Luke,

First thing will be building a 3d model of the board using FreeCAD. Not very easy to do accurately with the info I currently have. More photos and dimensional information would certainly be welcome!

I'll keep this list in the loop.

Kind regards,

Mark

2017-03-20 7:04 GMT+01:00 Luke Kenneth Casson Leighton lkcl@lkcl.net:

PLEASE KEEP ON-TOPIC on this subject, until at least 1st april, thx

question from mark so clarifying here, he asked "what mounting holes can be used from the PCB" and there aren't any, mark.

the stack that was done 2 years ago the micro-desktop PCB was "loose" inside the case, using close-fitting to ensure it didn't move about too much.

if you review the photos on crowdsupply (i have some more somewhere... chris can you remember where they are, after you did those beautiful drawings?) you can see that there are 4 screws which keep the entire stack together, these do *not* go through the PCB.

the baseline stack height is 3mm.

the new USB connector is 6mm high

the VGA connector and PJ-045 (DC jack) i believe are both 9mm

the micro-sd card slot (on the bottom, other end) is 2mm

the EOMA68 (PCMCIA) is 6mm.

so it all works out pretty well: you end up in some cases with two halves per horizontal stack but that's fine because there are 2 screws each side to hold them together.

what i can do, mark, is if we get the DXF files sorted i can turn them into 3D-printed "things" and then do a video (which you can publish on your website, yay!).

we can do a few rounds at it.

l.

---

arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk

On Tue, Mar 21, 2017 at 11:39 AM, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:

On Tue, Mar 21, 2017 at 10:33 AM, Mark Van den Borre mark@lusis.eu wrote:

...

Luke,

First thing will be building a 3d model of the board using FreeCAD. Not very easy to do accurately with the info I currently have. More photos and dimensional information would certainly be welcome!

let me take care of that, it should take me 20 mins with pyopenscad.

ok done.

http://hands.com/~lkcl/eoma/tablet/3dcase then you should find microdesktop_model2.stl

if you send me a first DXF file what i can do is throw it into that model and generate an STL and some pictures to see what it looks like. openscad can do "extrude DXF" to turn it 3D.

l.

On 03/21/2017 02:42 PM, Jonathan Frederickson wrote:

On Tue, Mar 21, 2017 at 3:10 PM, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:

...

ok done.

http://hands.com/~lkcl/eoma/tablet/3dcase then you should find microdesktop_model2.stl

http://hands.com/~lkcl/eoma/tablet/3dcase 404s here, and http://hands.com/~lkcl/eoma/microdesktop/ (which was my next guess) doesn't have a '3dcase' directory or 'microdesktop_model2.stl' file.

This appears to be the correct link:

http://hands.com/~lkcl/eoma/kde_tablet/3dcase/ http://hands.com/~lkcl/eoma/kde_tablet/3dcase/microdesktop_model2.stl

Hello Luke, hello list,

Here's a FreeCAD solid model:

http://glasno.st/microdesktop_model2.fcstd

Kind regards,

Mark

2017-03-21 21:54 GMT+01:00 Luke Kenneth Casson Leighton lkcl@lkcl.net:

On Tue, Mar 21, 2017 at 7:45 PM, Joseph Honold mozzwald@gmail.com wrote:

...

On 03/21/2017 02:42 PM, Jonathan Frederickson wrote:

...

On Tue, Mar 21, 2017 at 3:10 PM, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:

...

ok done.

http://hands.com/~lkcl/eoma/tablet/3dcase then you should find microdesktop_model2.stl

http://hands.com/~lkcl/eoma/tablet/3dcase 404s here, and http://hands.com/~lkcl/eoma/microdesktop/ (which was my next guess) doesn't have a '3dcase' directory or 'microdesktop_model2.stl' file.

This appears to be the correct link:

http://hands.com/~lkcl/eoma/kde_tablet/3dcase/ http://hands.com/~lkcl/eoma/kde_tablet/3dcase/microdesktop_model2.stl

thx for raising (and finding) that.

---

arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk

Hello Mark,

We had a short exchange on #freecad 2 days ago (my nick is normandc). I'm mostly active on the FreeCAD forum. Let me know if you need help, I'm pretty proficient with the GUI side of FreeCAD.

Cheers

Normand

On 2017-03-21 at 18:24, Mark Van den Borre wrote:

...of the board obviously. Even that took time fiddling with the (mesh) stl...

