Debian GNU/Linux on tablet hardware

[Rob van der Hoeven]

Hi,

Recently i bought a very nice tablet with the following specifications:

Chipset : InfoTMIC IMAPX210 ARM11
Speed : 1000 MHz
Memory : 512 MB
Storage : 4 GB.
Expansion : MicroSD™ expandable up to 64 GB total (2 x 32 GB)

Connectivity:

2 x USB host
1 x RJ45 Ethernet (100 MHz)
Mini HDMI 1.3 out
3.5 mm audio out

Full specs:

http://www.yarvik.com/en/products/tablets/tablets_10/TAB410/

I think the hardware of this tablet can also be used as a server or
desktop computer. The tablet is mass produced and very cheap (i got mine
for 149 euro). Maybe we can order only the circuit board and build a
nice Debian GNU/Linux ARM computer around it?

Cheers,
Rob van der Hoeven.

http://freedomboxblog.nl

[Arnaud Patard (Rtp)]

Post by Rob van der Hoeven
Hi,

Hi,

Post by Rob van der Hoeven
Chipset : InfoTMIC IMAPX210 ARM11
Speed : 1000 MHz
Memory : 512 MB
Storage : 4 GB.
Expansion : MicroSD™ expandable up to 64 GB total (2 x 32 GB)

I may be wrong but iirc, imapx210 is yet an other cheap cinese arm SoC
with *no* support in mainline kernel. This means that you may be able to
use a chroot or you can try booting debian armel with the kernel
provided on the system but not use a full debian armel (kernel+userspace
and maybe uboot). Expect also some troubles if it requires proprietary
drivers to work.
I can't say more as the way to do it and troubles you'll get into are
system specifics.

Post by Rob van der Hoeven
2 x USB host
1 x RJ45 Ethernet (100 MHz)
Mini HDMI 1.3 out
3.5 mm audio out
http://www.yarvik.com/en/products/tablets/tablets_10/TAB410/
I think the hardware of this tablet can also be used as a server or
desktop computer. The tablet is mass produced and very cheap (i got mine
for 149 euro). Maybe we can order only the circuit board and build a
nice Debian GNU/Linux ARM computer around it?

For that price, to make a server, I would rather buy a loco board or any
other development board rather than relying on hardware with a SoC
without manual and support in mainline kernel. For desktop, things may
be more complicated as most gpu on arm SoC are relying on proprietary
drivers for 2d/3d.


Arnaud

[Luke Kenneth Casson Leighton]

On Fri, Oct 28, 2011 at 12:57 PM, Arnaud Patard

Post by Arnaud Patard (Rtp)

Hi,

Hi,

Chipset   : InfoTMIC IMAPX210 ARM11
Speed     : 1000 MHz
Memory    : 512 MB
Storage   : 4 GB.
Expansion : MicroSD™ expandable up to 64 GB total (2 x 32 GB)

I may be wrong but iirc, imapx210 is yet an other cheap cinese arm SoC
with *no* support in mainline kernel.

i have the source code. i may have placed it into the arm-netbooks
alioth repository, it was a while back. quotes illegal quotes copies
of the infotmic source code can be downloaded from the usual quotes
illegal quotes web sites. i say quotes illegal quotes because the
fucking morons at infotmic forced everyone to sign NDAs before
receiving the BSP (including kernel source code) thus losing their
rights to actually distribute their *own* BSPs under the terms of the
GPLv2.

it's a dog's dinner situation, basically.

l.

[Bill Gatliff]

Post by Luke Kenneth Casson Leighton
i say quotes illegal quotes because the
fucking morons at infotmic forced everyone to sign NDAs before
receiving the BSP (including kernel source code) thus losing their
rights to actually distribute their *own* BSPs under the terms of the
GPLv2.

Oh, the irony of signing an NDA with a Chinese company!

Pardon me, I think my head just exploded...

b.g.

[Luke Kenneth Casson Leighton]

Post by Bill Gatliff
Oh, the irony of signing an NDA with a Chinese company!
Pardon me, I think my head just exploded...

ha ha :) yeah... unfortunately the company i'm associated with has
to actually respect things like that, regardless of whether it's
actually enforceable, because the scope of what we're doing is
international: the rules and regulations are just... nuts.

l.

[Rob van der Hoeven]

Post by Arnaud Patard (Rtp)

Post by Rob van der Hoeven
Chipset : InfoTMIC IMAPX210 ARM11
Speed : 1000 MHz
Memory : 512 MB
Storage : 4 GB.
Expansion : MicroSD™ expandable up to 64 GB total (2 x 32 GB)

I may be wrong but iirc, imapx210 is yet an other cheap cinese arm SoC
with *no* support in mainline kernel. This means that you may be able to
use a chroot or you can try booting debian armel with the kernel
provided on the system but not use a full debian armel (kernel+userspace
and maybe uboot). Expect also some troubles if it requires proprietary
drivers to work.
I can't say more as the way to do it and troubles you'll get into are
system specifics.

Hi Arnaud,

I'm new to the ARM and GNU/Linux world and would love to learn more
about the issues you mention. Any good links?

At the moment i'm working on a (my) FreedomBox, for this i have upgraded
my QNAP TS119 NAS with Debian GNU/Linux. Works like a charm!

Post by Arnaud Patard (Rtp)

Post by Rob van der Hoeven
I think the hardware of this tablet can also be used as a server or
desktop computer. The tablet is mass produced and very cheap (i got mine
for 149 euro). Maybe we can order only the circuit board and build a
nice Debian GNU/Linux ARM computer around it?

For that price, to make a server, I would rather buy a loco board or any
other development board rather than relying on hardware with a SoC
without manual and support in mainline kernel. For desktop, things may
be more complicated as most gpu on arm SoC are relying on proprietary
drivers for 2d/3d.

These boards are not mass produced which makes them relatively
expensive. Hardware that is not mass produced has some other issues,
namely availability and vendor lock-in (At the moment the FreedomBox
Foundation is using hardware from GlobalScale Technologies for their
prototype development. Things are not going smooth...)

I think it must be possible to buy an android motherboard for just a
fraction of the price that i paid for my tablet. Why is relying on
hardware with a SoC such a bad idea? If the SoC is popular it will not
go out of production for a long time. A motherboard without a SoC can
replace any IO chip with an undocumented chip for which there is no
driver.

B.T.W. I fully agree that a SoC for which there is no
manual/documentation or one that comes with an NDA is unacceptable.

Friendly greetings
Rob van der Hoeven

http://freedomboxblog.nl

[Phil Endecott]

Post by Rob van der Hoeven

Post by Arnaud Patard (Rtp)

Post by Rob van der Hoeven
I think the hardware of this tablet can also be used as a server or
desktop computer. The tablet is mass produced and very cheap (i got mine
for 149 euro).

For that price, to make a server, I would rather buy a loco board or any
other development board

These boards are not mass produced which makes them relatively
expensive.

The i.MX LOCO board, the OMAP panda board, and some of the others cost about the
same as your tablet.

Post by Rob van der Hoeven
Hardware that is not mass produced has some other issues,
namely availability and vendor lock-in

You think that your tablet is going to have better availability and less lock-in
than a board from Freescale or TI? That seems unlikely to me. Look at the
BeagleBoard; it would be hard to find any smartphone or tablet device that has
been available for as long as that has.

Post by Rob van der Hoeven
I think it must be possible to buy an android motherboard for just a
fraction of the price that i paid for my tablet.

Why do you think that? I have personally never seen an "Android motherboard"
offered for sale at all, let alone for a low price.

Post by Rob van der Hoeven
Why is relying on
hardware with a SoC such a bad idea? If the SoC is popular it will not
go out of production for a long time.

No, that's not how it works. Both popular and unpopular chips are replaced on a
schedule that's determined by advances in manufacturing technology. This also
applies to the consumer products that are made from them: even if a device is
popular, it will soon be replaced with something that is faster and cheaper.


Regards, Phil.

[Lennart Sorensen]

Post by Phil Endecott
Why do you think that? I have personally never seen an "Android motherboard"
offered for sale at all, let alone for a low price.

The i.MX53 LOCO board has an android BSP available. Does that qualify?

[Phil Endecott]

Post by Lennart Sorensen

Post by Phil Endecott
Why do you think that? I have personally never seen an "Android motherboard"
offered for sale at all, let alone for a low price.

The i.MX53 LOCO board has an android BSP available. Does that qualify?

It does to me, but Rob doesn't seem to like these "not mass produced",
"relatively expensive" boards; I think his idea of an "android
motherboard" is something from a no-name Chinese tablet manufacturer.


Regards, Phil.

[Luke Kenneth Casson Leighton]

[for the benefit of the freedombox list, cc'd here: a discussion is
taking place - again - on debian-arm about the cost and availability
of ARM-based systems. a chipset has been found which meets eben's
criteria in mass-volume. actually, is a whopping 25% _below_ eben's
expectations of $20, and is 40% lower than the raspberrypi price,
with, being a 1ghz Cortex A8 instead of a 700mhz ARM11, roughly 2 to
2.5x better performance.]

On Fri, Oct 28, 2011 at 4:07 PM, Phil Endecott

Post by Phil Endecott

Why do you think that?  I have personally never seen an "Android
motherboard"
offered for sale at all, let alone for a low price.

The i.MX53 LOCO board has an android BSP available.  Does that qualify?

It does to me, but Rob doesn't seem to like these "not mass produced",
"relatively expensive" boards; I think his idea of an "android motherboard"
is something from a no-name Chinese tablet manufacturer.

i am in touch with such a "no-name" chinese tablet manufacturer. i
can negotiate one of those "into existence"... *if* people are willing
to commit - publicly so that i can refer the factory to the archives -
to buying some.

let's work through the costs, and i'll ask the factory to confirm,
but it'll be roughly something like this:

* the mass-production BOM of a 1ghz Cortex A8 system using the
chiprise aka bmorn aka allwinner SoC will be - literally - around the
$15 mark.

* morphing the "Reference Board Design" into a "FreedomBox-like"
device will be about 10 days @ $150 per day (NOT $150 per hour, REALLY
$150 per day).

* back-caclulating that cost to include it in to the first.. say...
100 units, that's $15 extra per unit.

* scaling up on the BOM to 100-only quantities, you prooobbbably need
to add about 60% to 70%.

* $15 * 1.7 = $25.50 plus $15 amortised over 100 units is $40.

* add a reasonable profit margin - say 25% (that's *really* generous
btw), that's now $50.

* $50 plus shipping, tax and a box, that's about $80 to get it into your hands.

$80 is still a f*** of a lot lower than $125 *excluding* shipping for
a beagleboard.

and, once it hits mass-production and goes world-wide, that $80
becomes about $35, and if you do a "Special FreedomBox Special Order"
of 100,000 units, that $35 becomes about $25 *including* shipping and
tax.

so it's up to you, folks. if you want to see something like this
happen, you have to say so.

btw - if you want to take over negotiations with the factory, i am
quite happy to put anyone who wishes to take over in touch with the
factory. i will give you their email address, you can talk directly
with them.

l.

