How about an Open Source P2-HOT??
mindrobots
Posts: 6,506
Dear Parallax,
Since many of us had so much fun playing with the P2-HOT FPGA imagines (and some of us purchased expensive DE2-115 FPGA boards just to run it - no, I'm not bitter about that anymore...much).
Wouldn't it be cool/fun/exciting to release the Verilog for that wicked hot machine as Open Source?? The design of the new P2 is so radically different, it doesn't seem like you would be jeopardizing any potential P2 IP. It would be a great learning experience and give us some truly amazing features to work with and learn from. Making those of us that are trying to play along in the FPGA world all the more prepared for the P3 project.
Just a thought to keep us busy through the Spring!
What say ye???
Since many of us had so much fun playing with the P2-HOT FPGA imagines (and some of us purchased expensive DE2-115 FPGA boards just to run it - no, I'm not bitter about that anymore...much).
Wouldn't it be cool/fun/exciting to release the Verilog for that wicked hot machine as Open Source?? The design of the new P2 is so radically different, it doesn't seem like you would be jeopardizing any potential P2 IP. It would be a great learning experience and give us some truly amazing features to work with and learn from. Making those of us that are trying to play along in the FPGA world all the more prepared for the P3 project.
Just a thought to keep us busy through the Spring!
What say ye???
Comments
I thought a never to exist in silicon P2-HOT in the hand was better than nothing in the bush for a while. (Wow, that just didn't word-smythe well at all!
There is P1V there now, if you want to play on FPGA boards, and hopefully even small boards will be able to test New P2 peripherals to get maximum coal-face experience and get software interaction / user control thrashed out so the stuff is solid before being 'cast in Silicon' .
No hints on a Verilog image release time frame for Cooler-P2-V0.1 ?
I missed the announcement about the New P2 peripherals.....what are those going to be? Where are all these small boards coming from, ready for testing? Parallax? Isn't that a diversion of effort?
We don't want him spending his time here on the forums, either.
Ken Gracey
So I'm hoping and expecting a P2 image within the next 3 months based on what Chip and Ken have said recently.
EDIT: Speaking of which, what's the status of Parallax's own FPGA board? I haven't purchased any such boards as yet.
Ken's "schedule it" - http://forums.parallax.com/showthread.php/159884-Appreciation-and-Thanks?p=1313096#post1313096
I'm pretty confident all the FPGA work is prior the six month countdown.
EDIT: Maybe I misunderstand what you mean by "all the FPGA work". If you mean a complete FPGA image that implements all of the functionality, then I agree. However there will be additional FPGA work to fix problems that are found during testing. The additional FPGA work needs to be limited to bug fixes. If there is additional FPGA work to add new features the P2 will never be completed, and I don't think we'll ever see silicon. Parallax needs to put a stake in the ground and freeze all the features of the P2 if they ever want to actually produce the thing.
I was looking for a better description of the intent and purpose of the "get together" in Rocklin.
I did find this. In which Ken tells us:
So "schedule it" has three constraints:
- Propeller 2 code is running in full on the Parallax FPGA 1-2-3 Development Board
- Manual layout of Propeller 2 is done by Treehouse Designs
- Availability of the Propeller 1-2-3 FPGA board.
From what I recall, I see the event as a reception after the marriage of the FPGA image to the new board. The FPGA image and the new 1-2-3 board being the happy couple (boards for sale in the lobby).
I'm unsure if there will be an FPGA release made for testing on DE2 or DE0 or anything else prior to ALL of those 3 milestones being complete.
When it happens, it happens and whoever is around at the time can party like it's 1999!
All in all I feel you would be better off fetching a CPU design from opencores.org or perhaps the RISC V design.
And hey, there is the P1 HDL to play with.
That's before we get to the point of how much time it would waste for Parallax.
What would happen if parallax publish the details and someone fills the design into the patent office on the same day? They don't even need to fill the patent for the whole IC, just one small thing (like smartpins / smartio or anything else).
Or what would happen if they publish the opcode list and someone gets a trademark, or whatever be needed to prevent parallax to sell the design?
When Parallax were doing the layout by themselves, they could modify the design at will. But now that they get the design from Treehouse Designs and I think they don't have a second chance without paying big $$$.
Have we been able to document, completely understand, and verify the P1V code? How is it possible that we can help Parallax to test the P2 when we have not done that yet for the P1v? Does Parallax or the design house need any help from us? I think not.