Some very friendly people in irc.freenode.net#freecad http://irc.freenode.net#freecad helped me out.

Regards,

Mark

Hello Luke,

On 2017-03-20 at 02:04, Luke Kenneth Casson Leighton wrote :

the new USB connector is 6mm high

the VGA connector and PJ-045 (DC jack) i believe are both 9mm

the micro-sd card slot (on the bottom, other end) is 2mm

the EOMA68 (PCMCIA) is 6mm.

Could you confirm that your microdesktop_prog2.scad file (used to generated microdesktop_model2.stl) is _not_ up-to-date with these dimensions? Taking measurements from Mark's FreeCAD file converted from your stl:

- USB connectors are 6.65mm high rather than 6 - PJ-045 (thin tall one at rear extreme left?) is 11mm high rather than 9 - VGA is 9mm - microSD slot (at the front under the board I presume) is 1.8mm rather than 2

Mark, I imported the scad file directly into FreeCAD, import wasn't perfect but I fixed the issues, all shapes are separate primitive cubes that can be re-sized through their properties. I left the EOMA-68 card separate as a mesh to bring the file size down to <160KB.

https://drive.google.com/file/d/0B8EXnjMkh8zVNXFvRjA5Qm0tSlU/view?usp=sharin... (apologies for the Google hosting...)

Edit --> Project info reports license as GPLv3 (it's saved into the document).

On the manufacturing side, I wonder if it may be a good idea to check the real thickness of the 3mm plywood boards. Due to manufacturing tolerances, materials provided in sheets are often thinner than their nominal thickness be it for plywood, plastics etc.

As for the screws, difficult to evaluate from a couple of pictures but they seem to be M3 socket head, black oxide, with hex nuts at the bottom (possibly 2 per screw) with some kind of either plastic or vinyl cap. Nickel plated screws would be better IMHO (considering the hex nuts are either nickel or zinc) and not prone to rust like black oxyde screws, but they might be difficult to source and more expensive.

Cheers,

Normand

On 2017-03-21 23:45, Alexander Ross wrote :

heh well i dread to think how much javascript is loaded compared to 156kb file... so i dumped it at https://transfer.sh/kvi63/microdesktop-model2-normandc1.fcstd for easy china-net downloading, i hope. however it expires in 2weeks.

Thanks. Figured Github wouldn't be much better. Actually that's the purpose I plan for the microdesktop I pledged: self-hosting!

Forgot the link to the calculator for changing wood dimensions in changing climates:

http://owic.oregonstate.edu/wood-shrinkswell-estimator

Tor

On 21/03/17 17:56, Tor, the Marqueteur wrote:

On 21/03/17 17:07, Normand Chamberland wrote: ...snip...

...

On the manufacturing side, I wonder if it may be a good idea to check the real thickness of the 3mm plywood boards. Due to manufacturing tolerances, materials provided in sheets are often thinner than their nominal thickness be it for plywood, plastics etc.

Good point. Also, with plywood, the thickness /will/ change in different climates. Length and width should stay pretty stable, enough that I wouldn't consider worrying about anything other than perhaps two spaced solder joints directly being pushed/pulled by the case.

Taking quartersawn Birch (approximately correct for rotary cut hardwood plywood, going from an air conditioned desert conditions (70F, 10%RH) to an open tropical environment (80F, 90%RH), one could expect the case to swell .06" (or about 1/16") per inch of thickness.

...

As for the screws, difficult to evaluate from a couple of pictures but they seem to be M3 socket head, black oxide, with hex nuts at the bottom (possibly 2 per screw) with some kind of either plastic or vinyl cap. Nickel plated screws would be better IMHO (considering the hex nuts are either nickel or zinc) and not prone to rust like black oxyde screws, but they might be difficult to source and more expensive.

IME, (US based) zinc plate is usually easy to come by and fairly cheap. The nice HW that really won't rust and is expensive is SS.

Tor

...

Cheers,

Normand

Hello list, hello Normand,

2017-03-22 4:07 GMT+01:00 Normand Chamberland gemnoc@gmail.com:

On the manufacturing side, I wonder if it may be a good idea to check the real thickness of the 3mm plywood boards. Due to manufacturing tolerances, materials provided in sheets are often thinner than their nominal

thickness

be it for plywood, plastics etc.

