Hi,
Just wondering how much is it these days ball park to pay someone like witstech or a similar design outfit to design an A20 board (like a cubie) and make 10 samples?
Hi Joem, look at olimex boards at https://www.olimex.com/Products/OLinuXino/A20/open-source-hardware they are OSHW and so the complete design is open source.
On Wed, 9 Sep 2015 at 15:51 joem joem@martindale-electric.co.uk wrote:
Hi,
Just wondering how much is it these days ball park to pay someone like witstech or a similar design outfit to design an A20 board (like a cubie) and make 10 samples?
arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk
On Wed, 2015-09-09 at 14:53 +0000, Alejandro Mery wrote:
Hi Joem, look at olimex boards at https://www.olimex.com/Products/OLinuXino/A20/open-source-hardware they are OSHW and so the complete design is open source.
--
Hi, Just wondering how much is it these days ball park to pay someone like witstech or a similar design outfit to design an A20 board (like a cubie) and make 10 samples?
Hi Alejandro,
Thank you for that link. Do you think they can design a custom board to specifications and just hand over the design files for a fee? And make 10 samples to prove it works.
On Wed, Sep 9, 2015 at 3:53 PM, Alejandro Mery amery@geeks.cl wrote:
Hi Joem, look at olimex boards at https://www.olimex.com/Products/OLinuXino/A20/open-source-hardware they are OSHW and so the complete design is open source.
the designs are under the "attribution" style licenses, forcing you to advertise as part of the product. as such they are *not* libre licensed. if the designs were released under a GPL license it would be a different matter.
l.
On 2015-09-09 at 16:04:26 +0100, Luke Kenneth Casson Leighton wrote:
On Wed, Sep 9, 2015 at 3:53 PM, Alejandro Mery amery@geeks.cl wrote:
Hi Joem, look at olimex boards at https://www.olimex.com/Products/OLinuXino/A20/open-source-hardware they are OSHW and so the complete design is open source.
the designs are under the "attribution" style licenses, forcing you to advertise as part of the product. as such they are *not* libre licensed. if the designs were released under a GPL license it would be a different matter.
really?
"provided that you conspicuously and appropriately publish on each copy an appropriate copyright notice; keep intact all notices stating that this License and any non-permissive terms added in accord with section 7 apply to the code; keep intact all notices of the absence of any warranty; and give all recipients a copy of this License along with the Program." in GPLv3
"provided that you conspicuously and appropriately publish on each copy an appropriate copyright notice and disclaimer of warranty; keep intact all the notices that refer to this License and to the absence of any warranty; and give any other recipients of the Program a copy of this License along with the Program." in GPLv2
Not very different from the Attribution clause in https://creativecommons.org/licenses/by-sa/4.0/legalcode
The only widespread FLOSS licenses that I know of that doesn't include similar wording are a Public Domain dedication (such as CC0) and the WTFPL.
On Wed, Sep 9, 2015 at 6:20 PM, Elena ``of Valhalla'' elena.valhalla@gmail.com wrote:
On 2015-09-09 at 16:04:26 +0100, Luke Kenneth Casson Leighton wrote:
On Wed, Sep 9, 2015 at 3:53 PM, Alejandro Mery amery@geeks.cl wrote:
Hi Joem, look at olimex boards at https://www.olimex.com/Products/OLinuXino/A20/open-source-hardware they are OSHW and so the complete design is open source.
the designs are under the "attribution" style licenses, forcing you to advertise as part of the product. as such they are *not* libre licensed. if the designs were released under a GPL license it would be a different matter.
really?
yes.
"provided that you conspicuously and appropriately publish on each copy an appropriate copyright notice; keep intact all notices stating that this License and any non-permissive terms added in accord with section 7 apply to the code; keep intact all notices of the absence of any warranty; and give all recipients a copy of this License along with the Program." in GPLv3
"provided that you conspicuously and appropriately publish on each copy an appropriate copyright notice and disclaimer of warranty; keep intact all the notices that refer to this License and to the absence of any warranty; and give any other recipients of the Program a copy of this License along with the Program." in GPLv2
so that's in the source code.
