Shop OBEX P1 Docs P2 Docs Learn Events
I want a P2! And how are we going to get these to you? - Page 16 — Parallax Forums

I want a P2! And how are we going to get these to you?

11213141618

Comments

  • What development tools are currently available? I realize all of this is in a state of flux, but I haven't been working with any of the emulators so this is "square one" (or "square zero" :)) start for me if I purchase one.

  • Has an estimated shipping date been decided?

    Thanks!
  • Clock LoopClock Loop Posts: 2,069
    edited 2018-12-03 19:12
    Not buying this right now is one of the hardest things I've ever done. I hope these get in the right hands so progress can be quick!

    I locked my ccard in a box with a timer on it. And the timer is dated 2019, now I couldn't order one, even if I wanted.

    I wanted one just for the novely factor of owning a chip that is 27,000$

    :hushed:

    Some of us are not useful in pasm, and would just take up a board to "play" with the code you guys give us and teach us.
    I have my hands full with dev stuff in QT5, simpleide, and the rest of the "toys" you guys have given us for free..

    I can't even keep up with the p1 and all the various options and tools available.

    wmosscrop wrote: »
    What development tools are currently available? I realize all of this is in a state of flux, but I haven't been working with any of the emulators so this is "square one" (or "square zero" :)) start for me if I purchase one.

    It truly is in flux right now...
    If you work in pasm then you REALLY need to start reading other forum posts here... https://forums.parallax.com/discussion/164187/fastspin-compiler-for-p2/p1
    If you don't work in pasm or asm then you might want to wait till a higher development language is made for it.
  • ColeyColey Posts: 1,108
    potatohead wrote: »
    Got one on the way.

    Thank you Ken, Chip, everyone. It is amazing to place an order with "P2" in it.

    Got two on the way, Baggers and me are in!
  • Clock Loop wrote: »
    It truly is in flux right now...
    If you work in pasm then you REALLY need to start reading other forum posts here... https://forums.parallax.com/discussion/164187/fastspin-compiler-for-p2/p1
    If you don't work in pasm or asm then you might want to wait till a higher development language is made for it.
    I work in pasm, but usually do a prototype in Spin... that's all I would need for now (Spin/pasm)

  • K2K2 Posts: 691
    My last thought before dozing off last night was, "Oh Smile! I didn't check to see if P2 boards were available yet!" Then sheer weariness overtook me. (It has been a busy week.) Thankfully it wasn't too late today. I simply had to have one - at almost any price. The P1 was 100x better than I ever though it would be. They still power all manner of devices that surround me. The latest project on my workbench is also Propeller-powered and uses hundreds of electrolytic capacitors.
    (But that's all I'm saying about it.)

    Anyway, I expect to have this new P2 running full-tilt on all eight cogs the day it arrives, executing a set of algorithms I took from Donald Knuth and have implemented on many architectures as something of a benchmark suite.
  • cgraceycgracey Posts: 14,131
    edited 2018-12-05 05:43
    K2 wrote: »
    My last thought before dozing off last night was, "Oh Smile! I didn't check to see if P2 boards were available yet!" Then sheer weariness overtook me. (It has been a busy week.) Thankfully it wasn't too late today. I simply had to have one - at almost any price. The P1 was 100x better than I ever though it would be. They still power all manner of devices that surround me. The latest project on my workbench is also Propeller-powered and uses hundreds of electrolytic capacitors.
    (But that's all I'm saying about it.)

    Anyway, I expect to have this new P2 running full-tilt on all eight cogs the day it arrives, executing a set of algorithms I took from Donald Knuth and have implemented on many architectures as something of a benchmark suite.

    Great, K2! I feel like my head's going to explode, it's so full of tantalizing ideas revolving around Prop2. I have to make myself calm down, just to get work done. If you found Prop1 useful, your going to feel like you've been set free with Prop2. It's like a new personal spaceship to zip around in and explore ideas with.
  • evanhevanh Posts: 15,091
    Lots of caps .... /me jumps to conclusion - building a mini Z-machine!

  • K2K2 Posts: 691
    evanh wrote: »
    Lots of caps .... /me jumps to conclusion - building a mini Z-machine!
    That's a very good guess. Were it not for the inductive reactance of electrolytics, it might even be correct. You are certainly in the right postal code. :)

  • Clock Loop wrote: »
    It truly is in flux right now...
    If you work in pasm then you REALLY need to start reading other forum posts here... https://forums.parallax.com/discussion/164187/fastspin-compiler-for-p2/p1
    If you don't work in pasm or asm then you might want to wait till a higher development language is made for it.

