[Arm-netbook] updates from eoma68-a20

Luke Kenneth Casson Leighton lkcl at lkcl.net
Sun Jan 28 12:53:49 GMT 2018


---
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68


On Sun, Jan 28, 2018 at 9:12 AM, Tor, the Marqueteur
<Marqueteur at fineartmarquetry.com> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> On 01/27/2018 10:40 PM, zap wrote:
>> lkcl wrote:
>>> ... which wouldn't cost that much.  my main concern would
>>> actually be: we have quite a lot of empirical evidence to suggest
>>> that most people don't f*****g well read the f*****g updates.
>>> witness the last set of complaints - RIGHT HERE - where people
>>> said, "but but you didn't say anything about that!!!" and i
>>> pointed them DIRECTLY at MULTIPLE updates which specifically,
>>> specifically demonstrated that they had simply not been paying
>>> attention.  at all.
>> Hmm...  I have read most of them. Although I may have skimmed
>> through a few. Mostly because they had to do with hdmi which is not
>> my main interest.
>
> Unfortunately, if there is more than one topic in an email, post, etc.,
> many people seem incapable of noticing it.  Even when it seems they are
> being paid to read the thing.  I don't know of a good solution there.

 it's ok: right now the project is not targetted 100% at busy
average-use-case people.

>>>
>>> so, sad as it is to have to point it out, i really do not have
>>> *any* confidence that people will actually bother to engage with a
>>> poll in any meaningful way.
>>>
>> Hm, you may have a point. Its not like you have a forum anyways
>> after all right? and know who knows if they would after that
>> anyways.
>>
>> I am sorry if I came across as demanding, I just need 2gb of ram at
>> the least for it to be usable on a regular basis for myself.
>
> IIUC, the current state of affairs is that:
> 1.  The design costs have run high due to issues found in the prototype
> that was supposedly working after being designed by a competent designer.

 ok clarification:

 * 5 years ago the first version we paid USD 10,000 for them to fail
to fucking well listen.  i said the OUTER dimensions were 54 x 96,
sent them the datasheet with the part from china and expected them to
read it and communicate with me about the internal PCB dimensions
needed to fit into the casework.  luckily, when the PCB turned up and
was 54x96 mm in size, there were critical flaws which indicated that
the JUNIOR pcb engineer given the task had completely fucking well
failed to even run the Design Validation Checks.

 * 4.5 years ago the second version was handed over to a COMPETENT
engineer... we were ONLY charged USD 5,000 for this service.  i
decided that from this point onwards i would do the PCB layout myself.

 * 4 years ago the HDMI connector utilised on this second revision
went END OF LIFE.

 * 2 years ago the REPLACEMENT HDMI connector went END OF LIFE

 * 1 year ago yet *ANOTHER* replacement HDMI connector turned out not
to be available in reasonable quantities at reasonable prices.

 * 6 months ago the pressure on legacy TSSOP48 NAND (the A20 having a
BOOT ROM that only reads older NAND block sizes / speeds) became too
great and i went "fuck it, it's gone".

 *NONE OF THESE THINGS COULD HAVE BEEN AVOIDED*.  you cannot know what
you do not know that you do not know.


> 2.  The time required has similarly run high for the same reason.

 naturally.

> 3.  In that time prices on key components have risen, as much because
> of a fluke of timing as because of valid price changes.

 this is the latest development, yes.

> 4.  Given current prices, the only known way to make everything work
> with available resources is to put in lower capacity RAM ICs.

 we don't know that yet.  it is *suspected* that that is the case.


> While even a dedicated update to backers might get ignored, and it's
> not ideal in the first place, we obviously have some people prepared to
> pay significantly more than the price increase to get the 2GB RAM.

 .... and be prepared to cover the *INCREASED* cost due to reduced
quantities but not a reduction in the fixed overheads associated with
component sourcing, supply, and PCB production *and* PCB assembly...
yes.

> Should a crowdfunding, all or nothing, be considered to gauge how much
> people care about the difference in RAM?

it could... if someone else is prepared to take responsibility for it.
i'll be absolutely honest: i'm getting tired of handling this stuff on
my own.

l.



More information about the arm-netbook mailing list