Its sad that the RPi foundation didn't have the b***s to give even passing 'props' to Gumstix for doing this 10 or so years ago.
People were also putting FPGAs onto DIMMs back in the original dot com days. There was Pilchard which I think came from academia, and Nuron which was purchased by Intel. I had a friend at one of these places, and one of the fatal flaws was the PC motherboards all routed the data lines differently so they were effectively scrambled from system to system. Which made it a nightmare to deal with since you would need to spin the design for each, or put in the fully programmable mux and some sort of discovery mechanism. FPGAs were smaller back then...
Its sad that the RPi foundation didn't have the b***s to give even passing 'props' to Gumstix for doing this 10 or so years ago.
That's rather harsh. Who says they have even ever heard of a Gumstix? I'm mean, who has?
Besides people have been building Gumstix like modules since forever.
Also as far as I know Gumstix is not even similar, it does not come on a SODIMM connector. There are many other "ARM on DIMM module" products on offer from others now a days.
KeithE,
...and one of the fatal flaws was the PC motherboards all routed the data lines differently so they were effectively scrambled from system to system.
It should be pointed out that these ARM on DIMM products, the Pi and many others, are not intended to be plugged into the DIMM sockets in PCs. They are only making use of the connector standard as it is wide spread and cheap. They are intended to make it easy for users to create their own boards that can cheaply easily make use of the ARM.
However, Gumstix has been around forever, and I remember early 2000's there was one on a DIMM module as well.
Not sure why, however you seem to both agree with that statement and then want to contradict it as well.
I suppose it 'possible' that the idea somehow popped up without someone having ever heard of Gumstix, however I frankly doubt it.
Any and everyone looking for small computers from early 2000 onwards has heard of it, OCO, and several other credit card sized ones, as they were all one could realistically get.
Some of the younger hipsters involved in the project may in fact be oblivious, however anyone over the age of 35 interested in Linux and computers/electronics that I know all were talking about them 10+ years ago.
I can't believe there aren't some middle-aged guys involved with it who weren't aware of it.
I never even bought one, however as soon as I saw the Arduino one, it sure seemed like a direct rip.
Not saying Arduino is all Smile, just that is sure seems like they took someone's idea and recycled it without acknowledgement. Seemed kinda odd for an organization that seems to be so FOSS-aligned.
My aim with PropWare is to provide a similar experience to Arduino. I know there is a long way to go, but that was the point of the makefile system, the Eclipse IDE, the documentation, and the expanse of objects found within. Of course... it's a bit of my own taste added in. There are some things I don't like about Arduino that I did differently (no silly setup and loop functions)
C. Avoid the creation of one package that is supposed to do it all. It maybe possible, the resulting code will likely offer very little to most people that want to read code to learn. Only the most geekish will appreciate such orchestration.
@Loopy
this is something you posted a few pages ago. It caught my eyes. Can you expand on what you mean by "one package that is supposed to do it all?"
I don't mean for this to come off as defensive - if your statement was aimed at packages like PropWare, then I want to understand where you're coming from and then adjust my plans accordingly.
Someone mentioned 8-bit ports at some point. That can easily be simulated with PropWare's SimplePort class. You could, of course, even take it to another level and create PropWare::BytePort or something similar, which is just PropWare::SimplePort with an 8-pin restriction (possibly aligned to 8-bit frames).
On the subject of bringing prop-gcc into gcc mainline and I'm sure you guys are doing this already. It's better to have open-communication with the gcc project people waay before you're ready to merge.
Large project owners tend to prefer a series of smaller merges than one huge tarball. Add to that that they're likely going to have a bunch of changes they may want in order to have your code consistant with the wider project.
I'm excited to see the merge upstream, I have a perverse want to compile FORTRAN for the prop. Another retro language like FORTH :-)
If I'm not mistaken you can tweak some language selection option in the propgcc build scripts and get yourself a FORTRAN compiler already. Of course it won't have any Prop specific feature support so you then have a nice project to provide that on your hands
Another retro language like FORTH
Ahhggg! Retro perhaps, it was the mother of them all. "like FORTH", noooo...
I don't get the desire for "gumstix" attribution. Who says they invented the idea? I recall my father talking about replaceable electronic modules with me as kid back in the 1960's. i.e. circuit boards with edge connectors. It was a radical idea, things like TV's and radios were still point to point connected.