Indeed, no ply board thickness ever is exactly 3mm. Either over or under 3mm, plus climate variations. The good news is that the plywood is fairly dimensionally stable in width and length.

As for the screws, difficult to evaluate from a couple of pictures

Have you found the extra pictures? I've only seen the tiny extra picture at crowdsupply yet ...

seem to be M3 socket head, black oxide, with hex nuts at the bottom (possibly 2 per screw) with some kind of either plastic or vinyl cap.

Nickel

plated screws would be better IMHO (considering the hex nuts are either nickel or zinc) and not prone to rust like black oxyde screws, but they might be difficult to source and more expensive.

ISO 7380 bolts with either plastic cap nut or hex nut with plastic cap would be nicest I guess. I plan to drop by my excellent supplier early next week to see what they recommend.

Kind regards,

Mark

Hello Luke, list,

This is work by Normand, FreeCAD expert and kickstarter backer:

http://glasno.st/microdesktop_model2_normandc3.fcstd http://glasno.st/microdesktop_layers_v1.dxf

Kind regards,

Mark

2017-03-28 9:58 GMT+02:00 Luke Kenneth Casson Leighton lkcl@lkcl.net:

hiya mark (and list),

i started taking a look at creating case layers for the microdesktop, and it's not as straightforward as it seems: creating a stack that will ensure that the PCB doesn't drop out at one end is not as simple as it first seems. so i'll take care of it.

l.

---

arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk

On 29/03/17 17:34, Normand Chamberland wrote:

Hello,

Thanks Mark. Just to give some context to these files:

• There are 11 lasercut wood pieces. That's a lot. Is the layered design

definitive (I guess that's what the backers will expect since that's what's shown on Crowd Supply), or could this be discarded in favour of a new design optimizing fabrication? A more classic box with dovetail joints may be less expensive, and allow more control on the fitting of the board. Something akin to https://images-na.ssl-images-amazon.com/images/I/91H9TOVowCL._SL1500_.jpg (but hopefully with better looks!)

From a wood engineering standpoint I'd be in favour of looking into this. By using this method, wood movement with climate becomes negligible, which makes things a good deal easier there.

Artistic only: The burned edges are arguable in any logo cutouts, though less than optimal visually elsewhere (IMO). If laser-cutting, however, there is no choice, and the edges of a laser cut stack will exhibit the same burned look, just all in a band around the outside, rather than in scattered bits along the corners. The example on the Crowd Supply page appears to have been significantly sanded to minimize the burnt edges. This is a time consuming operation that will reduce the accuracy of edge alignment.

Any decent laser cutter can be used as a laser engraver to put (stopped) dadoes to hold PCBs anywhere you want them. You can also put a rabbet on any layer for the same purpose as needed in the layered option, so that isn't a serious obstacle. The only (minor) limitation with rabbets is that at the registration accuracy likely required, having a rabbet on both sides of a layer for different boards is probably unworkable.

Tor

...snip

On 30/03/17 03:46, Vincent Legoll wrote:

...

From a wood engineering standpoint I'd be in favour of looking into this. By using this method, wood movement with climate becomes negligible, which makes things a good deal easier there.

We're talking about plywood here, and smallish pieces, so I think the climate- induced wood movement will be negligible whatever we choose. Unless you sunk it. ;-)

Length/width, you're entirely correct, and that holds true for significantly larger pieces than we've got here. The laptop needn't worry.

Thickness is where we lose all plywood moisture stability with the stacked option. 8 layers, 3mm thick = ~24mm thick, or a little less than 1". As I posted earlier, between someone living in a desert, and someone living in a wet tropical area, the latter may have a case almost 1/16" or 1.5mm thicker than the desert person.

Tor

Idiot question from someone who does not work with wood very often.

Can't plywood, even little thin stuff like this, be sealed against moisture penetration? That stuff's gonna mold sooner or later otherwise, anyways...

I'm in North Carolina, USA. I guess mold and mildew aren't that big an issue after all... like I said, I don't work with wood much.

Hi Tor, hi list,

2017-03-30 23:04 GMT+02:00 Tor, the Marqueteur < Marqueteur@fineartmarquetry.com>:

I haven't put my dial caliper on enough pieces of similar plywood to make a statement about normal variation due to manufacturing tolerances.

I have. The 3mm birch plywood we use usually stays within +-0.1mm. Only rarely up to +-0.15mm.

