Curiously, one gets more glowing reports about X___ here than on X___ sites. Here are some excerpts from posts made last week about X___:
"When starting from scratch be prepared for a long haul of wading through treacle. Don't expect to be productive with them too quickly, you will die of frustration."
" X___ has all the hall marks of an academic startup, phenominaly clever, but they need to understand the needs of their potential customers a little better."
"If you can't tell from the above, having spent a disproportionate amount of time on it I am quite seriously disappointed. I hope sincerely that X___ can turn this around but will commit no more time to them until they have addressed the issue with their docs and tools."
I post this only because complexity issues are such a huge turn-off. The few times I've ventured into X-territory, I've been turned off very quickly. I'd have to have a desperate need for 400 MIPS before I'd spend a tuppence on X.
The Propeller won't work in that application, of course. The Prop II might, but I doubt it.
Really? Prop can do fast encoder reading, fast PWM generation, fast co-ordinate rotation (for vector control), PID, motion control profile generation. I would have thought that a single prop 1 could do a lot of it, and a second would help with the comms.
I've driven a brushless motor with a propeller (though not with a complex control algorithm) so I'm surprised you say this.
I would enlighten them but it would seem bad form :P
Really? Prop can do fast encoder reading, fast PWM generation, fast co-ordinate rotation (for vector control), PID, motion control profile generation. I would have thought that a single prop 1 could do a lot of it, and a second would help with the comms.
I've driven a brushless motor with a propeller (though not with a complex control algorithm) so I'm surprised you say this.
Graham
Can you get the same performance as in that application? I just don't see it.
Can you get the same performance as in that application? I just don't see it.
Probably can't emulate that board but the issue is can the propeller be used in the application. Can't get 500 MIPS for control or 1000MIPS for comms but I'm not sure it is needed? I'm happy to be enlightened, I'm not sure.
I think I have to disagree here.
I've been pleasantly surprised at how easy going pretty much everyone here has been.
I am on another forum, a freakish AVR one. People get wound up there for the same reasons as here: a competing product almost always seems to better in every way, and seems to be brought up at every opportunity...
There as here, higher ups have had to request people stay on topic as it concerns that site's primary focus.
I get the sense here that an occasional reference to PIC or AVR or ARM, either pro or con is not a hanging offense. I expect that here as in over there, it is only when it starts to take over a thread, or is becoming repetitive that people get annoyed.
@koehler - Stick around, it gets very interesting at times.
@everyone - It truly surprises me that Parallax is keeping such a low profile on this popular thread. I would have thought there would have been a lot more input.
1. This forum is congenial enough that people are not afraid to come here with their problems.
2. Our users have fewer proprietary restrictions on what they can ask in a public forum.
3. Our users are working on more ambitious projects than those other folks.
4. The other chip is so complicated to get started with that its users don't even know what to ask.
5. The Propeller attracts more users with less hardware or programming experience.
6. Or ... ?
The original poster got more than he asked for, don't think it was exactly what he asked for, but he sure got a lot of it.
C.W.
That is a understatement.
I asked the question because I want the limited time I have to learn a new uC to be put to good use; I also wanted to know if Spin and the Prop is worth the time. I have had my Prop for 2 days nows.
Within 24 hours and 3 tutorials I was able to reto-fit a OBJ and make a working PIR Intruder alarm with PST terminal messaging.
I have also written a 11 line program (a hack of the propeller help tutorials), which would (IMO) take probably a hundred lines (and a heck-of-alot of math) if you tried the same thing withoutr independent cogs operating in parallel.
I was on the fence between Arduino UNO and the Prop. Once I saw the Gadget Gangster usb Prop (which is priced right between a Arduino uno and a mega) I decided to take a chance.
So far, I am impressed with what the Prop can do.
(This will get me into trouble with some) I also think SPIN is a very decent language with some really nice features. It will take some time to become fluent, but Parallax has tons of resources.
I only mention it when people claim that the Propeller is unique in terms of parallel processing and deterministic operation. I suppose that the message will get across eventually.
Given the rate of activity on the forum and the constant flow of new members the message will never get across fully. There is inevitably going to be a string of occasions when you will feel the need to correct people on this issue, the problem is that everyone else has heard it all before ad nauseum.
Is this a noble quest for truth and justice or pedantry/OCD?