    Perhaps you missed it, but fastspin doesn't only support P2 PASM, it supports all of Spin (plus some proposed Spin2 features), and also compiles a version of BASIC. Granted, one might argue about how "high level" Spin or BASIC are, but they are higher level than PASM :).

  • Hi,

    I didn't know that the P2 Eval boards were already available (and, apparently, selling fast). Thanks God I had that sense to look out for them. I still managed to get my hands on one.

    I'll probably make a review in January, on my blog.

    Are there any C toolchains or compilers ready for the P2? I wanted to test my prime number calculation program. That would be a solid benchmark. I think I could do a YouTube video as well.

    Kind regards, Samuel Lourenço
  • samuell wrote: »
    Hi,

    I didn't know that the P2 Eval boards were already available (and, apparently, selling fast). Thanks God I had that sense to look out for them. I still managed to get my hands on one.

    I'll probably make a review in January, on my blog.

    Are there any C toolchains or compilers ready for the P2? I wanted to test my prime number calculation program. That would be a solid benchmark. I think I could do a YouTube video as well.

    Kind regards, Samuel Lourenço

    "Are there any ..." yes
    They are not GCC. See lots of good info here: https://forums.parallax.com/discussion/169023/status-of-p2-c-c-compiler#latest
    In short: checkout p2gcc or the spin2cpp suite of tools if you're impatient.
  • samuellsamuell Posts: 554
    edited 2018-12-05 20:04
    DavidZemon wrote: »
    samuell wrote: »
    Hi,

    I didn't know that the P2 Eval boards were already available (and, apparently, selling fast). Thanks God I had that sense to look out for them. I still managed to get my hands on one.

    I'll probably make a review in January, on my blog.

    Are there any C toolchains or compilers ready for the P2? I wanted to test my prime number calculation program. That would be a solid benchmark. I think I could do a YouTube video as well.

    Kind regards, Samuel Lourenço

    "Are there any ..." yes
    They are not GCC. See lots of good info here: https://forums.parallax.com/discussion/169023/status-of-p2-c-c-compiler#latest
    In short: checkout p2gcc or the spin2cpp suite of tools if you're impatient.
    Thanks David. Now there are only two (three) obstacles to overcome:
    - Learning how to use github to obtain the code for P2gcc and compile the tool for Linux (Debian).
    - How to compile and program?

    I usually use SimpleIDE, which is simple to use. I've compiled programs using GCC before, many many times. But how to transfer them via serial?

    I've checked the spin2cpp tool, but I think it is the p2gcc that I need.

    Kind regards, Samuel Lourenço
  • samuell wrote: »
    Hi,

    I didn't know that the P2 Eval boards were already available (and, apparently, selling fast). Thanks God I had that sense to look out for them. I still managed to get my hands on one.

    I'll probably make a review in January, on my blog.

    Are there any C toolchains or compilers ready for the P2? I wanted to test my prime number calculation program. That would be a solid benchmark. I think I could do a YouTube video as well.

    Kind regards, Samuel Lourenço

    Since I don't work for Parallax, I may be out of line but I don't think that Ken wanted to have public reviews written about the P2 ES since it is not the finished version, and still has some aspects that need to be revised in a future production. The P2 ES evaluation boards are primarily to increase the number of testers and tools building.

    To quote Ken : https://forums.parallax.com/discussion/comment/1455103/#Comment_1455103 "We won't be making any public hoopla and this is a quiet release for a couple of reasons you already appreciate. First, we don't want potential customers/reviewers to get the impression we're making an industry release, causing them to expect multi-platform software tools, Spin code examples, a staff of FAEs on board to answer questions, etc."
  • samuell wrote: »
    Thanks David. Now there are only two (three) obstacles to overcome:
    - Learning how to use github to obtain the code for P2gcc and compile the tool for Linux (Debian).
    - How to compile and program?

    I usually use SimpleIDE, which is simple to use. I've compiled programs using GCC before, many many times. But how to transfer them via serial?

    I've checked the spin2cpp tool, but I think it is the p2gcc that I need.