I don't want to belittle the gumstix but really such ideas have been around forever. Early 2000 is not forever it's very recent history.
Back in the day processors were put on their own cards with things like euro connector backplanes. When the Altair came out, for example, they wanted a cheap expansion system and ended up using that 100 pin connector just because it was plentiful and cheap.
Time progresses, things get smaller, when you want a cheap expantion connector what do you use? Stuff that's already used every where else, like that DIMM socket in every PC.
I would not class the guys on the Pi project as "young" or "hipsters". Many of them nostalgically recall their youth filled with C64, Sinclair Spectrums, BBC computers, and so on.
For timeline sake in regards to System on Module devices, ZFMicro starting making their first in 1995. They went quiet for a while thanks to lawsuits with the greedy folks at National Semiconductor, but are back in the game again with a 486 system based BGA. At my previous employer, we built a small board in 1999 that used a x86 System on Module from ZFMicro that went into ATM modem controllers.
As for Gumstix specifically, their $239 Overo Tidalstorm kit is close enough to a $40 RPi model B to compare the outrageous price difference as a factor in not considering Gumstix.
As for SPIN vs PASM vs anything else, I started micros with PBASIC and it was extremely easy for me to learn SPIN. I have yet to write any PASM, but leverage many objects that use PASM. I still prefer SPIN over anythin else because I can digest it easily and so far,without heartburn. Next teo languages on my list are C and Python (C for the Prop and AVR, Python for projects with my RPi)
Far as I know/recall, Gumstix were the first ones to have an actual full computer (Linux) on a DIMM chip. Electronics modules have indeed been well known forever, but not what could arguably be called the PSOC....
I also wouldn't classify the average Pi organization member as a hipster, I was giving some of them the benefit of the doubt, as I agree that most of them are probably graybeards, who probably had seen the gumstix.
IIRC, the gumstix were made on a DIMM exactly because they could then be slotted into standard/custom SBC.
Again, I never even bought one so I'm not exactly some PO'd fanboy. Just that as soon as I saw it I recognized it as copying the form/function of the Gumstix, and some of the selling points for it seemed pretty similar to the GS.
Really, really odd that the old neckbeards would have chanced the use the same DIMM FF, for the same reasons, and not have seen/known about the GS line. Hence my hipsters comment.
WBA- I don't think anyone is suggesting GS would make any financial sense in this day and age with the current RPi, or even BBB offerings.
I don't get the desire for "gumstix" attribution. Who says they invented the idea? I recall my father talking about replaceable electronic modules with me as kid back in the 1960's. i.e. circuit boards with edge connectors. It was a radical idea, things like TV's and radios were still point to point connected.
I don't want to belittle the gumstix but really such ideas have been around forever. Early 2000 is not forever it's very recent history.
.....
I would not class the guys on the Pi project as "young" or "hipsters". Many of them nostalgically recall their youth filled with C64, Sinclair Spectrums, BBC computers, and so on.
Now we have gone from the Pi team being "hipsters" to "greybeards". Neither of which is correct.
You make it sound like some insanely great invention by gumstix. I'm not convinced that the idea of a putting a circuit on a board with an edge connector needs attributing to anyone. It is a great idea but I think we have long since forgotten where it came from. Certainly not gumstix. And his been done my many, over and over again.
+1. When I started servicing minicomputers and instrumentation in the mid 70's circuit boards with edge connectors were old hat, and once memory was available on dimms the idea of a computer board using the same connectors and pin out for the address and data bus was pretty obvious.
Now we have gone from the Pi team being "hipsters" to "greybeards". Neither of which is correct.
You make it sound like some insanely great invention by gumstix. I'm not convinced that the idea of a putting a circuit on a board with an edge connector needs attributing to anyone. It is a great idea but I think we have long since forgotten where it came from. Certainly not gumstix. And his been done my many, over and over again.
+1. When I started servicing minicomputers and instrumentation in the mid 70's circuit boards with edge connectors were old hat, and once memory was available on dimms the idea of a computer board using the same connectors and pin out for the address and data bus was pretty obvious.
What ???