I asked the question because I want the limited time I have to learn a new uC to be put to good use; I also wanted to know if Spin and the Prop is worth the time. I have had my Prop for 2 days nows.
Within 24 hours and 3 tutorials I was able to reto-fit a OBJ and make a working PIR Intruder alarm with PST terminal messaging.
13 pages of discussion later, and I think that little comment answers it best of all. Here we have someone completely new to the propeller and within a short time they have managed to get something cool working. Great stuff!
To me, this says that the big secret of the propellers success may not not just be its parallel structure, but it might also be the Tutorials and the code in the Obex. And this forum.
With you being the initial creator of this interesting thread, I am glad to hear that you are doing so well after having your Prop for only 2 days. From what you described, it sounds like you picked it up rather quickly. Keep up the good work.
Is this a noble quest for truth and justice or pedantry/OCD?
Graham
It's neither - it's a simple misunderstanding. The "determinism" issue seems to arise frequently in these forums, because the Propeller has it as a fundamental attribute of the chip design, whereas other chips do not. With those chips you generally have to do something - often something messy and/or complex - to achieve it.
For a discussion about this specific to the X___ chip, including an excellent summary of a discussion on precisely this issue with David May (CTO of X___) see Heater's post on this page (about four from the top).
Ross.
P.S. I don't mean to derail the thread - in fact I'd like to see it get back to the original topic. Determinism as it relates to the Propeller, X___ and NXP can be further discussed in this thread.
And I might respectfully disagree with Leon.
1) ARM chips use too much power -they get hot, and any chip that gets hot makes it more difficult in battery and solar applications.
2) With microchip uCs there is an impression that a handful of bytes are enough memory, and that a few kilobytes is more than enough.
The propeller uses a tiny amount of power and has lots of memory.
They come in a wide range of memory sizes - the NXP LPC2178 has 512 kB of program memory and 64 kB of SRAM and does 100 MIPS. It doesn't even get warm. I just powered up a $30 LPCXpresso board to make sure.
I think the OP answered his own question in the end. Whatever else you say about the propeller it is a great chip to develop on and for those that really want to deal with the real world in real time the multi-core methodology is fantastic.
Never have so many had so much fun with so little in such a short a space of time.
A couple of loosely related thoughts on the thread.
I want to address the technical issues above, but it takes time and I don't want to convey what is going to happen in regards to specific items until they're underway. More than once I've talked about something that happened differently.
While reading it I've really wanted to reply but we're working on many projects for all of you. It would be fantastic to bring you inside of Parallax, especially considering you are Parallax - or at least that's the way it is supposed to be. This isn't a place where we go to work just to make our salary, but to work with the technology and people we enjoy a whole bunch. As a result the output from Parallax will often be a mix of what you want but with a few surprises mixed in along the way. And the traits you dislike may disappear and be replaced with better things along the way. No company makes a single chip and thinks they've got it all right; your feedback is part of our next design, tools, code examples and presentation. The research you provide on these forums has an impact on what we do.
Somewhere above there was discussion about Parallax Semiconductor.
Please keep in mind that no changes occur overnight; we're a small company with a team of 44 people. At this moment eight of us are working to develop the system to properly support this effort, including: FAEs and an easy way to reach them; a Drupal web site, demonstrations of customer success, AppNotes and distribution. Several of you are involved in a customer focus group and a few others have joined from other corners of the industry to bring your expertise.
When we mention that we'll be commercializing the Propeller as Parallax Semiconductor it's because our commercial customers need that from the company. Engineers who work for small/medium companies and wish to specify the Propeller demand a different style from our company and they're elated that we'll give them the validation to take to their managers. Entrepreneurs don't seem to mind either way - and like hippy said "are you not going to eat at your favorite restaurant just because it isn't a chain?"
The approach to Parallax Semiconductor is simple: make it easy, showing high-level integration of easy-to-use "gold-plated" objects that accompany AppNotes; and quickly connect you with an FAE via Skype, phone, or in-person as you desire to talk about your project. We'll help you past any hurdles and see you through to completion.
In our AppNotes you'll see some traditional microcontroller applications but many that need the Prop's multi-core design.
Chip design cycles will speed up at Parallax, too. We're in the process of porting Prop2 from AHDL to VHDL so we can skip the schematic layout and go straight to die fabrication. The tools for doing this are very expensive (did I hear $1M?) so we've retained a firm specifically for this purpose. Even though we'll be quicker we still won't be releasing new processors every six months.