Not very different from the Attribution clause in https://creativecommons.org/licenses/by-sa/4.0/legalcode
attribution clauses require you to advertise *on the product*.
that's utterly different from requiring to maintain the copyright notice and the fact that there is a license *in the source code*.
if you're not familiar with or don't clearly understand the difference, look up the history behind why the Debian Team renamed firefox to "iceweasel".
l.
On Wed, Sep 9, 2015 at 7:42 PM, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
if you're not familiar with or don't clearly understand the difference, look up the history behind why the Debian Team renamed firefox to "iceweasel".
here you go:
https://en.wikipedia.org/wiki/Mozilla_Corporation_software_rebranded_by_the_...
so, if joe were to use the CC-attribution design, he would be required to advertise the name of the creator on the OUTSIDE OF THE PRODUCT.
prominently. including on the casework if the PCB is inside a box.
so, when someone bought that product, despite the fact that joe would be the one that had put the effort into marketing, sales, spent the up-front cash on prototypes (which is a considerable amount)...
.... guess whom the customers are most likely to contact?
not joe - the one who put the effort into getting a polished professional product into their hands.
by complete contrast, for GPL'd products, all you have to do is put a little bit of paper in the box saying "contact us if you want the source code".
you *DO NOT* have to put "Copyright (C) Blah Blah" on the OUTSIDE OF THE PRODUCT, just because it's got some software in it (or in the case of hardware is manufactured from design files that are GPL licensed).
is that clearer?
btw i was amazed and deeply impressed when i bought a TP-Link router last year, because it contained *exactly that* in the box. finally - at long last - a large company that understands its obligations under the GPL.
l.
Luke Kenneth Casson Leighton lkcl@lkcl.net writes:
On Wed, Sep 9, 2015 at 7:42 PM, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
if you're not familiar with or don't clearly understand the difference, look up the history behind why the Debian Team renamed firefox to "iceweasel".
here you go:
https://en.wikipedia.org/wiki/Mozilla_Corporation_software_rebranded_by_the_...
That was nothing to do with copyright, nor attribution.
The problem is the way that Mozilla enforces its trademark.
Mozilla is (fairly reasonably) concerned that people might take one of it's trademarked programs, trojan it, and redistribute the result under the name of e.g. Firefox, thus tainting their good name. They therefore reserve the right to specify which code costitutes Firefox, etc. and want sight of any patches that are applied to allow them to determine whether they should withdraw the use of the name from the result of the patch.
Debian on the other hand wants to be able to apply security patches without needing to ask Mozilla for approval, and more importantly perhaps want not to impose such restrictions on their downstreams.
The use of the Ice* names is done to avoid the scenario where a security fix fails to meet with approval, and then the Debian maintainers being faced with the need to do an emergency trademark purge in order to deploy a security fix.
Cheers, Phil.
On Wed, Sep 9, 2015 at 8:59 PM, Philip Hands phil@hands.com wrote:
Luke Kenneth Casson Leighton lkcl@lkcl.net writes:
On Wed, Sep 9, 2015 at 7:42 PM, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
if you're not familiar with or don't clearly understand the difference, look up the history behind why the Debian Team renamed firefox to "iceweasel".
here you go:
https://en.wikipedia.org/wiki/Mozilla_Corporation_software_rebranded_by_the_...
That was nothing to do with copyright, nor attribution.
The problem is the way that Mozilla enforces its trademark.
Mozilla is (fairly reasonably) concerned that people might take one of its trademarked programs, trojan it, and redistribute the result under the name of e.g. Firefox, thus tainting their good name. They therefore reserve the right to specify which code costitutes Firefox, etc. and want sight of any patches that are applied to allow them to determine whether they should withdraw the use of the name from the result of the patch.
so mozilla have a total lack of trust of the debian team. that's the debian team who have software libre's interests, user's interests, their own long-standing reputation (backed up by GPG-signing) to protect, and the mozilla foundation's directors could not see fit to trust such reliable and reputable people to look after something as critical as security patches.