[Rob van der Hoeven]

Post by Phil Endecott

Post by Lennart Sorensen

Post by Phil Endecott
Why do you think that? I have personally never seen an "Android motherboard"
offered for sale at all, let alone for a low price.

The i.MX53 LOCO board has an android BSP available. Does that qualify?

It does to me, but Rob doesn't seem to like these "not mass produced",
"relatively expensive" boards; I think his idea of an "android
motherboard" is something from a no-name Chinese tablet manufacturer.

Mass produced hardware has a higher chance of being of good quality.
There is more money involved and a failure can be very expensive for a
brand or manufacturer.

Regards,
Rob van der Hoeven

http://freedomboxblog.nl

[Phil Endecott]

Post by Rob van der Hoeven
Mass produced hardware has a higher chance of being of good quality.

Absolute Rubbish. You are on a different planet.

[Luke Kenneth Casson Leighton]

On Fri, Oct 28, 2011 at 8:54 PM, Phil Endecott

Post by Rob van der Hoeven
Mass produced hardware has a higher chance of being of good quality.

Absolute Rubbish.  You are on a different planet.

oo, ouch phil. i did try to be tactful and explain! quite amusing -
normally it's me who says that, and i really _do_ come from a
different planet (*1)

l.

(*1 - no really. sadly i can't tell you where it is, due to the
mind-wipe that's inherent in the technical process known on earth as
"being born". for the first few seconds here on earth i could
probably have told you, if you understood what "ga ga" means)

[Rob van der Hoeven]

Post by Phil Endecott

Post by Rob van der Hoeven
Mass produced hardware has a higher chance of being of good quality.

Absolute Rubbish. You are on a different planet.

Mass production is expensive to start
Mass production is even more expensive when it fails

Greeting from the Earth.

Rob van der Hoeven
http://freedomboxblog.nl

[Luke Kenneth Casson Leighton]

On Fri, Oct 28, 2011 at 9:20 PM, Rob van der Hoeven

Post by Rob van der Hoeven

Post by Rob van der Hoeven
Mass produced hardware has a higher chance of being of good quality.

Absolute Rubbish.  You are on a different planet.

Mass production is expensive to start

rob. i just went to a lot of trouble to describe to you the
circumstances which i've spent nearly two years negotiating with
factories and also communicating directly with SoC vendors to find out
how they operate, which prove that *if* you are willing to do a deal,
mass production is *not* expensive to start.

to give you an idea of what it feels like, for me, to read that
blanket statement, i am actually feeling acute vertigo. the reason
for that acute vertigo is because i am trying to believe you, instead
of my own eyes and ears reading messages for a large fraction of my
life.

the acute vertigo is caused because you are asking me to believe that
the last two years of my life were entirely a dream - a fantasy - that
the people who told me that they could take the hardware design from
the SoC vendor and cut/paste it "at cost", in return for nothing more
than a committment by me that i will guarantee that they will get the
software somehow (even if i have to do it myself) do not exist.

now, whilst i'm trying to make a joke about it i'm actually dead
serious about the acute vertigo.

so i apologise if i haven't been able to get across to you that this
is real - it's not a fantasy. i'll leave it at that, because any
attempt to make jokes after that (i wrote several, and deleted them
all) appear to be "flippant" or "sarcastic".


perhaps, as is more likely, there is a misunderstanding, for which i
again apologise, because this is quite a complex bit of negotiation
that i'm involved in.

perhaps you may have misunderstood one of several things:

* that the factory is developing the entire hardware "from scratch" -
they are not.

* that the factory is using European or USA $150 per hour prices and
employees who wish to be paid $50 per hour - they are not.

* that the factory is supplying the software - they are not.

* that the factory is designing and supplying casework - they are
not: it's off-the-shelf casework.


now if any of those 4 things were part of the assumptions that you've
made, you would be absolutely correct.

* casework (see the OpenPandora project blogs for a good illustration)
typically costs upwards of $100,000 even using china factories. if
you _really_ want to get the price down, you have to pay a chinese
employee cash - about $7,000 will normally do it - to "moonlight", use
his employers' facilities and risk getting fired (he can always go
work at another factory, and he can do so once that $7000 gets a bit
low. c'est la vie...)

* a complete "from scratch" design typically takes 100 to 150 hours
(or access to software which costs $250,000 per week to license and
operate) - either way, you're looking at an insane amount of money
and, yes, taking an enormous risk.

i'll stop there, because it would i feel be counterproductive to keep
illustrating: i apologise for that but if anyone is interested i can
write further (off-list).

so i trust that we have a simple misunderstanding that you thought i
was talking about something different, when in fact i am talking about
taking advantage of some very very specific circumstances, criteria
and opportunities, all of which add up to a deal which does *not*
involve large sums of money changing hands, for any party involved.

l.

[Luke Kenneth Casson Leighton]

On Fri, Oct 28, 2011 at 9:20 PM, Rob van der Hoeven

Post by Rob van der Hoeven

Post by Rob van der Hoeven
Mass produced hardware has a higher chance of being of good quality.

Absolute Rubbish.  You are on a different planet.

Mass production is expensive to start
Mass production is even more expensive when it fails

rob. you - and anyone else who believes the above - need to read the
following:

http://quickembed.com/Tools/Shop/ARM/200908/43.html

The advantage of SBC is that when you define a new product, you have
only small change on baseboard/motherboard, no change on SBC/daughter
board, it saves great effort, and save much expense as the SBC is big
lots produced while baseboard can be produced in small lots, also this
helps improving the quality of your products. More detail read here.

where "here" is an article from just _one_ very very experienced PCB
design company.
http://www.quickembed.com/forum/viewtopic.php?f=6&t=7

* directinsight produce split-level modules for most of their iMX range.
* colibri produce split modules.
* cogcomp.com produce split modules.
* ziilabs supply their CPUs on an SO-DIMM split module.
* hardkernel.com produce split modules.

in the embedded world, the technique of splitting out the CPU+RAM+NAND
into its own separate board, thus reducing both cost and risk has been
done again, and again, and again.

l.

p.s. of course in the x86 world, with the ridiculous
northbridge-southbridge architecture and the insane power
requirements, it's completely impossible to do a full CPU+RAM+NAND
Flash in under 5 watts, let alone under 2, even if you could fit all
those ICs onto a 5cm x 7cm PCB.

[Bill Gatliff]

Post by Phil Endecott
The i.MX LOCO board, the OMAP panda board, and some of the others cost about the
same as your tablet.

Post by Rob van der Hoeven
Hardware that is not mass produced has some other issues,
namely availability and vendor lock-in

You think that your tablet is going to have better availability and less lock-in
than a board from Freescale or TI? That seems unlikely to me. Look at the
BeagleBoard; it would be hard to find any smartphone or tablet device that has
been available for as long as that has.

Post by Rob van der Hoeven
I think it must be possible to buy an android motherboard for just a
fraction of the price that i paid for my tablet.

Why do you think that? I have personally never seen an "Android motherboard"
offered for sale at all, let alone for a low price.

Yes, you have: LOCO, Panda, Beagleboard, etc.. The OP disagrees due to an
unclear definition of "motherboard".

And the nice thing about the aforementioned boards is, they come with
schematics. That erases the stability and availability problems by
significantly lowering the barriers to producing the boards yourself if the
need arises.

BeagleBoard is built by CircuitCo, and I'm sure they would be pleased to
quote any similar board you like. Getting boards is NOT the problem,
overcoming the psychology preventing one from asking for the boards is the
problem.

b.g.

[Luke Kenneth Casson Leighton]

Post by Bill Gatliff

Post by Rob van der Hoeven
I think it must be possible to buy an android motherboard for just a
fraction of the price that i paid for my tablet.

only if you're prepared to place a cash up-front order for a large
number of units, because you're dealing "direct" with the factory.
the exceptions to that, that i know of, will be companies like the one
i'm dealing with (because they trust me), and franson who runs
quickembed would almost certainly be happy to sell you an
off-the-shelf motherboard from something he's designed (or licensed)
already.

Post by Bill Gatliff

Why do you think that?  I have personally never seen an "Android
motherboard"
offered for sale at all, let alone for a low price.

Yes, you have: LOCO, Panda, Beagleboard, etc.. The OP disagrees due to an
unclear definition of "motherboard".
And the nice thing about the aforementioned boards is, they come with
schematics. That erases the stability and availability problems by
significantly lowering the barriers to producing the boards yourself if the
need arises.

well... that's only if there's no modifications required. IGEPv2 and
DEVKIT8000 are "clones" of the beagleboards: you can't get *their*
schematics.

the problem with using the pandaboard, beagleboard etc. is that the
components are *not* available at reasonable prices. the 1k price of
the OMAP3530 is... well, you can check for yourself on the web site.
and then these boards use POP RAM (package-on-package) which is
high-density RAM, is *not* a mass-produced part and thus is insanely
expensive (relatively speaking).

the project i'm working on, our goal is to only use parts that are
mass-produced, large volume, and thus the cost is jaw-droppingly low.
hence it's really important to be able to point this little factory at
the debian archives to show them "yes, there are people who are
interested in buying, here they are".

it's up to you, people. if you want a low-cost freedom box, or a
low-cost server, or a low-cost tablet or a low-cost nettop or whatever
but, unlike many offerings, with an actual decent CPU that won't have
you regretting spending the money, you have to say so, publicly, so
that i can refer the factory to the archive posts so that they have
the confidence to go ahead *without* you needing to put down any money
up-front.

alternatively, you can place an order with the factory directly, pay
their fees up-front (which aren't high) etc. etc. - i mention this
purely as an option which i'm sure you wouldn't _really_ want to do,
but i have to mention it for completeness.

alternatively, i'm quite happy to put people directly in touch with
the factory: anyone who wants to take over the negotiations and
maintain a dialog with them is entirely at liberty to do so. i will
then be able to buy whatever product results from the efforts of your
work.

l.

[Rob van der Hoeven]

Post by Luke Kenneth Casson Leighton

Post by Rob van der Hoeven
I think it must be possible to buy an android motherboard for just a
fraction of the price that i paid for my tablet.

only if you're prepared to place a cash up-front order for a large
number of units, because you're dealing "direct" with the factory.
the exceptions to that, that i know of, will be companies like the one
i'm dealing with (because they trust me), and franson who runs
quickembed would almost certainly be happy to sell you an
off-the-shelf motherboard from something he's designed (or licensed)
already.

What is needed for the FreedomBox is hardware that is already used by a
major brand. This hardware can be more trusted with regard of quality
and availability. It is not needed to place orders before testing. We
can just get the motherboard from a device that we buy. If things go
well, then we can contact the manufacturer.

I don't think we should try to specify our own hardware. Too risky for
me. We should liberate existing hardware from the android :-)

Regards,
Rob van der Hoeven

http://freedomboxblog.nl

[Luke Kenneth Casson Leighton]

On Fri, Oct 28, 2011 at 7:34 PM, Rob van der Hoeven

Post by Rob van der Hoeven

Post by Rob van der Hoeven
I think it must be possible to buy an android motherboard for just a
fraction of the price that i paid for my tablet.

 only if you're prepared to place a cash up-front order for a large
number of units, because you're dealing "direct" with the factory.
the exceptions to that, that i know of, will be companies like the one
i'm dealing with (because they trust me), and franson who runs
quickembed would almost certainly be happy to sell you an
off-the-shelf motherboard from something he's designed (or licensed)
already.