And because this is your company, we are using your investment to meet these goals. Parallax has no venture funding and the company reinvests more than most CPAs would imagine. Venture-funded companies may have more waste in decision making as a result of pressure to deliver, including more staff turnover, and don't easily generate customer loyalty as a result. I think hobbyists and commercial users would prefer our method. When we spend your money we must do it in a way that doesn't create unnecessary risk for you, which is why you've been able to count on us as a steady supplier for 20+ years. Now that we have our own chip you can count on us for at least 20 more years.
The target number is Propellers we should sell is peanuts to most semiconductor suppliers but it is more than we're selling today. As engineers you already know something about silicon margins compared to robot kits, BASIC Stamps and easy-to-use sensors.
From a sales perspective Parallax's revenue is equally derived from hobby, education and commercial customers yet the company looks like it serves the first two groups. We are very sensitive to the fact that while we undertake Parallax Semiconductor we must increase the product base and support for our education/hobby customers more than before. The sensors, motor controllers, WiFi, PropBOEs, educational kits and selection of simply useful designs from Parallax will continue at the same rate.
I just wanted to reaffirm that we're not sitting on our laurels around here - we've read and heard first-hand every statement from all of you and many of them are being built into our products. Parallax is an active place with enthusiasm and energy, with much investment in the future. We're more excited than ever about the changes that lie ahead!
Alas and alack, Propeller by itself's pretty sophisticated compared to the other microcontrollers being sold everywhere.
It's also handy for testing the other thing such as testing a dead switchmode power supply (function generator testing), as well as
testing few other things, like sticky logics (all of our favorite 74s family logics we love).
SPIN's pretty slow - but it's okay, as Chip haven't designed anything as serious as Propeller before then - everybody do make mistake and learn from it.
Same goes for the limited number of IO pins.
Propeller II is the example of what Chip learned from the first generation. But even then, Propeller 1 will still be available for up to 15 years - still plentiful of times for us to enjoy the new Propeller 1 from now on. Propeller 1 still have up to 80% (if I am not wrong) of horsepower left untapped - plentiful of room for us to pick its brains clean.
And, as for Parallax Inc, I am grateful about Parallax Semiconductors, Inc. I hope their new fab starts off great as it's a shame to put such a great microcontroller to bed.
Comments
"When starting from scratch be prepared for a long haul of wading through treacle. Don't expect to be productive with them too quickly, you will die of frustration."
" X___ has all the hall marks of an academic startup, phenominaly clever, but they need to understand the needs of their potential customers a little better."
"If you can't tell from the above, having spent a disproportionate amount of time on it I am quite seriously disappointed. I hope sincerely that X___ can turn this around but will commit no more time to them until they have addressed the issue with their docs and tools."
I post this only because complexity issues are such a huge turn-off. The few times I've ventured into X-territory, I've been turned off very quickly. I'd have to have a desperate need for 400 MIPS before I'd spend a tuppence on X.
Really? Prop can do fast encoder reading, fast PWM generation, fast co-ordinate rotation (for vector control), PID, motion control profile generation. I would have thought that a single prop 1 could do a lot of it, and a second would help with the comms.
I've driven a brushless motor with a propeller (though not with a complex control algorithm) so I'm surprised you say this.
Graham
That was funny.
Bruce
P.S. It is both amusing and interesting, I must admit.
Can you get the same performance as in that application? I just don't see it.
The original poster got more than he asked for, don't think it was exactly what he asked for, but he sure got a lot of it.
C.W.
Probably can't emulate that board but the issue is can the propeller be used in the application. Can't get 500 MIPS for control or 1000MIPS for comms but I'm not sure it is needed? I'm happy to be enlightened, I'm not sure.
Graham
That's because we are lucky enough to have Leon. Leon apparently spends ten times more time on this site than he does on the various X related sites:
Leon's posts on xcore.com: 384
Leon's posts on parallax.com: 4,386
Ros.
Ross.
I also like helping people with their problems, and people have more problems here.
I've been pleasantly surprised at how easy going pretty much everyone here has been.
I am on another forum, a freakish AVR one. People get wound up there for the same reasons as here: a competing product almost always seems to better in every way, and seems to be brought up at every opportunity...
There as here, higher ups have had to request people stay on topic as it concerns that site's primary focus.