Debian on the other hand wants to be able to apply security patches without needing to ask Mozilla for approval, and more importantly perhaps want not to impose such restrictions on their downstreams.
The use of the Ice* names is done to avoid the scenario where a security fix fails to meet with approval, and then the Debian maintainers being faced with the need to do an emergency trademark purge in order to deploy a security fix.
good for them. sounds like the right decision.
also sounds very much like i quoted _completely_ the wrong example. any other mistakes i made that you can see, phil? :)
l.
Luke Kenneth Casson Leighton lkcl@lkcl.net writes:
On Wed, Sep 9, 2015 at 8:59 PM, Philip Hands phil@hands.com wrote:
Luke Kenneth Casson Leighton lkcl@lkcl.net writes:
On Wed, Sep 9, 2015 at 7:42 PM, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:
if you're not familiar with or don't clearly understand the difference, look up the history behind why the Debian Team renamed firefox to "iceweasel".
here you go:
https://en.wikipedia.org/wiki/Mozilla_Corporation_software_rebranded_by_the_...
That was nothing to do with copyright, nor attribution.
The problem is the way that Mozilla enforces its trademark.
Mozilla is (fairly reasonably) concerned that people might take one of its trademarked programs, trojan it, and redistribute the result under the name of e.g. Firefox, thus tainting their good name. They therefore reserve the right to specify which code costitutes Firefox, etc. and want sight of any patches that are applied to allow them to determine whether they should withdraw the use of the name from the result of the patch.
so mozilla have a total lack of trust of the debian team.
No.
Mozilla has a trademark policy designed to deal with abusers.
Debian has a policy that requires any license to _not_ be exclusive to Debian, because that would cause trouble downstream.
These two things are both reasonable, but sadly incompatible.
that's the debian team who have software libre's interests, user's interests, their own long-standing reputation (backed up by GPG-signing) to protect, and the mozilla foundation's directors could not see fit to trust such reliable and reputable people to look after something as critical as security patches.
None of that is relevant.
Debian on the other hand wants to be able to apply security patches without needing to ask Mozilla for approval, and more importantly perhaps want not to impose such restrictions on their downstreams.
The use of the Ice* names is done to avoid the scenario where a security fix fails to meet with approval, and then the Debian maintainers being faced with the need to do an emergency trademark purge in order to deploy a security fix.
good for them. sounds like the right decision.
also sounds very much like i quoted _completely_ the wrong example. any other mistakes i made that you can see, phil? :)
Since you ask:
https://www.olimex.com/Products/OLinuXino/open-source-hardware
"The Hardware project is released under the Creative Commons Attribution-Share Alike 3.0 United States License."
http://www.gnu.org/licenses/license-list.html#ccbysa
"This is a copyleft free license that is good for ..."
So the olimex boards are under an FSF-approved copyleft license.
I think perhaps you've conflated the word "Attribution" with the BSD 4-clause license (with its obnoxious "Advertising" clause):
http://www.gnu.org/licenses/license-list.html#OriginalBSD
Cheers, Phil.
Mozilla has a trademark policy designed to deal with abusers.
Debian has a policy that requires any license to _not_ be exclusive to Debian, because that would cause trouble downstream.
These two things are both reasonable, but sadly incompatible.
For what it may be worth, my ultimate aim is to get the thing built, and redo it in KiCAD and eventually build it and release it under GPL with full documentation as I'm paying for it. The aim is to have a GPL'd base design out there and watch all those GPL'd derivatives being built to harness it for more projects.
The two or three dollar differences between boards doesn't go towards much in the way of profit today if anyone cares to look into it. So a fully GPL'd kicad design that anyone can customize won't be harming anyone with a head screwed on.