ICL and Singer/Frieden minis (and others) used pcbs with edge connectors from at least 1970 up until the early 90's. The ICL System 25 was released in 1981 and it had a standard backplane and plugin edge connector boards. In fact I built boards to plug into the bus in ~1990. The video terminals of 1970 and 1975, and disk drives also used gold edge connectors. The Centronics printers and compatibles through the 80's had sockets for interface boards with gold edge connectors (I built the interface boards to ICL minis).
What ???
ICL and Singer/Frieden minis (and others) used pcbs with edge connectors from at least 1970 up until the early 90's. The ICL System 25 was released in 1981 and it had a standard backplane and plugin edge connector boards. In fact I built boards to plug into the bus in ~1990. The video terminals of 1970 and 1975, and disk drives also used gold edge connectors. The Centronics printers and compatibles through the 80's had sockets for interface boards with gold edge connectors (I built the interface boards to ICL minis).
Once again, somebody used a colloquial expression, and another person took it the wrong way.
Your list of all the places it was used for so long is literally an example of the first definition of "old hat".
There must be a second or implied definition, meaning "out of fashion" which is not what was meant.
Once again, somebody used a colloquial expression, and another person took it the wrong way.
Your list of all the places it was used for so long is literally an example of the first definition of "old hat".
There must be a second or implied definition, meaning "out of fashion" which is not what was meant.
LOL - yes!
Similar expressions and context from different parts of the world (or even in the same country) have caught me out more than once!
LOL, not trying to be argumentative, honestly. Its obvious that edge connectors preceded both RPi and GS by decades.
Not arguing that.
My only 'complaint' was that the entire 'computer on a DIMM' had been seen before, and a lot of RPi'ers seem to be taking this as some massive FF coup or some such, and their blog posting about it sorta seeems to imply 'they' are geniuses for coming up with the idea...
When GS was announced, it made headlines all across the known interenet, on every SBC dist-list, Yahoo Groups and Linux forum.
And not specifically because it was Linux. My opinion was that RPi has a lot of greybeards who probably were familiar with the GS, and probably a lot of younger 'hipsters' who weren't, as likely 90% of RPi users aren't.
Nothing wrong with reusing a 10+(?) year idea of course, however tooting one's own horn while doing so is rather gouche, and just makes me think 'hipsters' must be involved.
Simply viewing this as yet another edge connection 'board', I agree it is nothing new.
Now we have gone from the Pi team being "hipsters" to "greybeards". Neither of which is correct.
You make it sound like some insanely great invention by gumstix. I'm not convinced that the idea of a putting a circuit on a board with an edge connector needs attributing to anyone. It is a great idea but I think we have long since forgotten where it came from. Certainly not gumstix. And his been done my many, over and over again.
it seems to be in the gens. Do you remember the time, we invented sex? Knowledge comes from ancient times. I know something, I know everything, I know everything better, I know that I don't know everything, I know that I know nothing. Lhotse. But education must take place in less and less time, so people in the end only know the resume, : I know nothing. Those that reach the third level of enlightenment are called to be professor at out universities ;-) By the way: "The journey is the reward." is not a book title, nor realized by Confucius, but the final proof, that hoover robots existed forever!
Never heard of ZFMicro. Amazing, you can get a dead slow and small x86 system on a module. Who is their target market for that now a days?
I've not heard of them either, but I'm also not in their target market, which is legacy system maintenance. It's a HUGE market, and it's worldwide. Now, before you say everyone should update their old systems to use modern hardware and OSs, I'll just remind you of a little thing called money. Replacing these systems outright is far more expensive than buying a replacement x86 SBC for $200 or $300.
By the way: "The journey is the reward." is not a book title, nor realized by Confucius, but the final proof, that hoover robots existed forever!
Well, it's also been used for the title of at least one book, an unauthorized biography of Steve Jobs. He HATED that book, even -- as I recall -- going so far as threatening to pull Apple product from any store that carried it.
Jobs also disliked the concept of "Hoover robots." Thought it was an unclassy use of the technology. Thought I'd throw that out.
I guess you are right. My observation though is that sometimes supporting such legacy systems ends up costing more than redoing it. Of course that is not always possible with the old world of closed source platform specific software. I guess those products have their place.
Once again, somebody used a colloquial expression, and another person took it the wrong way.
Your list of all the places it was used for so long is literally an example of the first definition of "old hat".
There must be a second or implied definition, meaning "out of fashion" which is not what was meant.