What is needed for the FreedomBox is hardware that is already used by a
major brand. This hardware can be more trusted with regard of quality
and availability. It is not needed to place orders before testing. We
can just get the motherboard from a device that we buy. If things go
well, then we can contact the manufacturer.

rob - you're not aware of the realities of the situation behind the
factories. the factories DO NOT HAVE the source code.

i repeat.

the factories do NOT have the source code.

you're therefore asking too much, and expecting too much, on several counts.

not only do the factories not have the source code, but they don't
actually do the full designs themselves. they *modify* designs (just
like the guys at quickembed modified the beagleboard "Open Source" PCB
designs to create the DEVKIT8000 product).

this is why certain companies can supply you with samples at very low
prices, whilst others want $150,000 up-front before you even start.

Post by Rob van der Hoeven
I don't think we should try to specify our own hardware. Too risky for
me. We should liberate existing hardware from the android :-)

rob, feel free. you will then end up with a device that is not only
completely unsuited to the task, because it's in the wrong
form-factor, but by the time you've finished negotiating (or
reverse-engineering) it will be end-of-lifed.

or, worse, superceded by a cheaper chipset, which *causes* everyone
to run like hell towards it, abandoning everything else that's even
vaguely comparable.

however: the point about needing testing is valid (in some ways) and
not valid in others.

the ways in which it is valid is if the SoC vendor themselves makes a
mistake. this has actually happened. three times that i know of.
the first was with the OMAP3530 (beagleboard). the second was with
the sPEAR1310. the third was with the OMAP4440 (pandaboard). ask
around: those people who got early pandaboards had to be provided with
replacements. in each case, there were errors *in the CPU*.

now, this is easily "solved", by picking a CPU that is in
mass-production that has, clearly by virtue of it being in
mass-production, has no errors. if it does, you can sue the SoC
vendor. so that's not a problem.

another way is that the factory makes a hardware mistake. this is
"their problem" and "their responsibility to fix". given the amounts
of money that they are likely to make, if you're "in early" and are
therefore helping them out to do software development, they will go
out of their way to get you a replacement board (revision N+1),
because you, in return, are helping them [remember: normally they have
to pay cash up-front for the ODM ready-resigned PCB as well as the GPL
software _and_ get themselves into GPL Hell over the Linux Kernel
_and_ the OS!]

so _that's_ not a problem.

however that does leave the case where there are people who would
just like to buy very-early-revision hardware, which may or may not
work, but who do not want to actually work on the software. such
people i am *not* interested in. at all. if you are one such person,
who is *solely* interested in obtaining early-revision-hardware and
who does not want to help get the product working, you need not read
any further.

so leaving out that small group of people, we can discard them and
move on to the next group.

we're now into the scenario where not only are we using a "reliable"
SoC, but also that small group of dedicated people have worked on the
software enough to be able to "prove" that the hardware works
reliably, and we can now proceed to the scenario of offering "beta"
versions to a wider group.

this is all absolutely standard practice, btw - goldelico are
following this procedure to the letter, for the GTA04 development.

so let me rephrase the question.

if there was a "heavily-beta-tested" PCB (as illustratively defined
above) which, in mass-volume would be around the $15 mark, was made
available with at least one working linux kernel and at least one
GNU/Linux OS, at an approximate guide-line price of $70 to $80 for the
first 100 and potentially dropping even further as sales picked up,
would anyone be interested enough to commit publicly to saying "yeah i
could go for that", such that i can then show the factory that there
is interest, such that it is not necessary to pay them up-front, cash,
for the PCB development? oh, proviso, of course: only if there are
enough *other* people interested.

also please bear in mind that the design i have in mind will be
EOMA/PCMCIA compliant. thus, far from being "problematic" if the CPU
card is quotes wrong quotes, actually it can just be... replaced.
and, precisely because it is EOMA/PCMCIA compliant, the exact same CPU
card (which is where most of the risk and hard work is) can be plugged
into a mass-produced tablet, or a FreedomBox chassis, or a NetTop,
or... whatever.

so, if you are in that group - who would like a heavily-beta-tested
PCB - please do say so, publicly, so that i can point the factory at
the responses. if you know of anyone or any group who would also be
interested, please do also contact them, asking them to either send me
a message with permission to pass it on to the factory, or ask them to
post to the debian-arm list, or subscribe and post to arm-netbooks.

remember that this is about gaining traction. traction cannot be
gained if there is nobody else willing to commit publicly in order to
gain traction.

i can, if you prefer, use a site like "pledgebank.com" - that might
help people to understand that i'm not asking people for a
"committment to buy". i'm asking people if they would be *willing* to
commit to buying, if a certain threshold of other people also publicly
made the exact same committment.

btw if anybody else would like to take over these negotiations, please
feel free to make yourself known and i will pass responsibility on to
you.

l.

(EOMA/PCMCIA specification:
http://elinux.org/Embedded_Open_Modular_Architecture/PCMCIA )

[Philippe Clérié]

Post by Luke Kenneth Casson Leighton
i can, if you prefer, use a site like "pledgebank.com" - that might
help people to understand that i'm not asking people for a
"committment to buy". i'm asking people if they would be *willing* to
commit to buying, if a certain threshold of other people also publicly
made the exact same committment.

At the prices you are talking about I am willing to commit to buy. I have a
couple of questions though.

1- I am not clear on the design you have in mind. Personally, I have three
use-cases: router, NAS, sound server (not Media server, I don't care about
video). I don't necessarily need them all in one system. Does your design
cover these uses?

2- I presume that you are talking about committing to quantities of one or
two systems. Could you please confirm? I really just need a couple of very
low power systems that I can setup and forget!

I've had occasion to deal with people who deal with Chinese manufacturers,
but in fields far removed from electronics. What you describe does sound
familiar.
--
Philippe

------
The trouble with common sense is that it is so uncommon.
<Anonymous>

[Luke Kenneth Casson Leighton]

Post by Philippe Clérié

Post by Luke Kenneth Casson Leighton
i can, if you prefer, use a site like "pledgebank.com" - that might
help people to understand that i'm not asking people for a
"committment to buy".  i'm asking people if they would be *willing* to
commit to buying, if a certain threshold of other people also publicly
made the exact same committment.

At the prices you are talking about I am willing to commit to buy.

fantastic to hear, philippe.

Post by Philippe Clérié
I have a
couple of questions though.
1- I am not clear on the design you have in mind. Personally, I have three
use-cases: router, NAS, sound server (not Media server, I don't care about
video). I don't necessarily need them all in one system. Does your design
cover these uses?

short answer: yes.... depending on what performance you're expecting.

the limit is actually due to having to pick "lowest common
denominator" interfaces for the EOMA/PCMCIA "meta-interface" aka
Standard.

but, the thing is that the "Lowest Common Denominator" is actually
incredibly high. not "mad" high, but still damn good.

* 10/100 ethernet has been around for... forever
* USB2 has been around for over a decade
* SATA-II likewise - decade.
* I2C i don't even know
* 24-pin RGB/TTL you don't care about, but it's there.

now, on *some* SoCs, there is no SATA-II port on-board, so this has to
be done with a USB-to-SATA converter. according to some quick google
searches (and based on empirical observation of these "Set-Top-Box"
style CPUs), the Chiprise aka BoxChip aka BMorn aka AllWinner CPU
_should_ have on-board SATA-II....

http://www.ebay.com/itm/1080P-HDMI1-3-2-5-SATA-HDD-Media-Player-DTS-H-264-NY52-/220879009813?pt=LH_DefaultDomain_0&hash=item336d6a3815

yep, all signs so far are that it's a built-in SATA-II. i'd fall off
my chair in shock if it doesn't have 10/100 Ethernet

(that's the best i can do without having access to the datasheet
because this is a STB-style CPU, they just... don't... hand... out...
datasheets... plus despite all the names, we can't find the real
company or its web site, in order to call them! the factory's on the
case, they'll get them - eventually - even if it's by word-of-mouth
inside china).

Post by Philippe Clérié
2- I presume that you are talking about committing to quantities of one or
two systems.

yes, that's right: it'll be a 1-or-2 price but that 1-or-2 price will
still be lower by virtue of being part of a "group order". this is...
unavoidable, but i think you'll find that even "worst case" it'll
still be well below any cut-off point you might have in mind [except
possibly the situation where only 10-20 units are wanted - then yep,
the sums definitely don't add up, go figure].

issues like putting into boxes that would get split at customs (to
save a bit of $), well... the percentages aren't entirely worth it,
because the value of the goods is so small as in _legitimately_ so
damn small not "let's put a fake value on the goods" small, it's...
yeah, hardly worth doing. if there were 500 units shipped world-wide,
now that's a different story.

Post by Philippe Clérié
Could you please confirm? I really just need a couple of very
low power systems that I can setup and forget!

:)

yes, i do mean 1 or 2. the weight's going to be tiny (please, _do_
source your own PSU, it's just not worth putting one in the box and
paying extra for the size and the weight), the price low so the amount
of customs duty is going to be low as well.

actually, my major concern is that customs in each country won't
_believe_ it's a fully-functioning computer with a replacement value
so damn small, it'll get "red flagged". it's probably a good idea
that this is pre-empted, and a letter put on the outside of each box,
for Customs to read.

btw... from experience: anyone in the UK, yes you can risk using
ParcelForce, if you don't mind that they charge a £13 flat-rate
"handling" fee on VAT collection. bastards. you would do well to
follow the advice of this guy. by following it you can successfully
tell ParcelForce to get stuffed, _and_ still get the goods:

http://www.consumeractiongroup.co.uk/forum/showthread.php?94811-Parcelforce-Fees-for-Collecting-VAT&p=3593303&viewfull=1#post3593303

there are also additional posts on that forum where Fedex have passed
over collection of "their" quotes fee quotes to Debt Collection
Agencies, but a simple letter to the DCA involved, stating that there
is no contract between Fedex and yourself regarding this "fee",
results in the DCA involved giving Fedex absolute hell for wasting
their time.

the important thing is to dispute the fees, immediately, and go from
there. there's some empirical evidence to support the hypothesis that
ParcelForce actually _delete_ the invoice (like, literally wipe the
SQL entry) from their database if you cause enough fuss. this has me
sufficiently alarmed to entertain the possibility that ParcelForce may
actually be committing fraud, or their Directors prosecuted for
extortion.

Post by Philippe Clérié
I've had occasion to deal with people who deal with Chinese manufacturers,
but in fields far removed from electronics. What you describe does sound
familiar.

mwahhhh :)

i ain't got to the bit about having to send some along to check that
the containers don't contain bricks, yet *lol*. i know someone who's
quite happy to go to china and literally kick their desks over until
they refund the money, if that happens, but it generally only occurs
on the really really big orders (otherwise it's not worth the thieves
time!). it'd serve them right for employing chinese shipping agents
who thought it was ok to embezzle from gwailo foreigners. but, yeah,
that only happens if you use untrustworthy shipping agents. the
lesson here is: don't use untrustworthy shipping agents!

l.