Its what you customise these boards into that matters. For example PCDuino is a wild success because although its a cubieboard equivalent, it has arduino compatibility and is sold as an Arduino on steroids which it is :)
There are a lot more projects like that waiting to happen me thinks.
On Wed, Sep 9, 2015 at 10:43 PM, Philip Hands phil@hands.com wrote:
I think perhaps you've conflated the word "Attribution" with the BSD 4-clause license (with its obnoxious "Advertising" clause):
that sounds about right.
ta for the clarification phil.
l.
On 2015-09-09 at 19:42:25 +0100, Luke Kenneth Casson Leighton wrote:
"provided that you conspicuously and appropriately publish on each copy an appropriate copyright notice; keep intact all notices stating that [...]
so that's in the source code.
also in non source: section 6 says
"You may convey a covered work in object code form under the terms of sections 4 and 5" (section 4 is the one quoted above)
similar wording is also in GPLv2
On Wed, Sep 9, 2015 at 3:48 PM, joem joem@martindale-electric.co.uk wrote:
Hi,
Just wondering how much is it these days ball park to pay someone like witstech or a similar design outfit to design an A20 board
about $USD 3-4k
(like a cubie) and make 10 samples?
about $USD 2.5k
l.
On Wed, 2015-09-09 at 16:02 +0100, Luke Kenneth Casson Leighton wrote:
On Wed, Sep 9, 2015 at 3:48 PM, joem joem@martindale-electric.co.uk wrote:
Hi,
Just wondering how much is it these days ball park to pay someone like witstech or a similar design outfit to design an A20 board
about $USD 3-4k
Thats a good number! Do you think they will hand over the design files? I don't want to spend another 2.5k on top for the 10 samples. (Cheaper companies out there that source and populate.)
(like a cubie) and make 10 samples?
about $USD 2.5k
On Wed, Sep 9, 2015 at 4:08 PM, joem joem@martindale-electric.co.uk wrote:
On Wed, 2015-09-09 at 16:02 +0100, Luke Kenneth Casson Leighton wrote:
On Wed, Sep 9, 2015 at 3:48 PM, joem joem@martindale-electric.co.uk wrote:
Hi,
Just wondering how much is it these days ball park to pay someone like witstech or a similar design outfit to design an A20 board
about $USD 3-4k
Thats a good number! Do you think they will hand over the design files?
they always do.
I don't want to spend another 2.5k on top for the 10 samples.
it was approximate.
... y'know... if you're looking for someone to design an A20 board, i could always do it for you, i have a working proven and very small footprint board.
l.
Hi,
Just wondering how much is it these days ball park to pay someone like witstech or a similar design outfit to design an A20 board
about $USD 3-4k
Thats a good number! Do you think they will hand over the design files?
they always do.
I don't want to spend another 2.5k on top for the 10 samples.
it was approximate.
... y'know... if you're looking for someone to design an A20 board, i could always do it for you, i have a working proven and very small footprint board.
If you are offering, then name a ball park price (send to research@enemygadgets.com) I'm happy to say you deserve it more than anyone else :)
On Wed, Sep 9, 2015 at 4:26 PM, joem joem@martindale-electric.co.uk wrote:
... y'know... if you're looking for someone to design an A20 board, i could always do it for you, i have a working proven and very small footprint board.
If you are offering,
i am :)
then name a ball park price (send to research@enemygadgets.com) I'm happy to say you deserve it more than anyone else :)
:)
done sah.
On Wed, Sep 9, 2015 at 4:26 PM, joem joem@martindale-electric.co.uk wrote:
Hi,
Just wondering how much is it these days ball park to pay someone like witstech or a similar design outfit to design an A20 board
about $USD 3-4k
Thats a good number! Do you think they will hand over the design files?
they always do.
I don't want to spend another 2.5k on top for the 10 samples.
it was approximate.
... y'know... if you're looking for someone to design an A20 board, i could always do it for you, i have a working proven and very small footprint board.