I get the sense here that an occasional reference to PIC or AVR or ARM, either pro or con is not a hanging offense. I expect that here as in over there, it is only when it starts to take over a thread, or is becoming repetitive that people get annoyed.
@everyone - It truly surprises me that Parallax is keeping such a low profile on this popular thread. I would have thought there would have been a lot more input.
1. This forum is congenial enough that people are not afraid to come here with their problems.
2. Our users have fewer proprietary restrictions on what they can ask in a public forum.
3. Our users are working on more ambitious projects than those other folks.
4. The other chip is so complicated to get started with that its users don't even know what to ask.
5. The Propeller attracts more users with less hardware or programming experience.
6. Or ... ?
-Phil
Some facts to take into consideration:
Xmos.com (Support Forum {entire forum))
Total posts 2602 • Total topics 433 • Total members 4211
Xcore.com (Support Form {entire forum})
Total posts 6800 | Total topics 942 | Total members 967
Forums.Parallax.Com (Support Forum)
Propeller : Posts: 195,143 / Threads: 16,182
Total members 22,677
Let's keep in mind that there is A LOT MORE going on here.
OBC
That is a understatement.
I asked the question because I want the limited time I have to learn a new uC to be put to good use; I also wanted to know if Spin and the Prop is worth the time. I have had my Prop for 2 days nows.
Within 24 hours and 3 tutorials I was able to reto-fit a OBJ and make a working PIR Intruder alarm with PST terminal messaging.
I have also written a 11 line program (a hack of the propeller help tutorials), which would (IMO) take probably a hundred lines (and a heck-of-alot of math) if you tried the same thing withoutr independent cogs operating in parallel.
I was on the fence between Arduino UNO and the Prop. Once I saw the Gadget Gangster usb Prop (which is priced right between a Arduino uno and a mega) I decided to take a chance.
So far, I am impressed with what the Prop can do.
(This will get me into trouble with some) I also think SPIN is a very decent language with some really nice features. It will take some time to become fluent, but Parallax has tons of resources.
Given the rate of activity on the forum and the constant flow of new members the message will never get across fully. There is inevitably going to be a string of occasions when you will feel the need to correct people on this issue, the problem is that everyone else has heard it all before ad nauseum.
Is this a noble quest for truth and justice or pedantry/OCD?
Graham
13 pages of discussion later, and I think that little comment answers it best of all. Here we have someone completely new to the propeller and within a short time they have managed to get something cool working. Great stuff!
To me, this says that the big secret of the propellers success may not not just be its parallel structure, but it might also be the Tutorials and the code in the Obex. And this forum.
With you being the initial creator of this interesting thread, I am glad to hear that you are doing so well after having your Prop for only 2 days. From what you described, it sounds like you picked it up rather quickly. Keep up the good work.
Bruce
It's neither - it's a simple misunderstanding. The "determinism" issue seems to arise frequently in these forums, because the Propeller has it as a fundamental attribute of the chip design, whereas other chips do not. With those chips you generally have to do something - often something messy and/or complex - to achieve it.
For a discussion about this specific to the X___ chip, including an excellent summary of a discussion on precisely this issue with David May (CTO of X___) see Heater's post on this page (about four from the top).
Ross.
P.S. I don't mean to derail the thread - in fact I'd like to see it get back to the original topic. Determinism as it relates to the Propeller, X___ and NXP can be further discussed in this thread.
When damagedcells created this thread, I wish he had included a poll with it pertaining to the original subject. Something along the lines of:
And I might respectfully disagree with Leon.
1) ARM chips use too much power -they get hot, and any chip that gets hot makes it more difficult in battery and solar applications.
2) With microchip uCs there is an impression that a handful of bytes are enough memory, and that a few kilobytes is more than enough.
The propeller uses a tiny amount of power and has lots of memory.
http://www.energymicro.com/
I've never seen an ARM chip running hot.
They come in a wide range of memory sizes - the NXP LPC2178 has 512 kB of program memory and 64 kB of SRAM and does 100 MIPS. It doesn't even get warm. I just powered up a $30 LPCXpresso board to make sure.
Never have so many had so much fun with so little in such a short a space of time.
Graham
That's exactly correct!
A couple of loosely related thoughts on the thread.
I want to address the technical issues above, but it takes time and I don't want to convey what is going to happen in regards to specific items until they're underway. More than once I've talked about something that happened differently.