But said that those seventy-five $500 fpga boards have not been made to be wasted, so someday maybe we will have the fpga image.
As for the FPGA boards, I think most people here will be testing with their applications. Documentation and understanding and bugs found will all be a side effect of questions being raised when trying to make it work.
Prior art no longer matters, the recent changes to patent law 'reform' moved from the "First to Invent" where prior art was considered to "First to File". Thus making it easier for Apple, and the other big corporations, to see something they like and patent it out from under smaller innovators.
In any case, this is how patent law has worked in other countries for a very long time already. Except for one difference - read about it on Wikipedia.
That is exactly the problem with moving to First to File. The US PTO does not look at prior art hardly at all, if at all when dealing with an Apple/MS type case. So if they see something they like, that has NOT been patented yet, they just file on it. They have the budgets and legal people on staff to do it.
You have to realize that in this country (USA) the only marginal benefit you get as a small company for patenting an idea is limited protection from a big corporation from patenting and suing you over your own invention. The Chinese will just knock you off regardless and if you spend the $$$ to try to block them, and get anyone to issue an injunction to stop them which is almost impossible, they simply rebrand it through another shell company and keep on selling. A big company can, and will, bankrupt any small company who tries to go after them in court especially if they hold the cards (patent).
I know that from first hand experience. Was Director of Engineering for a manufacturing company for 14 years and we had a number of patents, including some I invented. We had one really good one for a product that we showed Sears (big retailer here) and they wanted to use it in the Christmas specials. Only problem was that they wanted it for 25% cheaper than we priced it at. When our Owner pointed out that it was a unique item covered by our patent the Executive Buyer actually LAUGHED at him. Turns out they had already shopped the product samples we had given them to china and had 3 quotes direct from there. He then proceeded to tell us that "your patent means nothing to us as our legal department's budget is more than you make in a year. Give me the price I want or I buy elsewhere, sue me and I will drop you as a vendor and legal will keep the case in court until you are bankrupt."
THAT is what our patent system has boiled down to here. You spend a bunch of money for one, and then the deck is stacked against you in trying to enforce it.
First to File simply makes it even easier for an Apple/MS to ram a questionable patent through. Which is why they pushed so hard (and paid a lot of $$$ to politicians) to get it done. Prior art means nothing if Apple claims they invented it 'in parallel' in the same time frame as someone else THEY WIN right off the bat with First to File since at that point they have the patent and were clearly First to File.
On the flip side, in niches, patent pending has a little more teeth.
The stuff I'm working on is novel mechanical. They do work considerably harder than the do the software stuff. We've seen claims from the 1900's cited frequently.
First to file is a land grab in tech patents. For us, and mechanical, it means you can't even let a whiff leave the building. Prior art helped with this considerably. Since the land grab has happened to a large degree in the mechanical space, novel is hard to prove, and since so much is happening in tech, it also means getting less attention and or it has been harder to source really great people too.
Prior art should still work if the evidence existed prior to the patent filing. Which obviously occurs when the source code is freely available.
Have you tried using that as a defense in court?
Prior art in any published form, or disclosed in previously granted patent, is still a principal part of patent discovery. FITF doesn't change that. Applicants are still expected to be thorough and disclose their discovery to the fullest extent, though sometimes a patent is granted that should not have been given the existence of prior art. But that was also a problem with first to invent. The patent office didn't, and still doesn't, keep tabs on forums like this one.
One major change in favor of the independent inventor is the fee structure. It cost very little for a small entity to initially file. The playing field is much more level now.
But all this begs the question: what about the P2 is patentable, and would they care? I see Parallax as a company that places far more value in customer support and education of its products. It's not just a simple question of someone else seeing what the P2 can do, and realizing it's not yet filed, send in a claim. It (still) doesn't work that way.
For providing a dependable date-stamp of the original webpage, http://archive.org/web/web.php is a good start. How much of the original content gets mirrored might depend on how the content is managed though. I'm no web expert.
Here's an example - http://web.archive.org/web/20130713003048/http://forums.parallax.com/showthread.php/125543-Propeller-II-update-BLOG
From the link you can see the snapshot was taken July 2013 and that dates are present on the topic posts. That's pretty good data to work with if one was trying to piece together believable evidence.
https://developer.valvesoftware.com/wiki/Valve_Time
Excellent!
"Coincident with the Rapture" as a target date almost got Pepsi on the keyboard!
I'll have to work that into one of my project schedules and see if anyone notices.