    You could try spin2gui, which is a *very* simple IDE for both P1 and P2. It supports Spin, BASIC, and a very limited dialect of C (I'm still working on the C compiler, which is not based on GCC). Since it works for both P1 and P2 you can do some testing of it now. The reason the compiler is not based on GCC is that it's intended to allow all 4 languages (PASM, Spin, BASIC, and C) to call each other easily. Also, I got very burned out on working on GCC (other people are taking that over now) and wanted to try out something different.

    It won't compile your prime numbers code yet, because it's missing a lot of the standard library (fgets, sqrt, and so on). You could probably get a simpler version of your code to work though -- it does support printf.

    spin2gui is located at:
    https://github.com/totalspectrum/spin2gui/releases

    Eric
  • K2K2 Posts: 691
    cgracey wrote: »
    If you found Prop1 useful, your going to feel like you've been set free with Prop2. It's like a new personal spaceship to zip around in and explore ideas with.
    I love the simile. Still, to be better than the P1 in terms of time-to-a-working-prototype (and utter precision and dependability of the same) would be quite an accomplishment. In those areas, you're just competing with yourself. No one else is close.


  • This is my first post on these forums.

    I am a friend of Cluso99 for 30+ years, and discovered the P2 from him.

    I am quite excited to (hopefully) receive the 2 x prop2 boards that i ordered.

    There are lots of projects that i would like to do with them, such as
    o An implementation of Apple Pascal (UCSD Pascal II.0)
    o An implementation of Oberon 07 (which currently runs on an FPGA spartan 9 board using a 32 bit risc cpu via verilog)
    I suspect that it would be easy to modify the Oberon 07 compiler to produce P2 machine code as these cpu's are quite
    similar.

    It is also likely that i will contribute to a number of the existing opensource compiler/assembler/spin projects currently
    underway.

    regards Mark Fawcett
  • TubularTubular Posts: 4,620
    edited 2018-12-06 02:38
    Welcome. Mark. You're going to have a lot of fun with Prop2. As are the rest of us.

    Its really interesting to see the range of things people are going to tackle with P2 in hand
  • jmgjmg Posts: 15,140
    markmpx wrote: »
    This is my first post on these forums.

    I am a friend of Cluso99 for 30+ years, and discovered the P2 from him.

    I am quite excited to (hopefully) receive the 2 x prop2 boards that i ordered.

    There are lots of projects that i would like to do with them, such as
    o An implementation of Apple Pascal (UCSD Pascal II.0)
    o An implementation of Oberon 07 (which currently runs on an FPGA spartan 9 board using a 32 bit risc cpu via verilog)
    I suspect that it would be easy to modify the Oberon 07 compiler to produce P2 machine code as these cpu's are quite
    similar.
    There is also work that emulates RISC MCUs on P2 - not as fast as native, but it could get you a working base to improve on ?

    Would you include Project Oberon in that work ?
    Porting the Verilog peripherals in Project Oberon to P2 Smart Pin Cells+COG would be somewhat more effort, but the info I find suggests the 1024*768 monochrome display would easily fit into P2 memory.

    There is also a BASIC compiler in the works for P2 ( broadly follows FreeBASIC subset ) see http://forums.parallax.com/discussion/164187/fastspin-compiler-for-p2/p1
    You may be able to take some of the code generation work from that, and splice into the Oberon Compiler ? - and use the assembler, to process your P2 Asm generated code.
  • jmg wrote: »
    There is also work that emulates RISC MCUs on P2 - not as fast as native, but it could get you a working base to improve on ?

    Would you include Project Oberon in that work ?
    Porting the Verilog peripherals in Project Oberon to P2 Smart Pin Cells+COG would be somewhat more effort, but the info I find suggests the 1024*768 monochrome display would easily fit into P2 memory.

    There is also a BASIC compiler in the works for P2 ( broadly follows FreeBASIC subset ) see http://forums.parallax.com/discussion/164187/fastspin-compiler-for-p2/p1
    You may be able to take some of the code generation work from that, and splice into the Oberon Compiler ? - and use the assembler, to process your P2 Asm generated code.

    Jmg,

    Firstly, i am surprised to find others familiar with Project Oberon ! (http://www.projectoberon.com/ and wirth's pages at ETH).
    I have experimented on project Oberon using a pair of "Pepino Oberon" (Spartan 6-LX25) boards from Saanlima Electronics.
    These boards contain 1 Megabyte of memory, the current P2 has only half of that amount.