While reading it I've really wanted to reply but we're working on many projects for all of you. It would be fantastic to bring you inside of Parallax, especially considering you are Parallax - or at least that's the way it is supposed to be. This isn't a place where we go to work just to make our salary, but to work with the technology and people we enjoy a whole bunch. As a result the output from Parallax will often be a mix of what you want but with a few surprises mixed in along the way. And the traits you dislike may disappear and be replaced with better things along the way. No company makes a single chip and thinks they've got it all right; your feedback is part of our next design, tools, code examples and presentation. The research you provide on these forums has an impact on what we do.
Somewhere above there was discussion about Parallax Semiconductor.
Please keep in mind that no changes occur overnight; we're a small company with a team of 44 people. At this moment eight of us are working to develop the system to properly support this effort, including: FAEs and an easy way to reach them; a Drupal web site, demonstrations of customer success, AppNotes and distribution. Several of you are involved in a customer focus group and a few others have joined from other corners of the industry to bring your expertise.
When we mention that we'll be commercializing the Propeller as Parallax Semiconductor it's because our commercial customers need that from the company. Engineers who work for small/medium companies and wish to specify the Propeller demand a different style from our company and they're elated that we'll give them the validation to take to their managers. Entrepreneurs don't seem to mind either way - and like hippy said "are you not going to eat at your favorite restaurant just because it isn't a chain?"
The approach to Parallax Semiconductor is simple: make it easy, showing high-level integration of easy-to-use "gold-plated" objects that accompany AppNotes; and quickly connect you with an FAE via Skype, phone, or in-person as you desire to talk about your project. We'll help you past any hurdles and see you through to completion.
In our AppNotes you'll see some traditional microcontroller applications but many that need the Prop's multi-core design.
Chip design cycles will speed up at Parallax, too. We're in the process of porting Prop2 from AHDL to VHDL so we can skip the schematic layout and go straight to die fabrication. The tools for doing this are very expensive (did I hear $1M?) so we've retained a firm specifically for this purpose. Even though we'll be quicker we still won't be releasing new processors every six months.
And because this is your company, we are using your investment to meet these goals. Parallax has no venture funding and the company reinvests more than most CPAs would imagine. Venture-funded companies may have more waste in decision making as a result of pressure to deliver, including more staff turnover, and don't easily generate customer loyalty as a result. I think hobbyists and commercial users would prefer our method. When we spend your money we must do it in a way that doesn't create unnecessary risk for you, which is why you've been able to count on us as a steady supplier for 20+ years. Now that we have our own chip you can count on us for at least 20 more years.
The target number is Propellers we should sell is peanuts to most semiconductor suppliers but it is more than we're selling today. As engineers you already know something about silicon margins compared to robot kits, BASIC Stamps and easy-to-use sensors.
From a sales perspective Parallax's revenue is equally derived from hobby, education and commercial customers yet the company looks like it serves the first two groups. We are very sensitive to the fact that while we undertake Parallax Semiconductor we must increase the product base and support for our education/hobby customers more than before. The sensors, motor controllers, WiFi, PropBOEs, educational kits and selection of simply useful designs from Parallax will continue at the same rate.
I just wanted to reaffirm that we're not sitting on our laurels around here - we've read and heard first-hand every statement from all of you and many of them are being built into our products. Parallax is an active place with enthusiasm and energy, with much investment in the future. We're more excited than ever about the changes that lie ahead!
I'll add in some more details soon, I promise.
Sincerely,
Ken Gracey
Parallax Inc.
(916) 624-8333 ext 5510
kgracey@parallax.com
It's also handy for testing the other thing such as testing a dead switchmode power supply (function generator testing), as well as
testing few other things, like sticky logics (all of our favorite 74s family logics we love).
SPIN's pretty slow - but it's okay, as Chip haven't designed anything as serious as Propeller before then - everybody do make mistake and learn from it.
Same goes for the limited number of IO pins.
Propeller II is the example of what Chip learned from the first generation. But even then, Propeller 1 will still be available for up to 15 years - still plentiful of times for us to enjoy the new Propeller 1 from now on. Propeller 1 still have up to 80% (if I am not wrong) of horsepower left untapped - plentiful of room for us to pick its brains clean.
And, as for Parallax Inc, I am grateful about Parallax Semiconductors, Inc. I hope their new fab starts off great as it's a shame to put such a great microcontroller to bed.