I think I misremembered, and combined both this book and a later book, iCon: Steve Jobs, The Greatest Second Act in the History of Business. Both were written by Jeff Young. The NY Times ran this article when the iCon book was about to come out:
I don't mean to sidetrack the discussion with Jobs, but someone mentioned the quote often attributed to him. And as a tech leader I think it's curious he had little or no interest in open source hardware, and would not see the value of Arduino, Pi, or any of the others. It's curious because Apple's annual revenue, primarily from closed source hardware, far outstrips Google's and anyone else playing in the commercial/open source space.
Why would Jobs be interested in open hardware? Apple lives by selling hardware. Would Intel publish it's CPU designs?
Comparing Google to Apple is a bit odd in my mind. One is an advertising company and the other is a computer/phone/music player company. Might as well compare Apple to Ford or McDonalds.
Why would Jobs be interested in open hardware? Apple lives by selling hardware. Would Intel publish it's CPU designs?
Comparing Google to Apple is a bit odd in my mind. One is an advertising company and the other is a computer/phone/music player company. Might as well compare Apple to Ford or McDonalds.
We often hear the Arduino is popular because it's open source. It's not true. Well, it is true, to a point, but far from the full story. Apple is proof that you can have a completely closed system, be popular, and make oodles of money.
You haven't been paying attention to Android or the billions Google has spent acquiring hardware technology companies. A bit more than an advertising company.
Never heard of ZFMicro. Amazing, you can get a dead slow and small x86 system on a module. Who is their target market for that now a days?
As Gordon mentioned, you would be surprised at how much machinery uses DOS based operating systems. The DEK 265 Screen Printers I replaced two years ago at work ran DOS 6.2.2. When one went down hard, I had to dig deep to pull out all those DOS commands. The machine software was on 3.5" floppy disks.
Even after buying two new AOI machines, we kept our older machine because it is a workhorse. However, I dread the day it goes down.... it runs on Windows NT 3.51.
Not all legacy hardware has gone to the boneyard. In fact, about once every three months, I build a batch of about 250 industrial 28.8k modems at work.
Similar expressions and context from different parts of the world (or even in the same country) have caught me out more than once!
Indeed! Many years ago I was nearly slapped by a young woman in your country for using a term that we consider polite. She asked me what her best physical feature was and I told her she had a nice fanny, immediately remembering that word means something else down under (double-entendre not intended). As her eyes flashed wide I realized my mistake and frantically said, "I meant bum! I meant bum! I meant bum!"
When I was an undergraduate at Case Institute of Technology, the university computer center had a Univac 1107. They also had an operational Univac I which one of the railroads leased for keeping track of their freight cars. They had their own Univac I which was running all the time and not up to handling their freight volume, so they went to multiprocessing (two Univac Is) for the throughput and redundancy. Talk about legacy maintenance. The Univac I was actually up more than the 1107.
Comments
People were also putting FPGAs onto DIMMs back in the original dot com days. There was Pilchard which I think came from academia, and Nuron which was purchased by Intel. I had a friend at one of these places, and one of the fatal flaws was the PC motherboards all routed the data lines differently so they were effectively scrambled from system to system. Which made it a nightmare to deal with since you would need to spin the design for each, or put in the fully programmable mux and some sort of discovery mechanism. FPGAs were smaller back then...
Besides people have been building Gumstix like modules since forever.
Also as far as I know Gumstix is not even similar, it does not come on a SODIMM connector. There are many other "ARM on DIMM module" products on offer from others now a days.
KeithE, It should be pointed out that these ARM on DIMM products, the Pi and many others, are not intended to be plugged into the DIMM sockets in PCs. They are only making use of the connector standard as it is wide spread and cheap. They are intended to make it easy for users to create their own boards that can cheaply easily make use of the ARM.
However, Gumstix has been around forever, and I remember early 2000's there was one on a DIMM module as well.
Not sure why, however you seem to both agree with that statement and then want to contradict it as well.
I suppose it 'possible' that the idea somehow popped up without someone having ever heard of Gumstix, however I frankly doubt it.
Any and everyone looking for small computers from early 2000 onwards has heard of it, OCO, and several other credit card sized ones, as they were all one could realistically get.
Some of the younger hipsters involved in the project may in fact be oblivious, however anyone over the age of 35 interested in Linux and computers/electronics that I know all were talking about them 10+ years ago.
