume market take on code protection is but it's no longer a problem for me, thankfully.
All the new products only have "blocky" or "C" drivers. I haven't seen SPIN or PASM code for any new products for quite a while.
From the web site, it looks like they are catering to the grade-school educational market. I'm not sure how the P2 fits into that ???
From what I've seen, the P2 is an very complicated beast. I hope there is some very detailed documentation for it when it becomes available.
Bean
Where can I sign the P8x64B petition?
That's what's been paying the bills. We haven't been much into the hobby or industrial market for a while. With the P2, we can go back into those areas. The educational market can keep going like it has been, alongside the new.
The P1 is still a made-in-heaven chip for industrial applications, and "Parallax Semiconductor" was a great idea; but IMO the execution fell a little short of what it would've taken to make it fly.
-Phil
Yes, we kind of abandoned the "Parallax Semiconductor" marketing effort. It didn't have any effect on our current customers, at least.