[Dominique Dumont]

Post by Luke Kenneth Casson Leighton
it's up to you, people. if you want a low-cost freedom box, or a
low-cost server, or a low-cost tablet or a low-cost nettop or whatever
but, unlike many offerings, with an actual decent CPU that won't have
you regretting spending the money, you have to say so, publicly, so
that i can refer the factory to the archive posts so that they have
the confidence to go ahead without you needing to put down any money
up-front.

I hope I'm not too late following up on this thread.

I'm interested. I pledge to buy at least one device.

In fact, I'm looking for a digital photo frame device. A device like your
efika smartbook minus the keyboard, camera but with a RC5 infrared receiver
would be a great opportunity.

All the best

Dominique
--
http://config-model.wiki.sourceforge.net/ -o- http://search.cpan.org/~ddumont/
http://www.ohloh.net/accounts/ddumont -o- http://ddumont.wordpress.com/

[Dominique Dumont]

Hello Luke

about the digital photo frame idea: that's an even simpler version of
a tablet, effectively. are you looking to build it yourself? if so,
below

No really build it myself.

First, we need a prototype to develop our software on it. The proto can be any
ARM based card plugged to a screen. We're currently playing with a beagleboard
and an igep 0020 card. We can can plug an USB wifi and a USB IR receiver to
play with (may be a home theater remote like this [1]).

Then, we'd need to field test our device. This could be done with netbooks or
small tablets, but not with a bunch of boards and wires ;-) . We'd need ~ 10
of them to lend to family and neighbours and get feedbacks.

Once this is done, we'll need a small serie (~ 100 depending on price) to
really get serious.

Regarding the device, we'd need something quite standard. The screen should be
at least 800x600. Bigger would be better, but the device price for the small
serie will be the main limiting factor.

Regarding user input, I'm wondering if a touchscreen would not be better than
buttons behind the frame. Navigating in a menus with these rear buttons is not
user friendly. But buttons are needed in case customer loose the remote
control. May be a touchscreen as a fallback is better for users. Still
scratching head there.

the right kind of low-speed, low-resolution LCD panel (800x600,
800x480, 480x320 etc.) you could almost connect it up directly to the
EOMA-PCMCIA connector. if you didn't need a brightness control, that
would bring the cost down as well. the IR receiver you can get as a
USB device. if you kept it _that_ simple you'd hardly need a PCB at
all!

That would be better: I've not done HW design for a while (~ 20 years).

All the best

Dominique

[1] http://www.ldlc.com/fiche/PB00101489.html
--
http://config-model.wiki.sourceforge.net/ -o- http://search.cpan.org/~ddumont/
http://www.ohloh.net/accounts/ddumont -o- http://ddumont.wordpress.com/

[Luke Kenneth Casson Leighton]

Post by Dominique Dumont
Hello Luke

about the digital photo frame idea: that's an even simpler version of
a tablet, effectively.  are you looking to build it yourself? if so,
below

No really build it myself.

great!

batteries / power. ah. when you say "digital photo frame", do you
mean "portable device" or do you mean "put it on the wall or
mantlepiece"?

Post by Dominique Dumont
Then, we'd need to field test our device. This could be done with netbooks or
small tablets, but not with a bunch of boards and wires ;-) . We'd need ~ 10
of them to lend to family and neighbours and get feedbacks.

ok, well we have about 17 committed so far. the hardware development
costs (PCB, tooling) are about $USD 2,000. with 10 more, that would
be about $75 each for one (fully-populated, fully-functional but with
no external case) CPU card.

Post by Dominique Dumont
Once this is done, we'll need a small serie (~ 100 depending on price) to
really get serious.

ok, once the first $2000 of NREs is done, that's it: it doesn't have
to be paid again (NREs: non-recurring expenses. go figure).
subsequent PCBs therefore can be much reduced pricing.

Post by Dominique Dumont
Regarding the device, we'd need something quite standard. The screen should be
at least 800x600. Bigger would be better, but the device price for the small
serie will be the main limiting factor.
Regarding user input, I'm wondering if a touchscreen would not be better than
buttons behind the frame. Navigating in a menus with these rear buttons is not
user friendly. But buttons are needed in case customer loose the remote
control. May be a touchscreen as a fallback is better for users. Still
scratching head there.

an overlay resistive touchscreen is about $5, typically. they do
however require calibration: they're also sensitive to temperature,
humidity etc.

capacitive ones made of glass are *MORE* expensive than the
same-sized LCD! absolutely ridiculous, but that's the way it goes.
they're also heavier (because they have to be structurally strong).
there do exist plastic capacitive touchpanels, they're still in
development, the price _will_ come down... at some point.

bottom line is: touchscreens sound great until you look into the cost
and useability.

there are some options however for very coarse-grained touchscreens.
if you can get away with "button-sized" fixed areas (such as used on
microwave ovens) then costs are dramatically reduced.

Post by Dominique Dumont

the right kind of low-speed, low-resolution LCD panel (800x600,
800x480, 480x320 etc.) you could almost connect it up directly to the
EOMA-PCMCIA connector.  if you didn't need a brightness control, that
would bring the cost down as well.  the IR receiver you can get as a
USB device.  if you kept it _that_ simple you'd hardly need a PCB at
all!

That would be better: I've not done HW design for a while (~ 20 years).

fortunately, electrons haven't changed charge or anything like that:
the speed of light is still the same constant, as best we are aware,
and V still equals I times R :)

l.

p.s. i've successfully compiled the allwinner a10 reference source
code, using the emdebian cross-compile toolchain.

[Rüdiger Leibrandt]

As for Buttons and I/O:
My experiences with digital photoframes that I reworked using regular
wooden frames are limited ( 2 units only ) but placing the buttons on
the lower underside and using those metal contacts which you bridge with
your finger ( you know, those TTL/CMOS Touchbuttons like on old TV-Sets
) is a great way.
You make a row of nails as touchbuttons all along the side or bottom of
the frame, and thats it.

A single MCU can act as the keyboard-encoder, and you can actually
translate the "buttons" on the side to scrolling through list entries,
buttons on the lower bottom of the frame to select options or some such.

It's a cheap technique material wise, and when you have a PCB which is
simply laid into the inside of the wooden frame and then have metal wire
run through drilled holes in the wood and solderd to the PCB, it's a
low-manual-labor issue.

Post by Luke Kenneth Casson Leighton

Post by Dominique Dumont
Hello Luke

about the digital photo frame idea: that's an even simpler version of
a tablet, effectively. are you looking to build it yourself? if so,
below

No really build it myself.

great!
batteries / power. ah. when you say "digital photo frame", do you
mean "portable device" or do you mean "put it on the wall or
mantlepiece"?

Post by Dominique Dumont
Then, we'd need to field test our device. This could be done with netbooks or
small tablets, but not with a bunch of boards and wires ;-) . We'd need ~ 10
of them to lend to family and neighbours and get feedbacks.

ok, well we have about 17 committed so far. the hardware development
costs (PCB, tooling) are about $USD 2,000. with 10 more, that would
be about $75 each for one (fully-populated, fully-functional but with
no external case) CPU card.

Post by Dominique Dumont
Once this is done, we'll need a small serie (~ 100 depending on price) to
really get serious.

ok, once the first $2000 of NREs is done, that's it: it doesn't have
to be paid again (NREs: non-recurring expenses. go figure).
subsequent PCBs therefore can be much reduced pricing.

Post by Dominique Dumont
Regarding the device, we'd need something quite standard. The screen should be
at least 800x600. Bigger would be better, but the device price for the small
serie will be the main limiting factor.
Regarding user input, I'm wondering if a touchscreen would not be better than
buttons behind the frame. Navigating in a menus with these rear buttons is not
user friendly. But buttons are needed in case customer loose the remote
control. May be a touchscreen as a fallback is better for users. Still
scratching head there.

an overlay resistive touchscreen is about $5, typically. they do
however require calibration: they're also sensitive to temperature,
humidity etc.
capacitive ones made of glass are *MORE* expensive than the
same-sized LCD! absolutely ridiculous, but that's the way it goes.
they're also heavier (because they have to be structurally strong).
there do exist plastic capacitive touchpanels, they're still in
development, the price _will_ come down... at some point.
bottom line is: touchscreens sound great until you look into the cost
and useability.
there are some options however for very coarse-grained touchscreens.
if you can get away with "button-sized" fixed areas (such as used on
microwave ovens) then costs are dramatically reduced.

Post by Dominique Dumont

the right kind of low-speed, low-resolution LCD panel (800x600,
800x480, 480x320 etc.) you could almost connect it up directly to the
EOMA-PCMCIA connector. if you didn't need a brightness control, that
would bring the cost down as well. the IR receiver you can get as a
USB device. if you kept it _that_ simple you'd hardly need a PCB at
all!

That would be better: I've not done HW design for a while (~ 20 years).

the speed of light is still the same constant, as best we are aware,
and V still equals I times R :)
l.
p.s. i've successfully compiled the allwinner a10 reference source
code, using the emdebian cross-compile toolchain.

--
Rüdiger Leibrandt

Technologie-Zentrum Informatik
Intelligente Systeme
Universität Bremen
Postfach 330440
D-28334 Bremen

Cel: +49 - 170 - 688 - 5664
Tel: +49 - 421 - 218 - 64012 (Sekretariat: 64000)
Fax: +49 - 421 - 218 - 64095


Email : ***@tzi.de
Website : http://ag-ki.tzi.de/~quitex

[Phil Endecott]

I use a Gyration gyroscopic mouse.

(These devices are also great if you need to give a software demo or similar on
a projector.)


Phil.

[Ruediger Leibrandt]

Gyro-Mouse?

That'd be a seperate I/O device. In case of requiring or wanting such, one
could always use a standard bluetooth mouse & keyboard, which's use is more
common than those "free-flight-mouses".

Integrating such a gyromouse inside the photoframe-tablet seems of little use,
either, as the normal accelerometers inside a tablet do somewhat the same as
the mouses gyro's do.
As the photoframe itselfs remains stationary on the wall, however, the only
use of such sensors is to detect the orientation of the frame or sense whole-
environment-effects, e.g. earthquakes.
Something I don't wish ever to be told about by my tablet...

The Gyromouses are for presentations, or as an intermediary between more
sophisticated gesture recognition or spatial movement detecting devices and
classical mouses and tablets. That's at least what I think they are for, as ,
so far, I never was able to check out one such device for myself and see which
applications make proper use of the additional axes of movement.

Post by Phil Endecott
I use a Gyration gyroscopic mouse.
(These devices are also great if you need to give a software demo or
similar on a projector.)
Phil.

[Dominique Dumont]

Post by Ruediger Leibrandt
Integrating such a gyromouse inside the photoframe-tablet seems of little
use, either, as the normal accelerometers inside a tablet do somewhat the
same as the mouses gyro's do.