I can't believe there aren't some middle-aged guys involved with it who weren't aware of it.
I never even bought one, however as soon as I saw the Arduino one, it sure seemed like a direct rip.
Not saying Arduino is all Smile, just that is sure seems like they took someone's idea and recycled it without acknowledgement. Seemed kinda odd for an organization that seems to be so FOSS-aligned.
@Loopy
this is something you posted a few pages ago. It caught my eyes. Can you expand on what you mean by "one package that is supposed to do it all?"
I don't mean for this to come off as defensive - if your statement was aimed at packages like PropWare, then I want to understand where you're coming from and then adjust my plans accordingly.
Someone mentioned 8-bit ports at some point. That can easily be simulated with PropWare's SimplePort class. You could, of course, even take it to another level and create PropWare::BytePort or something similar, which is just PropWare::SimplePort with an 8-pin restriction (possibly aligned to 8-bit frames).
Large project owners tend to prefer a series of smaller merges than one huge tarball. Add to that that they're likely going to have a bunch of changes they may want in order to have your code consistant with the wider project.
I'm excited to see the merge upstream, I have a perverse want to compile FORTRAN for the prop. Another retro language like FORTH :-)
If I'm not mistaken you can tweak some language selection option in the propgcc build scripts and get yourself a FORTRAN compiler already. Of course it won't have any Prop specific feature support so you then have a nice project to provide that on your hands
Ahhggg! Retro perhaps, it was the mother of them all. "like FORTH", noooo...
I don't get the desire for "gumstix" attribution. Who says they invented the idea? I recall my father talking about replaceable electronic modules with me as kid back in the 1960's. i.e. circuit boards with edge connectors. It was a radical idea, things like TV's and radios were still point to point connected.
I don't want to belittle the gumstix but really such ideas have been around forever. Early 2000 is not forever it's very recent history.
Back in the day processors were put on their own cards with things like euro connector backplanes. When the Altair came out, for example, they wanted a cheap expansion system and ended up using that 100 pin connector just because it was plentiful and cheap.
Time progresses, things get smaller, when you want a cheap expantion connector what do you use? Stuff that's already used every where else, like that DIMM socket in every PC.
I would not class the guys on the Pi project as "young" or "hipsters". Many of them nostalgically recall their youth filled with C64, Sinclair Spectrums, BBC computers, and so on.
As for Gumstix specifically, their $239 Overo Tidalstorm kit is close enough to a $40 RPi model B to compare the outrageous price difference as a factor in not considering Gumstix.
As for SPIN vs PASM vs anything else, I started micros with PBASIC and it was extremely easy for me to learn SPIN. I have yet to write any PASM, but leverage many objects that use PASM. I still prefer SPIN over anythin else because I can digest it easily and so far,without heartburn. Next teo languages on my list are C and Python (C for the Prop and AVR, Python for projects with my RPi)
Never heard of ZFMicro. Amazing, you can get a dead slow and small x86 system on a module. Who is their target market for that now a days?
I also wouldn't classify the average Pi organization member as a hipster, I was giving some of them the benefit of the doubt, as I agree that most of them are probably graybeards, who probably had seen the gumstix.
IIRC, the gumstix were made on a DIMM exactly because they could then be slotted into standard/custom SBC.
Again, I never even bought one so I'm not exactly some PO'd fanboy. Just that as soon as I saw it I recognized it as copying the form/function of the Gumstix, and some of the selling points for it seemed pretty similar to the GS.
Really, really odd that the old neckbeards would have chanced the use the same DIMM FF, for the same reasons, and not have seen/known about the GS line. Hence my hipsters comment.
WBA- I don't think anyone is suggesting GS would make any financial sense in this day and age with the current RPi, or even BBB offerings.
Now we have gone from the Pi team being "hipsters" to "greybeards". Neither of which is correct.
You make it sound like some insanely great invention by gumstix. I'm not convinced that the idea of a putting a circuit on a board with an edge connector needs attributing to anyone. It is a great idea but I think we have long since forgotten where it came from. Certainly not gumstix. And his been done my many, over and over again.