    Project Oberon comprises a number of hardware items, the 32 bit RISC cpu, the PS2 keyboard/mouse interface,
    the bit mapped graphics VGA display (1024 x 768) and the SD card interface, (and perhaps the networking interface also)

    Software for Project Oberon includes the base operating system, the compiler, and various utilities like the text editor, graphical editor etc etc

    I was hoping initially to port the Oberon compiler to the P2 processor with a minimum operating system runtime, so that
    i could produce standalone code for P2 boards. I was not intending to port the rest of Project Oberon to the P2, as it has
    limited usefulness for myself. After i have ported the compiler, it should be fairly easy for others to complete porting the
    rest of the project including replacing the keyboard/vga/sd card interfaces etc to be P2 compatible. The biggest problem
    is likely to be that the current P2 boards have only half of the memory. (Perhaps there will be P2 variants at a later date
    with the full 1 megabyte) (yes i know, some of the FPGA boards have 1MB and 16 cogs)
    I plan provide all the source code on github, so others can benefit.

    The way that i expect to bootstrap the Oberon compiler may surprise you :smile:

    I expect to port Wirth's Oberon compiler to Nim and then modify it to produce P2 object code directly. (google nim-lang)
    My preferred development environment is Ubuntu, but Nim is available on windows, mac etc, (ie: almost anything)
    The Nim compiler produces C/C++ source code and a C compiler produces the object code, so effectively the Oberon P2
    compiler is a C program. I expect to port a minimal set of the Oberon runtime and i expect i will have to write a linker
    (as the oberon system does this dynamically at runtime).

    Given that the above requires an intimate knowledge of the P2 processor, i expect to write an interpreter for p-code
    as produced by the Apple 2 pascal compiler (very similar to UCSD pascal version II.0) initially and use this as learning
    exercise. This is task i am have done numerous times.

    So...

    My main interest in the P2 is PASM, and C.
    Perhaps one day it would also be possible to write code for the P2 in Nim (if the P2 C compilers are good enough)

    Sorry for such a long post
  • jmgjmg Posts: 15,140
    Maybe a moderator can fork this into a separate thread, from your first post, about Apple Pascal and Oberon 07, to keep it separate from P2 supply lines... ?

  • Cluso99Cluso99 Posts: 18,066
    Hi Mark.
    Welcome to P2 - you certainly will have fun with it!
  • samuellsamuell Posts: 554
    edited 2018-12-06 14:01
    twm47099 wrote: »
    samuell wrote: »
    Hi,

    I didn't know that the P2 Eval boards were already available (and, apparently, selling fast). Thanks God I had that sense to look out for them. I still managed to get my hands on one.

    I'll probably make a review in January, on my blog.

    Are there any C toolchains or compilers ready for the P2? I wanted to test my prime number calculation program. That would be a solid benchmark. I think I could do a YouTube video as well.

    Kind regards, Samuel Lourenço

    Since I don't work for Parallax, I may be out of line but I don't think that Ken wanted to have public reviews written about the P2 ES since it is not the finished version, and still has some aspects that need to be revised in a future production. The P2 ES evaluation boards are primarily to increase the number of testers and tools building.

    To quote Ken : https://forums.parallax.com/discussion/comment/1455103/#Comment_1455103 "We won't be making any public hoopla and this is a quiet release for a couple of reasons you already appreciate. First, we don't want potential customers/reviewers to get the impression we're making an industry release, causing them to expect multi-platform software tools, Spin code examples, a staff of FAEs on board to answer questions, etc."
    Sorry, I didn't know that. Thanks for informing me! It seems I have to ask Ken.

    Anyway, I think it is good and free publicity. I'll mention that the P2 is still under development, and that it is just a pre-release. I like to publicize the P1, and sure I would love to be the first in Portugal to announce the P2, because of the quality of the chip, because of how it is easy to use and also due to the fact that the toolchains are not proprietary (for instance, Michochip requires you to pay if you want a more optimized compiler).
    ersmith wrote: »
    samuell wrote: »
    Thanks David. Now there are only two (three) obstacles to overcome:
    - Learning how to use github to obtain the code for P2gcc and compile the tool for Linux (Debian).
    - How to compile and program?

    I usually use SimpleIDE, which is simple to use. I've compiled programs using GCC before, many many times. But how to transfer them via serial?