Well, actually, I just need to know if the frame is in landscape mode or
portrait mode. Usually, it's a small component mounted at 45° on a PCB which
does a faint "tick" sound when moved (probably a small metallic ball in
there). That's low tech, but good enough for my purpose ;-)

All the best

Dominique
--
http://config-model.wiki.sourceforge.net/ -o- http://search.cpan.org/~ddumont/
http://www.ohloh.net/accounts/ddumont -o- http://ddumont.wordpress.com/

[Dominique Dumont]

Post by Rüdiger Leibrandt
It's a cheap technique material wise, and when you have a PCB which is
simply laid into the inside of the wooden frame and then have metal wire
run through drilled holes in the wood and solderd to the PCB, it's a
low-manual-labor issue.

This is a nice idea for a DIY frame. On the other hand, I'm a bit worried that
static electricity could fry the circuit. And, I don't know if this solution
could be applied on an industrial frame.

All the best

Dominique
--
http://config-model.wiki.sourceforge.net/ -o- http://search.cpan.org/~ddumont/
http://www.ohloh.net/accounts/ddumont -o- http://ddumont.wordpress.com/

[Dominique Dumont]

Post by Luke Kenneth Casson Leighton
batteries / power. ah. when you say "digital photo frame", do you
mean "portable device" or do you mean "put it on the wall or
mantlepiece"?

I mean always on power socket. No batteries

Post by Luke Kenneth Casson Leighton

Post by Dominique Dumont
Then, we'd need to field test our device. This could be done with
netbooks or small tablets, but not with a bunch of boards and wires ;-)
. We'd need ~ 10 of them to lend to family and neighbours and get
feedbacks.

ok, well we have about 17 committed so far. the hardware development
costs (PCB, tooling) are about $USD 2,000. with 10 more, that would
be about $75 each for one (fully-populated, fully-functional but with
no external case) CPU card.

This thread started on tablet hardware. Do you have a price range for such a
device ? (i.e. with screen and case)

Post by Luke Kenneth Casson Leighton
an overlay resistive touchscreen is about $5, typically. they do
however require calibration: they're also sensitive to temperature,
humidity etc.

ok I've seen this kind of device while repairing my kid's gameboy.

Post by Luke Kenneth Casson Leighton
capacitive ones made of glass are *MORE* expensive than the
same-sized LCD! absolutely ridiculous, but that's the way it goes.
they're also heavier (because they have to be structurally strong).
there do exist plastic capacitive touchpanels, they're still in
development, the price _will_ come down... at some point.

ok let's forget these.

Post by Luke Kenneth Casson Leighton
bottom line is: touchscreens sound great until you look into the cost
and useability.

The alternative is a remote control. Generic ones are used to navigate in
menus. But this is not user friendly for older people. I'm thinking about
having a more specific remote control, but I'm worried about the development
cost (PCB, plastic molds...)

Post by Luke Kenneth Casson Leighton
there are some options however for very coarse-grained touchscreens.
if you can get away with "button-sized" fixed areas (such as used on
microwave ovens) then costs are dramatically reduced.

I guess that such buttons could be drawn on the screen when required. One
touch anywhere will trigger drawing buttons on screen.

Post by Luke Kenneth Casson Leighton
the speed of light is still the same constant, as best we are aware,
and V still equals I times R :)

;-)

All the best

Dominique
--
http://config-model.wiki.sourceforge.net/ -o- http://search.cpan.org/~ddumont/
http://www.ohloh.net/accounts/ddumont -o- http://ddumont.wordpress.com/

[Luke Kenneth Casson Leighton]

Post by Dominique Dumont

 batteries / power.  ah.  when you say "digital photo frame", do you
mean "portable device" or do you mean "put it on the wall or
mantlepiece"?

I mean always on power socket. No batteries

whewww :)

Post by Dominique Dumont
This thread started on tablet hardware. Do you have a price range for such a
device ? (i.e. with screen and case)

achh... mass-volume pricing? 7in 800x480 LCDs are around $12, a case
would be around... $2 if that. motherboard (in this instance) 2-layer
single-sided including components: probably about... $2.50. bits from
the EOMA-PCMCIA card: about $15 - total comes to about $30 to $35.

it's the lack of a battery (in this instance) that does it (normally
an extra $6 to $10 or so). non-mass-volume pricing would
approximately double things.

profit, tax, shipping etc. usually adds at least 100% on top.

Post by Dominique Dumont

 an overlay resistive touchscreen is about $5, typically.  they do
however require calibration: they're also sensitive to temperature,
humidity etc.

ok I've seen this kind of device while repairing my kid's gameboy.

achh, pffh, if you've handled those, you're onto a winner - you know
the score. btw you can actually get 7in LCDs which have resistive
touchpanels already attached.

Post by Dominique Dumont

 bottom line is: touchscreens sound great until you look into the cost
and useability.

The alternative is a remote control. Generic ones are used to navigate in
menus. But this is not user friendly for older people. I'm thinking about
having a more specific remote control, but I'm worried about the development
cost (PCB, plastic molds...)

the PCB is peanuts. the circuit's simple enough to use a breadboard.
i think... you _may_ get away with an LCD Driver IC that has legs
(DIP) - can't remember the name of the one that's so low-cost that
everyone keeps using it, even though it's not SMT, it does
voltage-multiplying up to 28v, for arrays of LED backlights.... achh,
apologies.


you do _not_ want to go down the route of having your own case
manufactured. find a box that is "close enough", or spend vast
amounts of time finding something off-the-shelf. despite spending
vast amounts of time, it is guaranteed to be less money than paying to
have a case developed.

if you do not believe this, please review the openpandora blogs.

another option for low-volume is acrylic (which is clear), then paint
it. specialcomp got a case designed that way for the beagleboard.

Post by Dominique Dumont

there are some options however for very coarse-grained touchscreens.
if you can get away with "button-sized" fixed areas (such as used on
microwave ovens) then costs are dramatically reduced.

I guess that such buttons could be drawn on the screen when required. One
touch anywhere will trigger drawing buttons on screen.

yes - and if you keep it coarse-grained enough, you shouldn't need to
do too much in the way of calibration of the resistive touchpanel.
calibration is only kinda needed if you want millimetre / centimetre
accuracy. if you only care about "the top quadrant" you're laughing.

l.

[Jeremiah Foster]

Post by Phil Endecott

Post by Rob van der Hoeven
Hardware that is not mass produced has some other issues,
namely availability and vendor lock-in

You think that your tablet is going to have better availability and less lock-in
than a board from Freescale or TI? That seems unlikely to me. Look at the
BeagleBoard; it would be hard to find any smartphone or tablet device that has
been available for as long as that has.

Absolutely. And in the near term future the Beagleboards main CPU (an OMAP 3) is going to continue to be used commercially in Nokia's phones likely ensuring that the OMAP 3 CPU is supported for a number of years. A beagleboard is a good investment with a healthy software and hardware ecosystem. And they're cheap ~$125 US.

Post by Phil Endecott

Post by Rob van der Hoeven
I think it must be possible to buy an android motherboard for just a
fraction of the price that i paid for my tablet.

Why do you think that? I have personally never seen an "Android motherboard"
offered for sale at all, let alone for a low price.

Android's Linux kernels are supported (maintained?) by Linaro. Anything that runs Android can run GNU/Linux. Android is pretty much just Java stuff on the Linux kernel, so I'm not sure what a "Android motherboard" would be.

Regards,

Jeremiah

[Jeremiah Foster]

Wysłane z iPhone'a

^^ heh

Post by Jeremiah Foster
Android's Linux kernels are supported (maintained?) by Linaro.

With my Linaro hat on I must object. Depending on what you meant the statement above is either highly inaccurate or simply untrue.

Hence the question mark. :)

Android kernel situation is complicated and varies per board/SoC. What Linaro does is try to upstream and unify the kernel for Linaro member companies SoCs.

What does that mean in practice?

This is far from finished and uniform. The "BSP" kernel that hardware vendors provide is not supported by Linaro and in fact often contains code that cannot go upstream.

What does it use this proprietary code for? To know the APIs or to get other hardware interface info? Isn't that a little risky? Won't proprietary, and potentially patented IP leak into the Linaro work? (Not that I believe in IP.)

Linaro has several trees, including a grand unification tree that tries to support all the member companies chips in one tree (and one binary, thanks to device trees) but this effort is years away (my personal estimate, I don't speak for the organization). In addition we have several trees for normal/androidized kernel for each board. In the latest 2011.10 release hardware was not supported in 100% on any board that I'm aware of.
Having said that the term "supported" seems inappropriate to me. We do work on those boards though.

How would you define it?

Post by Jeremiah Foster
Anything that runs Android can run GNU/Linux.

This is a gross oversimplification IMHO. You usually get androidized BSP kernel from a few months/years ago with binary parts that have no corresponding source code. Good luck booting vanilla kernel there.

But it appears to me that all the official boards that are targets for Linaro can run a vanilla kernel, is that not the case? If not, what BSP stuff are you referring to - graphics acceleration?

[Zygmunt Krynicki]

Post by Jeremiah Foster

Post by Jeremiah Foster
Android's Linux kernels are supported (maintained?) by Linaro.

With my Linaro hat on I must object. Depending on what you meant the statement above is either highly inaccurate or simply untrue.

Hence the question mark. :)

I think what I originally meant is that we don't focus solely on Android.

"Improved" is the word I would use, that implies neither support nor
maintenance as we are between the vendors (that are also part of Linaro)
and upstream kernel community (that we are a part of) and we have no
control of either side and their actions. As to what we do check our FAQ
(http://www.linaro.org/faqs/) and read on.

Post by Jeremiah Foster

Android kernel situation is complicated and varies per board/SoC. What Linaro does is try to upstream and unify the kernel for Linaro member companies SoCs.

What does that mean in practice?

Disclaimer: I'm not a kernel developer. I have experience in the
non-Intel part of the world but I'm not the sort of person with
up-to-date hands-on experience. For those folks please look at traffic
in linaro-***@lists.linaro.org and at our patchwork instance at
patches.linaro.org. You can see what kind of patches we push upstream
and if they have landed yet.

As for your question, read on.

A lot of ARM devices have a BSP kernel (android or not) that is prepared
by some 3rd party (sometimes also by the vendor themselves if the device
is a clone of the reference design) and that kernel is generally not
pushed upstream.

This affects, by far, the vast majority of devices out there (I'd say
that nothing gets pushed by those companies simply because their work
mode does not require such a step - we are working on educating them in
the benefits of working both towards products and common code base).

The ARM tree in the upstream kernel is, again, by far, the largest of
all the other architectures. If I remember correctly it is in fact
larger than *all* the other trees combined. The reasons for this are
complicated but can be generally simplified to code duplication between
the different devices and greater diversity in the actual hardware as
compared to other platforms.

To get ARM Linux healthy we need to reduce that clutter and make sure
that support for the latest and greatest hardware is upstream quickly
and the code is being actively maintained by everyone.

Post by Jeremiah Foster

This is far from finished and uniform. The "BSP" kernel that hardware vendors provide is not supported by Linaro and in fact often contains code that cannot go upstream.

What does it use this proprietary code for? To know the APIs or to get other hardware interface info? Isn't that a little risky? Won't proprietary, and potentially patented IP leak into the Linaro work? (Not that I believe in IP.)