ICL and Singer/Frieden minis (and others) used pcbs with edge connectors from at least 1970 up until the early 90's. The ICL System 25 was released in 1981 and it had a standard backplane and plugin edge connector boards. In fact I built boards to plug into the bus in ~1990. The video terminals of 1970 and 1975, and disk drives also used gold edge connectors. The Centronics printers and compatibles through the 80's had sockets for interface boards with gold edge connectors (I built the interface boards to ICL minis).
Once again, somebody used a colloquial expression, and another person took it the wrong way.
Your list of all the places it was used for so long is literally an example of the first definition of "old hat".
There must be a second or implied definition, meaning "out of fashion" which is not what was meant.
Similar expressions and context from different parts of the world (or even in the same country) have caught me out more than once!
Not arguing that.
My only 'complaint' was that the entire 'computer on a DIMM' had been seen before, and a lot of RPi'ers seem to be taking this as some massive FF coup or some such, and their blog posting about it sorta seeems to imply 'they' are geniuses for coming up with the idea...
When GS was announced, it made headlines all across the known interenet, on every SBC dist-list, Yahoo Groups and Linux forum.
And not specifically because it was Linux. My opinion was that RPi has a lot of greybeards who probably were familiar with the GS, and probably a lot of younger 'hipsters' who weren't, as likely 90% of RPi users aren't.
Nothing wrong with reusing a 10+(?) year idea of course, however tooting one's own horn while doing so is rather gouche, and just makes me think 'hipsters' must be involved.
Simply viewing this as yet another edge connection 'board', I agree it is nothing new.
I don't see any on the announcement page: http://www.raspberrypi.org/raspberry-pi-compute-module-new-product/
I've not heard of them either, but I'm also not in their target market, which is legacy system maintenance. It's a HUGE market, and it's worldwide. Now, before you say everyone should update their old systems to use modern hardware and OSs, I'll just remind you of a little thing called money. Replacing these systems outright is far more expensive than buying a replacement x86 SBC for $200 or $300.
Well, it's also been used for the title of at least one book, an unauthorized biography of Steve Jobs. He HATED that book, even -- as I recall -- going so far as threatening to pull Apple product from any store that carried it.
Jobs also disliked the concept of "Hoover robots." Thought it was an unclassy use of the technology. Thought I'd throw that out.
I guess you are right. My observation though is that sometimes supporting such legacy systems ends up costing more than redoing it. Of course that is not always possible with the old world of closed source platform specific software. I guess those products have their place.
I think I misremembered, and combined both this book and a later book, iCon: Steve Jobs, The Greatest Second Act in the History of Business. Both were written by Jeff Young. The NY Times ran this article when the iCon book was about to come out:
http://www.nytimes.com/2005/04/30/technology/30apple.html?_r=0
I don't mean to sidetrack the discussion with Jobs, but someone mentioned the quote often attributed to him. And as a tech leader I think it's curious he had little or no interest in open source hardware, and would not see the value of Arduino, Pi, or any of the others. It's curious because Apple's annual revenue, primarily from closed source hardware, far outstrips Google's and anyone else playing in the commercial/open source space.
Comparing Google to Apple is a bit odd in my mind. One is an advertising company and the other is a computer/phone/music player company. Might as well compare Apple to Ford or McDonalds.
We often hear the Arduino is popular because it's open source. It's not true. Well, it is true, to a point, but far from the full story. Apple is proof that you can have a completely closed system, be popular, and make oodles of money.
You haven't been paying attention to Android or the billions Google has spent acquiring hardware technology companies. A bit more than an advertising company.
As Gordon mentioned, you would be surprised at how much machinery uses DOS based operating systems. The DEK 265 Screen Printers I replaced two years ago at work ran DOS 6.2.2. When one went down hard, I had to dig deep to pull out all those DOS commands. The machine software was on 3.5" floppy disks.
Even after buying two new AOI machines, we kept our older machine because it is a workhorse. However, I dread the day it goes down.... it runs on Windows NT 3.51.
Not all legacy hardware has gone to the boneyard. In fact, about once every three months, I build a batch of about 250 industrial 28.8k modems at work.
Indeed! Many years ago I was nearly slapped by a young woman in your country for using a term that we consider polite. She asked me what her best physical feature was and I told her she had a nice fanny, immediately remembering that word means something else down under (double-entendre not intended). As her eyes flashed wide I realized my mistake and frantically said, "I meant bum! I meant bum! I meant bum!"