Not Open Source Hardware.

Cabe,

You've completely missed the point of open source. It's not about what you or I think anyone might want to do, or about our notions of what is and isn't economically viable. The fact is that the term "open source hardware" means something very specific and given where we are on the adoption curve, that tons of the stuff is out there and well known definitions exist, there is no reasonable excuse for bending that defition so far from its intended, and widely accepted, meaning.

I also hear the point about the Broadcom SoC, however, if this were an open source design I could rip that out and replace it with another piece of silicon. Whether or not this would be economically viable, or the RP Foundation or Broadcom would be pleased, would be neither here nor there: with an open source hardware design I would be free to do as I please.

Regards,

Andrew

I actually find all this discussion about open hardware and what is or isn't really funny. How many of these open source hardware are really as open as they expound. Those that use electronic component do they make available the actual die designs and lithograph plates or are they closed design proprietry designs? Afterall if you have the capabilities to make chips those would allow you to make your own. If not why is it OK to use closed source non CPU chips when it isn't alright to use a closed source CPU chip?

You raise a good point - almost all Open Source Hardware project I know of are only open down to some level of abstraction.  That tends to be the silicon.  There are efforts like OpenCores to even open up that aspect but I think that is quite a bit far off for many applications. 

There was a good presentation on this topic at the Open Source Hardware Summit:

http://www.openhardwaresummit.org/wp-content/uploads/2011/09/Open_Sourcing_the_IC-s.pdf

[PDF]

Cheers,

Drew

How many silicon dies and/or chips have you made recently in your home lab?

For most of the chips used in OSHW designs information on how to use particular chip is available (read - data-sheet) whereas Broadcom is notorious for consealing all information about it's chips from mere mortals. OK, in this case Raspberry Pi Foundation found the way to publish most of the datasheet, except GPU part.

I haven't made any PCBs in my home lab either. I couldn't recreate proprietary software libraries either. Some people get very evangelical about open hardware or open source but seemingly choose to ignore their own arguments when it doesn't suit their particular view of the world. The simplistic view of open hardware is that you have ALL the information to recreate it for yourself without any proprietry items getting in the way. If you don't have the die design and die lithographs then you can't recreate it. Similarly if you don't have the doping information for the silicon then again you can't recreate it. The fact that you need specialist equipment or a heck of a lot of money isn't in question, maybe I am a chip manufacture and have the facilities. (I'm not but I'm making the point in a silly way.) I've never seen open hardware defined as open only as far as you have the money.

Practically you have to draw the line somewhere but the various arguments do seem to be somewhat arbitrary at times. Another silly example say you have an open hardware device which is truely open all the way down to the raw ingredients. Lets say it is a computer like the Pi. What happens if you connect it to a TV which isn't open hardware does that make no longer open hardware because it includes a proprietary element. If that is OK why is a proprietary graphics chip different? You just substituted an proprietary element without knowing about its innards so why is the graphics chip any different.

I'm sure one argument could be you can operate a Pi headless without a display, but there again if you can operate it headless without a display why do you need a graphics chip (even if it is part of the SOC).

By the way just for clarification I'm not treating this seriously I'm not hung up on open hardware etc just interested in the conversation.

I haven't made any PCBs in my home lab either. I couldn't recreate proprietary software libraries either. Some people get very evangelical about open hardware or open source but seemingly choose to ignore their own arguments when it doesn't suit their particular view of the world. The simplistic view of open hardware is that you have ALL the information to recreate it for yourself without any proprietry items getting in the way. If you don't have the die design and die lithographs then you can't recreate it. Similarly if you don't have the doping information for the silicon then again you can't recreate it. The fact that you need specialist equipment or a heck of a lot of money isn't in question, maybe I am a chip manufacture and have the facilities. (I'm not but I'm making the point in a silly way.) I've never seen open hardware defined as open only as far as you have the money.

Practically you have to draw the line somewhere but the various arguments do seem to be somewhat arbitrary at times. Another silly example say you have an open hardware device which is truely open all the way down to the raw ingredients. Lets say it is a computer like the Pi. What happens if you connect it to a TV which isn't open hardware does that make no longer open hardware because it includes a proprietary element. If that is OK why is a proprietary graphics chip different? You just substituted an proprietary element without knowing about its innards so why is the graphics chip any different.

I'm sure one argument could be you can operate a Pi headless without a display, but there again if you can operate it headless without a display why do you need a graphics chip (even if it is part of the SOC).

By the way just for clarification I'm not treating this seriously I'm not hung up on open hardware etc just interested in the conversation.