The term proprietary is a bit misleading, the code IS released as GPL.
It is simply there to support some parts (often userspace or "firmware")
that is not open sourced and cannot be for all practical considerations.

As for the patches in general, there are different reasons why they are
not suitable for being proposed and included upstream:

1) Shabby code, against old trees, copy-pasted from another similar
device, maintenance hell. This is, by my unqualified judgment, the vast
majority of the problem.

2) Code that has no good place in the kernel just yet because the kernel
interfaces are insufficient for the extremely complicated world of ARM
SoCs. Off the top of my, unqualified, head: power management, memory
management, everything related to graphics and probably many more. Here
the reason for not being upstream is that there is no consensus on how
to do something (how to support a certain class of SoC components) that
could be applied to many vendors and non-arm world as well. Here the
people that write the BSP cannot solve the problem and just implement
their own solution to meet the deadline. Such code is often very good
but there are many similar solutions that are quite nontrivial to merge
into one sensible codebase. One such example is memory management where
we have no less than 3 or 4 competing interfaces from various companies
and there is a working group inside Linaro and the greater Linux
community that tries to solve this problem.

3) Bits that enable proprietary userspace drivers. The reasons are
obvious. This could be related to lots of different things, not only
graphics as people often think. IP and software edge (optimizations that
make otherwise identical hardware perform better than competition) is
probably a big motivation here. The IP protection is not only used as in
"don't steal our stuff" but rather "hey, with this being binary it is
harder to prove that we violate a specific patent you have". In
retrospective this is a thing those companies obviously need. Just look
at how many Android handset vendors pay to, for example, Microsoft, for
patents that allegedly apply there. The world of graphics is riddled
with patents and I'm sure that a big money-laden hardware vendor is a
good target for whoever owns the patents.

Post by Jeremiah Foster

Linaro has several trees, including a grand unification tree that tries to support all the member companies chips in one tree (and one binary, thanks to device trees) but this effort is years away (my personal estimate, I don't speak for the organization). In addition we have several trees for normal/androidized kernel for each board. In the latest 2011.10 release hardware was not supported in 100% on any board that I'm aware of.
Having said that the term "supported" seems inappropriate to me. We do work on those boards though.

How would you define it?

By "work" I meant "we are *working* on making the kernel and userspace
on those boards better in each release". Better is shared amongst:

1) More patches landed upstream, thus less delta.
2) Less duplication within the kernel (better code), more device tree
usage, closer to having one kernel binary that supports several
different boards.
3) Better power management, stability, performance, more features, bells
and whistles.
4) Less delta from the android variant to the normal variant. More
discussion and more consensus on how to join the two worlds.

And let's not forget, my own personal favorite, more validation. The
code is tested both manually and automatically and the scope, coverage
and quality is pushed forward each time.

Post by Jeremiah Foster

Post by Jeremiah Foster
Anything that runs Android can run GNU/Linux.

This is a gross oversimplification IMHO. You usually get androidized BSP kernel from a few months/years ago with binary parts that have no corresponding source code. Good luck booting vanilla kernel there.

But it appears to me that all the official boards that are targets for Linaro can run a vanilla kernel, is that not the case? If not, what BSP stuff are you referring to - graphics acceleration?

No I don't think that is the case. A quick glance at
http://git.linaro.org/gitweb will show you how many trees we have.
Except for the explicit upstream trees that Nico maintains none of the
important changes in other trees are upstream today (at least not yet).

Even the Linaro various kernels (which are _not_ the BSP kernels) fail
to work sensibly on all of the boards today. Next-gen boards are usually
the ones with weakest support (although that is rapidly changing, thanks
to what we are doing). Often most primitive board features work (like,
it kind of boots with a specific boot loader and the CPU runs) but
anything you definitely want those boards to do: power management,
stability, sound, graphics, 3D & multimedia, wifi/bluetooth, FM radio,
GPS(?), DSP suppoet, ARMv7 optimizations, is simply not there.

Best regards
ZK

PS: I think that you should ask those questions on @linaro-dev. I could
be talking nonsense here and the people that really know simply did not
see this message. Therefore I'm cross-posting to linaro-dev.

[Luke Kenneth Casson Leighton]

On Fri, Oct 28, 2011 at 5:59 PM, Zygmunt Krynicki

Disclaimer: I'm not a kernel developer. I have experience in the non-Intel
part of the world but I'm not the sort of person with up-to-date hands-on
experience. For those folks please look at traffic in
patches.linaro.org. You can see what kind of patches we push upstream and if
they have landed yet.

ok - this is potentially misleading, zygmunt, at best irrelevant.
ok, shall i point out a correction, and you, or someone else from
linaro, can, if it is incorrect, provide the required corrections,
yes?

by mentioning the above patch queue (on the debian-arm list), then
*despite* previously mentioning that linaro is "between the vendors
and the kernel developers", there is a risk that people could not
connect the dots between this and the previous paragraph (seen that
happen so many times it's unreal) and thus it would appear - to them -
that linaro is *still* seen to quotes be an authority unquotes
regarding patches [for specific hardware]. the correction - if any is
needed - is that linaro is NOT an "authority" of ANY KIND regarding
"definitive patchsets" or in fact an authority of any kind PERIOD.

linaro is, in fact an "accelerator", helping SoC vendors to push
"lowest common denominator" code into the linux kernel for the
convenience of *MULTIPLE* hardware and software developers using that
PARTICULAR SoC, but linaro are not, repeat NOT "direct" suppliers of
linux kernel source code for a specific device.

perhaps if any code for a particular device _is_ pushed upstream by
linaro, it is almost "by accident", by nature of it, for example,
being a particular example "BSP" or convenient "Reference Platform".

putting it into context: linaro is paid by _SoC Vendors_. linaro is
*not* paid by individual Hardware Factories (afaik) to do
hardware-specific, device-specific linux kernel development.

thus in that context, whilst it is nice that linaro is doing upstream
patches, and it's nice that you mentioned it, it is in the context of
this discussion, "off-topic".

would that be a correct assessment?

i apologise a) if it is not - please do correct things b) for feeling
obligated to point out that linaro's patches and patch development is
"off-topic".

l.

[Zygmunt Krynicki]

Post by Luke Kenneth Casson Leighton
On Fri, Oct 28, 2011 at 5:59 PM, Zygmunt Krynicki

Disclaimer: I'm not a kernel developer. I have experience in the non-Intel
part of the world but I'm not the sort of person with up-to-date hands-on
experience. For those folks please look at traffic in
patches.linaro.org. You can see what kind of patches we push upstream and if
they have landed yet.

ok - this is potentially misleading, zygmunt, at best irrelevant.
ok, shall i point out a correction, and you, or someone else from
linaro, can, if it is incorrect, provide the required corrections,
yes?

I was just trying to state my opinion that Linaro "supporting android
kernels" is kind of misleading in its own way. While I'm not a kernel
developer I'm deep in Linaro and I try to stay informed about what is
going on. I would love to see responses from other Linaro engineers as
that would correct my personal opinion if I got something wrong.

Note: if the discussion below is off-topic to the original thread please
feel free to ignore me.

Post by Luke Kenneth Casson Leighton
by mentioning the above patch queue (on the debian-arm list), then
*despite* previously mentioning that linaro is "between the vendors
and the kernel developers", there is a risk that people could not
connect the dots between this and the previous paragraph (seen that
happen so many times it's unreal) and thus it would appear - to them -
that linaro is *still* seen to quotes be an authority unquotes
regarding patches [for specific hardware]. the correction - if any is
needed - is that linaro is NOT an "authority" of ANY KIND regarding
"definitive patchsets" or in fact an authority of any kind PERIOD.

Linaro is more than a random collection of kernel hackers. The key
advantage "we" have is that those hackers come from hardware vendors
directly (and here I mean the people that actually design the SoC and
then produce chips you can buy). Linaro is in a special position as it
can attempt cross-vendor solutions on a scale nobody has attempted
before. That is much stronger than any other group that I know of. We
can, in a way, discuss and implement de-facto standards that the main
ARM vendors will use in BSPs for subsequent products.

Post by Luke Kenneth Casson Leighton
linaro is, in fact an "accelerator", helping SoC vendors to push
"lowest common denominator" code into the linux kernel for the

Yes, Linaro is an accelerator. I would not, however, limit us to the
lowest common denominator. We are working on solving the big problems
that ARM kernel faces. That affects and involves everyone but is not
equivalent to lowest common denominator in my opinion.

Post by Luke Kenneth Casson Leighton
convenience of *MULTIPLE* hardware and software developers using that
PARTICULAR SoC, but linaro are not, repeat NOT "direct" suppliers of
linux kernel source code for a specific device.

No, that is not our goal (to be the supplier of kernel trees for random
hardware). What you may find interesting however is that some hardware
device manufactures (not SoCs vendors) want to use or sometimes even
_require_ Linaro-based trees for their devices.

I think that is a fantastic validation of our efforts. Working with
Linaro trees is faster and thus cheaper for those vendors.

Post by Luke Kenneth Casson Leighton
perhaps if any code for a particular device _is_ pushed upstream by
linaro, it is almost "by accident", by nature of it, for example,
being a particular example "BSP" or convenient "Reference Platform".

Not by accident and not for a particular device. Our intention is to
support the _next_ set of SoCs in the upstream kernel. When a company
goes out to search for devices to build their next product those SoC
will be already enabled and not only in a random BSP package but
straight in the upstream vanilla kernel.

Post by Luke Kenneth Casson Leighton
putting it into context: linaro is paid by _SoC Vendors_. linaro is
*not* paid by individual Hardware Factories (afaik) to do
hardware-specific, device-specific linux kernel development.

Some of the device vendors are coming closer to Linaro. I can cite one
example from memory, Genesi, the manufacturer of several interesting
freescale based products has joined Linaro to improve cooperation
between their engineers working on next Genesi products and Linaro's
common vision on how things should work.

Post by Luke Kenneth Casson Leighton
thus in that context, whilst it is nice that linaro is doing upstream
patches, and it's nice that you mentioned it, it is in the context of
this discussion, "off-topic".

As stated at the beginning of this message I was attempting to correct
an inaccurate, in my opinion, assessment of what Linaro is doing.

In a finishing note I'd like to state that the devices that we work on
daily (and improve GNU/Linux experience on) are being used to form
products we'll see on the market.

Initially "development boards" were a queer concept for the SoC vendors
but I think that is rapidly changing and more vendors start to see the
effect making those devices available on the market has on their
ecosystem. Devices are turned into prototypes and products, more
smaller-scale companies can participate, more code is written in the
community to support those boards. The ecosystem grows faster.

I can point to a simple recent example: the ST-E snowball board is
actually a full-featured product platform that can be taken from a
"prototype" to "product" without complex hardware redesign. This means
that one wishing to obtain a supply of hardware for a free ARM PC like
product will find that there are many virtually identical (and well
supported by free software) devices on the market. I think that is a
good thing.

Best regards
Zygmunt

[Christian Robottom Reis]

Post by Zygmunt Krynicki
I was just trying to state my opinion that Linaro "supporting
android kernels" is kind of misleading in its own way. While I'm not
a kernel developer I'm deep in Linaro and I try to stay informed
about what is going on. I would love to see responses from other
Linaro engineers as that would correct my personal opinion if I got
something wrong.