If you are offering, then name a ball park price (send to research@enemygadgets.com) I'm happy to say you deserve it more than anyone else :)
thought about this a bit more: i need 10 samples of a revised EOMA68-A20, after i just changed the standard (again. for the last time. again).
i'll do you a deal: if you can pay for the 10 samples, i'll do the design of an A20 board for you. that way, apart from anything, both of us would get a better deal on sample quantity.
l.
Just wondering how much is it these days ball park to pay someone like witstech or a similar design outfit to design an A20 board
about $USD 3-4k
(like a cubie) and make 10 samples?
about $USD 2.5k
The first batch funding of 1.6 kilo dollars on its way Luke :)
The board I want is designed to run on 18650 or 26650 batteries (very similar to a UPS configuration) and have capacitive touch LCDs.
Its just right for building a million dollar 6 petabyte SSD server that fits in a space of 2 cubic meters or thereabouts consuming 10kW.
You may not want to believe all that right now.. :)
On Tue, Sep 29, 2015 at 9:38 AM, joem joem@martindale-electric.co.uk wrote:
Just wondering how much is it these days ball park to pay someone like witstech or a similar design outfit to design an A20 board
about $USD 3-4k
(like a cubie) and make 10 samples?
about $USD 2.5k
The first batch funding of 1.6 kilo dollars on its way Luke :)
The board I want is designed to run on 18650 or 26650 batteries (very similar to a UPS configuration) and have capacitive touch LCDs.
Its just right for building a million dollar 6 petabyte SSD server that fits in a space of 2 cubic meters or thereabouts consuming 10kW.
You may not want to believe all that right now.. :)
no i get it - i added up once how many CPU Cards you could get into a single rack-mount space, it was absolutely mad.
l.
The board I want is designed to run on 18650 or 26650 batteries (very similar to a UPS configuration) and have capacitive touch LCDs.
Its just right for building a million dollar 6 petabyte SSD server that fits in a space of 2 cubic meters or thereabouts consuming 10kW.
You may not want to believe all that right now.. :)
no i get it - i added up once how many CPU Cards you could get into a single rack-mount space, it was absolutely mad.
Its how to wire all that that is difficult.
I'm building a hypercube (with emphasis on hype :) ) with extravagantly lit LED conduits for the wiring to be passed through to reach all computing elements in a 3D mesh. The idea being if all is well, the conduits are a pleasant colour while if there are problem, the conduits light the way to the problem areas.
On Wed, Sep 30, 2015 at 9:12 AM, joem joem@martindale-electric.co.uk wrote:
no i get it - i added up once how many CPU Cards you could get into a single rack-mount space, it was absolutely mad.
Its how to wire all that that is difficult.
I'm building a hypercube (with emphasis on hype :) ) with extravagantly lit LED conduits for the wiring to be passed through to reach all computing elements in a 3D mesh. The idea being if all is well, the conduits are a pleasant colour while if there are problem, the conduits light the way to the problem areas.
niiice.
btw do consider this as well, it expands up to infinite size: https://en.wikipedia.org/wiki/Nonblocking_minimal_spanning_switch
assuming a 2-bar crossbar as the "base unit", basically to create the next power of 2 up, you:
* duplicate the entire network of switches made so far * put them side-by-side * put in another "layer" * connect every ODD numbered output from the last of the old layers "straight through" * connect every EVEN numbered output to a port CROSS-WISE by adding N/2 to its port number
(N is obviously the current total number of inputs and outputs)
if you have a 4-port router then obviously you would quadruplicate the entire network so far, then connect the first one straight, the second one increase the port number by (N/4), the third by (N/4)*2, the fourth by (N/4)*3
if you have a 32-port hub as the base unit you could go straight to 32x32 nodes (1024) with only 2 layers.
i pretty much guarantee though that regardless of what you do, you'll end up with more wires and more network switches than anything else :)
l.
arm-netbook@lists.phcomp.co.uk
-
Alejandro Mery -
Elena ``of Valhalla'' -
joem -
Luke Kenneth Casson Leighton -
Philip Hands