    I've checked the spin2cpp tool, but I think it is the p2gcc that I need.

    You could try spin2gui, which is a *very* simple IDE for both P1 and P2. It supports Spin, BASIC, and a very limited dialect of C (I'm still working on the C compiler, which is not based on GCC). Since it works for both P1 and P2 you can do some testing of it now. The reason the compiler is not based on GCC is that it's intended to allow all 4 languages (PASM, Spin, BASIC, and C) to call each other easily. Also, I got very burned out on working on GCC (other people are taking that over now) and wanted to try out something different.

    It won't compile your prime numbers code yet, because it's missing a lot of the standard library (fgets, sqrt, and so on). You could probably get a simpler version of your code to work though -- it does support printf.

    spin2gui is located at:
    https://github.com/totalspectrum/spin2gui/releases

    Eric
    Well, the lack of an sqrt function can be compensated, at cost of efficiency. No problems there. I can either "mis"calculate the sqrt first by successive approximations, or use a for loop using a stop condition where the iterative variable is multiplied by itself (less efficient). The first approach doesn't have to be exact. I only need to calculate a number that exceeds the sqrt result by a bit. The second approach will be slower.

    Kind regards, Samuel Lourenço
  • __red____red__ Posts: 470
    edited 2018-12-06 19:38
    samuell wrote: »
    Well, the lack of an sqrt function can be compensated, at cost of efficiency. No problems there.

    To be clear, the P2 has an assembly function for calculating sqrts. Support from compilers to be added one would presume.

    QSQRT {#}D,{#}S

    If your compiler supports it, you could always inline the ASM.

    @ersmith would have orders of magnitude more knowledge on the feasibility of this than I.
  • __red__ wrote: »
    samuell wrote: »
    Well, the lack of an sqrt function can be compensated, at cost of efficiency. No problems there.

    To be clear, the P2 has an assembly function for calculating sqrts. Support from compilers to be added one would presume.

    QSQRT {#}D,{#}S

    If your compiler supports it, you could always inline the ASM.

    @ersmith would have orders of magnitude more knowledge on the feasibility of this than I.

    What's the precision of that instruction? Even though the P1 has those log and antilog tables, PropGCC didn't use them by default because they weren't as precise as what most people would expect from the log functions in the math library. I wouldn't be surprised if the same holds true for any built-in sqrt instruction. But it sure would be cool if it is high enough precision!
  • Clock LoopClock Loop Posts: 2,069
    edited 2018-12-06 20:27
    THE COUNTDOWN BEGINS!

    In Stock: 20
  • __red__ wrote: »
    samuell wrote: »
    Well, the lack of an sqrt function can be compensated, at cost of efficiency. No problems there.

    To be clear, the P2 has an assembly function for calculating sqrts. Support from compilers to be added one would presume.

    QSQRT {#}D,{#}S

    Yes, and the BASIC and Spin parts of fastspin already have that hooked up to their respective functions. I just haven't got to the C runtime library yet. It should come soon though.


  • cgraceycgracey Posts: 14,131
    edited 2018-12-06 21:13
    QSQRT {#}D,{#}S takes a 64-bit unsigned value and computes its 32-bit root.
  • Clock Loop wrote: »
    THE COUNTDOWN BEGINS!

    In Stock: 20

    Speaking of - do we have an ETA for shipping for any of them yet?
    (I know you guys have a ton of QA to do with it being a completely new board - so not hurrying up, just wanting to schedule some vacation / playtime) ;-)


  • localrogerlocalroger Posts: 3,451
    edited 2018-12-07 01:20
    samuell wrote: »
    twm47099 wrote: »
    Anyway, I think it is good and free publicity. I'll mention that the P2 is still under development, and that it is just a pre-release. I like to publicize the P1, and sure I would love to be the first in Portugal to announce the P2, because of the quality of the chip, because of how it is easy to use and also due to the fact that the toolchains are not proprietary (for instance, Michochip requires you to pay if you want a more optimized compiler).

    The reason they don't want reviews of this release is that it is known to be a flawed chip -- it has known bugs and possibly more bugs to be discovered. But it mostly works, and is better than the FPGA at doing a lot of Propellery things, so they offered to have some eval boards made up for those of us who wanted to drive real silicon. But this is in no way an actual P2 as it will be released when it is better debugged and has all the little extra features and tweaks that will go into the next revision.
Sign In or Register to comment.