If the context in this message hadn't been butchered (and the message
itself wasn't the size of the Britannica) I would risk a more complete
answer; however, generally Linaro a) does intend its kernel releases and
distribution images to be runnable and useful to developers using our
members' development boards and b) we do significant work testing,
fixing and maintaining said kernel releases and distribution images.

If that wasn't the question, can somebody restate it clearly?
--
Christian Robottom Reis, Engineering VP
Brazil (GMT-3) | [+55] 16 9112 6430 | [+1] 612 216 4935
Linaro.org: Open Source Software for ARM SoCs

[David Given]

On 28/10/11 17:59, Zygmunt Krynicki wrote:
[...]

Post by Zygmunt Krynicki
Here the
people that write the BSP cannot solve the problem and just implement
their own solution to meet the deadline.

Just to expand on this: right now the BSPs are mostly written by
hardware manufacturers, and hardware manufacturers have a radically
different mindset to software manufacturers. Where we're mostly
interested in provided generic solutions that can be applied to a wide
range of devices, they care only about specific deliverables, and once
the deadline has been met they no longer care about it --- because they
have a new deliverable deadline to meet.

As an example: I recently worked with a god-awful television
motherboard. (It actually ran MIPS, but that's irrelevant to this.) For
the audio system, instead of writing a /dev/dsp driver, they instead
wrote a library which used IPC to talk to a userspace daemon which
mmapped /dev/kmem and poked the DSP registers directly. Why? Because the
only software that was *ever* going to need to play audio was their own,
and this approach was more flexible, simplified project dependencies,
and above all saved valuable development time.

[Zygmunt Bazyli Krynicki]

Wysłane z iPhone'a

Post by Jeremiah Foster

Post by Phil Endecott

Post by Rob van der Hoeven
Hardware that is not mass produced has some other issues,
namely availability and vendor lock-in

You think that your tablet is going to have better availability and less lock-in
than a board from Freescale or TI? That seems unlikely to me. Look at the
BeagleBoard; it would be hard to find any smartphone or tablet device that has
been available for as long as that has.

Absolutely. And in the near term future the Beagleboards main CPU (an OMAP 3) is going to continue to be used commercially in Nokia's phones likely ensuring that the OMAP 3 CPU is supported for a number of years. A beagleboard is a good investment with a healthy software and hardware ecosystem. And they're cheap ~$125 US.

Post by Phil Endecott

Post by Rob van der Hoeven
I think it must be possible to buy an android motherboard for just a
fraction of the price that i paid for my tablet.

Why do you think that? I have personally never seen an "Android motherboard"
offered for sale at all, let alone for a low price.

Android's Linux kernels are supported (maintained?) by Linaro.

With my Linaro hat on I must object. Depending on what you meant the statement above is either highly inaccurate or simply untrue.

Android kernel situation is complicated and varies per board/SoC. What Linaro does is try to upstream and unify the kernel for Linaro member companies SoCs. This is far from finished and uniform. The "BSP" kernel that hardware vendors provide is not supported by Linaro and in fact often contains code that cannot go upstream. Linaro has several trees, including a grand unification tree that tries to support all the member companies chips in one tree (and one binary, thanks to device trees) but this effort is years away (my personal estimate, I don't speak for the organization). In addition we have several trees for normal/androidized kernel for each board. In the latest 2011.10 release hardware was not supported in 100% on any board that I'm aware of.

Having said that the term "supported" seems inappropriate to me. We do work on those boards though.

Post by Jeremiah Foster
Anything that runs Android can run GNU/Linux.

This is a gross oversimplification IMHO. You usually get androidized BSP kernel from a few months/years ago with binary parts that have no corresponding source code. Good luck booting vanilla kernel there.

Best regards
Zygmunt Krynicki

[Rob van der Hoeven]

Post by Phil Endecott

Post by Rob van der Hoeven

Post by Arnaud Patard (Rtp)

Post by Rob van der Hoeven
I think the hardware of this tablet can also be used as a server or
desktop computer. The tablet is mass produced and very cheap (i got mine
for 149 euro).

For that price, to make a server, I would rather buy a loco board or any
other development board

These boards are not mass produced which makes them relatively
expensive.

The i.MX LOCO board, the OMAP panda board, and some of the others cost about the
same as your tablet.

Panda board has very nice specs.

Post by Phil Endecott

Post by Rob van der Hoeven
Hardware that is not mass produced has some other issues,
namely availability and vendor lock-in

You think that your tablet is going to have better availability and less lock-in
than a board from Freescale or TI? That seems unlikely to me. Look at the
BeagleBoard; it would be hard to find any smartphone or tablet device that has
been available for as long as that has.

The FreedomBox project is looking for very cheap hardware. This hardware
exists today, but it is used for running Android. It would be very nice
if we could liberate this hardware and use it for our own computing.

Beagle board and Panda board are very nice but i don't think they will
become cheap enough for the FreedomBox (one monopolistic manufacturer,
low volume - "only" 8900 Panda boards sold)

If the FreedomBox would use a popular SoC then the manufacturer of the
motherboard seems less important to me. All the major functionalities
for which drivers are needed are on one chip. We could simply switch to
an other board with the same SoC and still run our software (maybe with
some minor adjustments, please correct me if i am wrong...)

Post by Phil Endecott

Post by Rob van der Hoeven
I think it must be possible to buy an android motherboard for just a
fraction of the price that i paid for my tablet.

Why do you think that? I have personally never seen an "Android motherboard"
offered for sale at all, let alone for a low price.

You find out the manufacturer of the motherboard inside a tablet. Then
you contact this manufacturer and say: Hi, i know you are making 10000
motherboards for Yarvik, if i were to order 1000 of these boards what
would the price be? I think the manufacturer will be happy with an extra
order. Mass produced boards are well tested (the manufacturer simply
can't afford mass problems) and cheap.

Post by Phil Endecott

Post by Rob van der Hoeven
Why is relying on
hardware with a SoC such a bad idea? If the SoC is popular it will not
go out of production for a long time.

No, that's not how it works. Both popular and unpopular chips are replaced on a
schedule that's determined by advances in manufacturing technology. This also
applies to the consumer products that are made from them: even if a device is
popular, it will soon be replaced with something that is faster and cheaper.

Not quite true i think, especially for SoC. My FreedomBox has an Marvell
6281 (Kirkwood) SoC inside. This chip has been around for a long time.
You are right for non-SoC boards, they can more easily change for
example the graphics chip and spoil our fun.

Regards,
Rob van der Hoeven.

http://freedomboxblog.nl

[Luke Kenneth Casson Leighton]

On Fri, Oct 28, 2011 at 6:50 PM, Rob van der Hoeven

Post by Rob van der Hoeven
The FreedomBox project is looking for very cheap hardware. This hardware
exists today, but it is used for running Android. It would be very nice
if we could liberate this hardware and use it for our own computing.

rob - i don't know if you're aware but the conversations with these
factories typically go like this, usually with at least a 48-hour
delay due to timezone differences.

so the alternative is this: is $15 (in mass-volume) for a device with
a 1ghz Cortex A8 considered sufficiently cheap?

if "yes" then i am willing to help carry out the negotiations - for
zero monetary up-front cost (because i can get a royalty on hardware
products sold outside of the People's Republic of China) - in order to
_get_ the full GPL Source Code for the device.

for zero up-front cost i will get you what you seek. full GPL
compliant source code. 1ghz Cortex A8 device which will be around the
$15 mark, FOB, in large volumes.

if enough people say "yes, sure, if you manage to pull that off, i'll
place an order for X units at non-mass-volume pricing of $NN", where
NN is likely to be... guessing here (i'll get the factory to confirm
it)... $60 for the first 100 units, then i can move things forward.

i apologise for the apparent complexity but the alternative is "cash
up-front". this is at least a way that i can negotiate a "no money
down" deal. not least because the hard part is not the hardware, it's
the software [which i can handle if there's nobody else volunteering].

anyway, if you're feeling particularly masochistic, feel free to read
on further, to gain some insight into how discussions with factories
in china actually end up going, and why. please feel free to skip it.

@begin illustration of insane discussions with factory X


[day 1] Q: we would like to buy a sample of your product
[day 3] A: thank you! here is our bank account details.
[day 10] Q: thank you! we have received the sample by DHL. where is
the GPL source code?
[day 12] A: we do not know what you are talking about, sorry
[day 14] Q: linux kernel. u-boot. tool-chain. you know. you
developed the software, yes?
[day 16] A: sorry, we really do not know. let us get back to you.

at this point, they go off and contact their ODM, who were the *real*
people who developed the software, but they don't tell you or even
admit that. when they do "answer", it is actually an answer from the
ODM (!) but relayed via the factory to make them "look competent".
the delays are them talking behind the scenes to the ODM...

[day 25] Q: sorry we did not receive a response.
[day 27] A: apologies, yes, we cannot give you the GPL source code.
[day 29] Q: huh? i do not understand. here is the GPL license. read it!
[day 31] A: sorry, we do not know. let us get back to you.

further discussions occur whilst they talk to the ODM. the factory
and the ODM agree a "strategy" to basically tell you to f*** off,
without actually saying "f*** off". we're now a full month since the
first contact, and the end is *nowhere near* in sight.

[day 32] Q: sorry we did not receive a response.
[day 34] A: apologies, yes, we can only give you the source code if
you order 10,000 units
[day 36] Q: wtf?? did you actually _read_ the GPL??
[day 38] A: sorry, it is not our problem
[day 40] Q: yes it is - would you like me to alert gpl-violations.org?
[day 42] A: apologies - let us get back to you

now, depending on whether you choose the strategy of being carrot or
stick, you can sometimes threaten them to the point where they will
give you the name of the ODM. sometimes this works, sometimes it
doesn't.

[day 50] Q: sorry, we did not receive a response
[day 52] A: apologies, no we cannot comply with your request. please
place an order.
[day 54] Q: are you mad?? we're not going to place an order without
the source code!!!
[day 56] A: sorry but ...

and this is where things begin to degenerate. what the situation is,
we are guessing, is that the factory hasn't actually received the
source code... at all! and, worse, they negotiated a deal where the
GPL source code was in fact ***SOLD*** to them, on a royalty basis.

thus, the f**** little w****** at the f***** ODM ***** software
company won't _give_ the factory the GPL source code until the factory
has paid them their royalties!

and _that_ means that the factory has to sell at least 10,000 units in
order to get enough profit in order to pay the ODM in order to get the
f*****g source code which they should have f*****g well been given in
the first place.

the only factory who actually sent me the GPL source code got into
deep shit with the ODM _and_ the SoC company for quotes violating the
NDA quotes which is really funny because both the ODM and the SoC
company actually lost the right to distribute that very same source
code _to_ the factory because of the conflicts with the GPL!

hilarious in a sort of "only a geek who knows the GPL" kind of way...

anyway, enough of that. thank god.

