[Arm-netbook] fosdem2016

[GaCuest]

En 4 de febrero de 2016 en 0:47:41, Luke Kenneth Casson Leighton (lkcl at lkcl.net) escrito:
> On Wed, Feb 3, 2016 at 10:52 PM, Wookey wrote:
> > +++ Luke Kenneth Casson Leighton [2016-02-03 19:02 +0000]:
> >> > I had a quick look but failed to find details of the screen, keyboard,
> >> > boards and 3D parts online, other than scattered through many mailing
> >> > list-messages. Never mind info like the above. Is there a page that
> >> > actually has the info someone keen would need to get started?
> >>
> >> eek - sorry :) i usually maintain a page that has them but haven't
> >> put it together yet - give me a mo and it'll be at
> >> http://rhombus-tech.net/community_ideas/laptop_15in/
> >
> > OK, cheers. So battery and touchpad not available in ones and keyboard
> > suppiers are idiots. So some associated faff there :-)
>  
> yyeahh... there's a whole stack of that
>  
> >> the boards i want to do another run in the next few weeks, wookey, so
> >> if you'd like to buy some you'd be most welcome. i don't want to get
> >> too many done in case they need modifications. they're only
> >> single-sided 2-layer 1.5mm thick so are "bog standard" i.e. dirt
> >> cheap. i think this time i'll get them made up rather than do the
> >> component assembly myself.
> >
> > I am interested, but I was a little put off by the prices you listed
> > last time at $120-150 per board and 3 boards.
>  
> a qty 5 figure is $1700 for the CPU Cards (including components and
> assembly). a qty 5 figure for these 2-layer single-sided PCBs is
> going to be waaay less than that. i'd put a guess of around $60 for
> PCB1, $50 for PCB2 and $50 for PCB3. massive difference - just
> because of using simple 2-layer and 1.5mm.
>  
> the $120-150 was because i was using eurocircuits. i think it was
> around that much for PCB1 (qty 2). i'll get everything done in china
> this time.
>  
> > I don't call that 'dirt
> > cheap'. That's $430 + screen+keyboard+panel+printing, which is a
> > little more than I want to pay for just 'mucking about'.
>  
> i added up a rough MOQ 200-1k figure today and it came out to a BOM
> of around $190, excluding assembly costs. which honestly isn't that
> big a difference from the qty5 figure.
>  
> > I could
> > afford it but a) I'm tight and b) I don't like buying electronics
> > unless I'm fairly sure I'm going to get decent use out of it (all that
> > eco-thinking).
>  
> good for you! so the question becomes: is it worthwhile for you to
> spend the time as an early adopter, to help "prove the concept" - i'm
> pretty sure it'd be possible to find a home for the end result (i have
> to give one to dr stallman for example).
> 

Well, maybe the problem is that the project initially was too ambitious 
for a small company.

I remember when EOMA-68 would be sold in stores and you could 
put it on any kind of device. It was a very good idea, but very difficult 
to do (at least without the money of a big company).

The problem is that people will be reluctant to buy a computer with 
Allwinner A20. Even the people will be reluctant to buy a computer 
without Windows or Linux (x86).

Perhaps it would be interesting to establish requirements for 
software and minimum hardware requirements as did 96boards.

> > And ultimately a 2G RAM laptop is 'toy' these days,
> > because 'browsers'.
>  
> *sigh* tell me about it... bear in mind this is only a 1366x768 LCD.
>  
> > So I was waiting to see if the upgradability
> > aspect looked likely to actually solve this issue, and I have a pile
> > of other half-started projects so don't _actually_ need any more :-)
> >
> haha
>  
> > So, er. 'maybe' :-) How much and when?
>  
> let me work it out more accurately, likely timeframe 2-3 months. i'd
> like it to be before 2 months as i'm leaving den haag end of march.
>  
> >> so if someone can confirm whether this is true or not, i'll
> >> re-prioritise the allwinner A64 board back to the top of the TODO
> >> list.
> >
> > Karsten Merker explained this at some length after my talk (he has a
> > pine). Allwinner's first-stage (non free, probbaly not even
> > redistributable) bootloader initialises the RAM, but we have no docs
> > to do it in uboot/uefi. Someone cut out the blob and linked it in
> > which works, but that's not redistributable either. So yes RAM init is
> > a roadblock until we can get someone at AW to tell us how to do it, or
> > it's otherwise revenged.
>  
> you've seen the lichee A64 source code from the a64 sdk, right?
> links and mirrors were discussed here about 2 months ago, but i'm
> seeing full source including "init_DRAM" which all looks fine... i
> mean they forgot (again) to put a GPL header on the file (mctl_hal.c)
> but other than that it looks fine... let me just upload the u-boot
> source that i have here to hands.com... here y'go:
>  
> http://hands.com/~lkcl/u-boot-2014.07.tgz
>  
> can you put me in touch with karsten?
>  
> > I was going to try and lean on them from the Linaro end (Connect in
> > March) and see if we can get any joy, but it may well be difficult.
>  
> well let's see if that source (which includes boot0 full source, it
> seems - no .o or .a files) does the trick, first.
>  
> l.
>  
> _______________________________________________
> arm-netbook mailing list arm-netbook at lists.phcomp.co.uk
> http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook
> Send large attachments to arm-netbook at files.phcomp.co.uk