@end illustration of insane discussions with factories

ok, so you skipped it, or can still read whilst manic-laughing
hysterically. this is why i have been working extremely hard to
negotiate with factories to provide *them* in advance with the
required linux software.

as an example: can you see why it is so utterly utterly critical, to
e.g. the success of the FreedomBox Foundation's goals, that the tables
are turned, and we (Free Software Community plural we) supply *the
factories* with the GPL-compliant Software, rather than letting a
GPL-violating third party somewhere in China do it?

for these chinese "software developer" companies, i've mentioned it
before but it's worth reiterating: they are actually in a very
unstable situation, because the moment their employees finish on one
project, they quit and go and get a job elsewhere for 3x the salary!

the actual number of software developers in china is too low to go
round all the ODMs that want to employ them. so, once a product goes
end-of-life, you're REALLY in the shit because the developer didn't
take backups, or nobody knows where the backups are, and their only
answer is "well, tough - go buy the next product, that old one is
rubbish now anyway, here's our list of fees and royalties".

it's an insane situation - i'm offering Free Software Developers a
"way out" of this insanity.

l.

[Michelle Konzack]

Comments on the "Adventures of Luke", ;-)

Hello Rob,

I am currenly in contact with 17 Enterpises in Asia which are
manufacturing TablePC, PanelPC and Mainboards and I have the same
experience as Luke. The fool each other.

Regarding the GPL-Violations, I had to contact the the "Chinese Chamber
of Industry" in Germany which took over the GPL-Violation. So, I have
created some trouble for SOME chinese (not very big) manufacturers.


I am Electronic Engineer and Programmer and try to build two TabletPC
(8" and 12" with capacitive Touchscreen) and a PanelPC (19" AUO G-Series
with resitive Touchscreen) using ARM1176J-ZFS, Cortex A8/A9.

Why?

I need OutDoor TabletPC and LowPower Industrial PanelPC.

Since April I have solved the Problem with the Marvel Discovery MV78200
(1200 MHz) and after I have burned 4 Mainboards (each 2500 Euro materiel
and soldering) I hade to send the for mainboards to Marvel/USA and they
have cheked it directly.

I say FUCK! Anything was right, BUT, two address lines to the DDR2 RAM
where 2,8mm to short and one 3,2mm to long. It created interferences!

Now the 5th Mainbaord is working since April 2011...

The problem is now, manufacturing 1000 Mainboards for a PanelPC using:

1) Marvel Discovery MV78200 (1200 MHz, Dual-Core)
2) 1 GByte DDR2
3) 2 x MiniSDHC 32 GByte (it support SDXC too and this is, WHY I had to
choose the "MiniSD"and not "MicroSD")
4) 2 GigaEthernet
5) 3 x USB 2.0
6) 2 x SATA 2.0
7) Useing an FTDI2232 for JTAG and UART for easier flashing
8) 2 x PCIe Port (x4 or Quad 1x)

a) 1x PCIe 4x => Marvel SAS/SATA Controller (8 Drives)
b) 1x PCIe 1x => USB 3.0 2-Port Controller + two 4 Port Hubs
c) 1x PCIe 1x => NVidia TFT/HDMI Grafik Driver
d) 2x PCIe 1x => PCIe Device controller
(I use it like a 8051 with MANY I/O-Ports)

cost arround 280 Euro including my development costs and excluding VAT!
The advantage this Mainboard is that I know the hardware and can install
Debian GNU/Linux ARMEL.

If you realy think you can get a full featured computer for 149US$ which
allow you t install whatever you want, forget it!

I have te same Experienc ad Luke and it is frustrating!

It does not mather in which Asian country (China, Hongkong, Taiwan,
Corea, Malaisia, Singapur, Pakistan) you go, they will fool you in all
directions!

However, I was switching on 2011-05-23 to my new location in Germany and
running into the next problem with the french authorities which
confiscatd my Truck including my whole Electronic Lab (~80.000 Euro) an
the shortly before buyed Microchips (1000x Marvel Armada 300, 2000x
Marvel Discovery MV78100, 200x Marvel Discovery MV78200, 10000 DRR2 RAM,
2400 NAND Flash, PMICS, Grafic Chips, etc.)

Currently I am more or less Jobless... and can not do anything anymore.

However, if someone is interested Mainboards for industrial PanelPC or
TabletPC, please let me know.

I will restart my development shortely (need to survive next 3 month)
and have created for each Project a mailinglist...

Note1: I evaluate following Microcontrollers:
1) TI Sitara AM 3517
2) TI OMAP 3530
3) TI OMAP 4430 (if TI does not lower the equiremens
for developers, forget this µC)
4) Marvel Armads 100
5) Marvel Armada 300
6) Marvel Discovery MV78100
7) Marvel Discovery MV78200 (Dual-Core)
8) Marvel Kirkwood 88F6281 (same µC as the Seagate DockStar)
I have ALL Development Kits (>18.000 euro) here and not only the
inexpensive Evaluation Kits. So I think, it is time to earn money
with it. :-S

Note2: I have signed the hell of NDAs with Texas Instruments and Marvel,
which does NOT mean, I can not have GPLed drivers...

Thanks, Greetings and nice Day/Evening
Michelle Konzack
--
##################### Debian GNU/Linux Consultant ######################
Development of Intranet and Embedded Systems with Debian GNU/Linux
Internet Service Provider, Cloud Computing
<http://www.itsystems.tamay-dogan.net/>
<http://www.debian.tamay-dogan.net/>

***@tdnet Jabber ***@jabber.ccc.de
Owner Michelle Konzack

Gewerbe Strasse 3 Tel office: [phone]
77694 Kehl Tel mobil: [phone]
Germany Tel mobil: [phone] (France)

USt-ID: DE 278 049 239

Linux-User #280138 with the Linux Counter, http://counter.li.org/

[David Goodenough]

Post by Rob van der Hoeven

Post by Phil Endecott

Post by Rob van der Hoeven

Post by Arnaud Patard (Rtp)

Post by Rob van der Hoeven
I think the hardware of this tablet can also be used as a server or
desktop computer. The tablet is mass produced and very cheap (i got
mine for 149 euro).

For that price, to make a server, I would rather buy a loco board or
any other development board

These boards are not mass produced which makes them relatively
expensive.

The i.MX LOCO board, the OMAP panda board, and some of the others cost
about the same as your tablet.

Panda board has very nice specs.

Post by Phil Endecott

Post by Rob van der Hoeven
Hardware that is not mass produced has some other issues,
namely availability and vendor lock-in

You think that your tablet is going to have better availability and less
lock-in than a board from Freescale or TI? That seems unlikely to me.
Look at the BeagleBoard; it would be hard to find any smartphone or
tablet device that has been available for as long as that has.

The FreedomBox project is looking for very cheap hardware. This hardware

What about the shortly to be released Raspberry Pi?

David

Post by Rob van der Hoeven
exists today, but it is used for running Android. It would be very nice
if we could liberate this hardware and use it for our own computing.
Beagle board and Panda board are very nice but i don't think they will
become cheap enough for the FreedomBox (one monopolistic manufacturer,
low volume - "only" 8900 Panda boards sold)
If the FreedomBox would use a popular SoC then the manufacturer of the
motherboard seems less important to me. All the major functionalities
for which drivers are needed are on one chip. We could simply switch to
an other board with the same SoC and still run our software (maybe with
some minor adjustments, please correct me if i am wrong...)

Post by Phil Endecott

Post by Rob van der Hoeven
I think it must be possible to buy an android motherboard for just a
fraction of the price that i paid for my tablet.

Why do you think that? I have personally never seen an "Android
motherboard" offered for sale at all, let alone for a low price.

You find out the manufacturer of the motherboard inside a tablet. Then
you contact this manufacturer and say: Hi, i know you are making 10000
motherboards for Yarvik, if i were to order 1000 of these boards what
would the price be? I think the manufacturer will be happy with an extra
order. Mass produced boards are well tested (the manufacturer simply
can't afford mass problems) and cheap.

Post by Phil Endecott

Post by Rob van der Hoeven
Why is relying on
hardware with a SoC such a bad idea? If the SoC is popular it will not
go out of production for a long time.

No, that's not how it works. Both popular and unpopular chips are
replaced on a schedule that's determined by advances in manufacturing
technology. This also applies to the consumer products that are made
from them: even if a device is popular, it will soon be replaced with
something that is faster and cheaper.

Not quite true i think, especially for SoC. My FreedomBox has an Marvell
6281 (Kirkwood) SoC inside. This chip has been around for a long time.
You are right for non-SoC boards, they can more easily change for
example the graphics chip and spoil our fun.
Regards,
Rob van der Hoeven.
http://freedomboxblog.nl

[Rob van der Hoeven]

Post by David Goodenough

Post by Rob van der Hoeven
The FreedomBox project is looking for very cheap hardware. This hardware

What about the shortly to be released Raspberry Pi?

I like the Raspberry Pi model B ($35 with 10/100Mb Ethernet). I will
probably buy one. But... the subject of this tread is: Debian GNU/Linux
on tablet hardware. The discussion is becoming much to broad now (all my
fault, i started using the FB word :-)

Post by David Goodenough
From what i understand it is not easy to use GNU/Linux on tablet

hardware. Even if this hardware is running a Linux kernel. Such a
shame...

Regards,
Rob van der Hoeven.

http://freedomboxblog.nl

[Luke Kenneth Casson Leighton]

On Fri, Oct 28, 2011 at 8:42 PM, Rob van der Hoeven

Post by Rob van der Hoeven

Post by David Goodenough

Post by Rob van der Hoeven
The FreedomBox project is looking for very cheap hardware. This hardware

What about the shortly to be released Raspberry Pi?

I like the Raspberry Pi model B ($35 with 10/100Mb Ethernet). I will
probably buy one.

yes... that's not $25, is it?

Post by Rob van der Hoeven
But... the subject of this tread is: Debian GNU/Linux
on tablet hardware. The discussion is becoming much to broad now (all my
fault, i started using the FB word :-)

:) ok then strictly speaking i'm off-topic, too. ok, not exactly,
because of this:
http://elinux.org/Embedded_Open_Modular_Architecture/PCMCIA/Tablet

right - there you go: discussion back "on topic" :)

so, yep, EOMA/PCMCIA compliant tablet. mass-volume BOM: about $40
(using chiprise aka born aka allwinner chipsets). $3 for casework,
$15 for the PCB, $15 for an 800x480 7in LCD with resistive touch
panel, $7 for a battery. those are _very_ approximate prices, btw,
and exclude profit margin, a box and a PSU (none of which will be a
lot of money on a mass-volume product).

so that would be a "split" PCB (ok maybe it's $16 instead of $15)
which would _happen_ to be split along the EOMA/PCMCIA lines...

@BEGIN OFF-TOPIC
such that it would _happen_ that you could have a "FreedomBox Chassis"
and casework for about $2.
@END OFF-TOPIC.

Post by Rob van der Hoeven

Post by David Goodenough
From what i understand it is not easy to use GNU/Linux on tablet

hardware. Even if this hardware is running a Linux kernel. Such a
shame...

that's usually because of GPL violations involving the bootloader.

so, ye gods, let me correct the request again, ok just assume that i
put "and the bootloader and the toolchain and the kitchen sink" into
the request.

l.