Some extra stabilisation can also be designed into the z dimension by doing things like inserting (laser cut) plywood "grip beaks" along the z axis.

Another laser cutting technique that may come in handy is the "accordeon curve" or "kerf bend".

Kind regards,

Mark -- http://lusis.eu full service b2b laser cutting

Hello,

2017-03-30 21:36 GMT-04:00 Tor, the Marqueteur < Marqueteur@fineartmarquetry.com>:

This could be used in conjunction with a dado in the top and bottom pieces and rabbeting the sides. This would keep the unmarked top and bottom (with optional venting laser fretwork logos, etc.), and still allow the sides to be held. This would require flipping the sides to laser the kerfs for the bends, and alignment of PCB mounting dadoes (on the second laser operation) might be tricky at best.

The other option would be to dado the top/bottom for the full thickness of the sides and let the top/bottom extend past the sides.

If a pair of internal pieces are used for mounting the PCB/card, then you're up to eight pieces, but still less than the 11 I recall being mentioned, and probably at better material efficiency.

Tor

If I understand you, you mean something like this?

https://github.com/normandc/EOMA68/tree/master/Microdesktop/Images

Please note I didn't add the dadoes nor vents and kerfing for the curved sides.

There's really not a lot holding the PCB, about 0.6mm due to the PCMCIA Connector being so close to the edge on the right (as can be shown in images 06 & 07 which has the front plate hidden). Maybe pieces oriented sideways would be best as they could provide more support width.

Also, with the bottom plate exceeding by 3mm all around this may make it hard to insert/remove the SD card, the plate might need to be notched.

Normand

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

On Fri, Mar 31, 2017 at 6:28 AM, Normand Chamberland gemnoc@gmail.com wrote:

Hello,

2017-03-30 21:36 GMT-04:00 Tor, the Marqueteur Marqueteur@fineartmarquetry.com:

...

This could be used in conjunction with a dado in the top and bottom pieces and rabbeting the sides. This would keep the unmarked top and bottom (with optional venting laser fretwork logos, etc.), and still allow the sides to be held. This would require flipping the sides to laser the kerfs for the bends, and alignment of PCB mounting dadoes (on the second laser operation) might be tricky at best.

The other option would be to dado the top/bottom for the full thickness of the sides and let the top/bottom extend past the sides.

If a pair of internal pieces are used for mounting the PCB/card, then you're up to eight pieces, but still less than the 11 I recall being mentioned, and probably at better material efficiency.

Tor

If I understand you, you mean something like this?

https://github.com/normandc/EOMA68/tree/master/Microdesktop/Images

normand, can you please not use non-free proprietary git services for this project, it sends completely the wrong message. please delete that github repository immediately, it brings the EOMA68 project's reputation into disrepute and that's completely unacceptable.

if you do not have a savannah or other libre hosting service please send me an ssh public key and i will set you up with a gitolite repository.

l.

On 31/03/17 11:34, Normand Chamberland wrote:

FYI My repo had a README file for which content specifically said

...

This repository hosts my personal design files for the EOMA68 project. This is not the official project repository.

I fail to see how they could bring "disrepute" to it.since these files were clearly *PERSONAL* and *NON-ASSOCIATED TO THE OFFICIAL PROJECT*.

I spent valuable time setting this up but fine, I just deleted it.

if you do not have a savannah or other libre hosting service please
send me an ssh public key and i will set you up with a gitolite
repository.

No, right now I'm too fucking pissed off by your zealotry to do anything resembling that and I might very well stop my contributions right there.

I understand your frustration, Normand, but I hope you will continue to contribute to the project. It's important work and your contributions, and those of the community as a whole, are extremely valuable.

May I suggest a reading of [ETHICALREPO] so that you may see that "zealotry" is an unfair categorisation: github (and others) exhibit problems that have a real impact on freedom.

You should of course have the freedom to publish your personal design files wherever you wish, but the files in question represent, in some way, your contribution to a public project. As such, consideration must be given to the goals of the project and the ideals that underpin it.

Please keep up your good work! The success of the project depends on contributions like yours.

jah

[ETHICALREPO]: https://www.gnu.org/software/repo-criteria-evaluation.html

On Sun, Apr 2, 2017 at 11:45 PM, Normand Chamberland gemnoc@gmail.com wrote:

2017-03-31 7:14 GMT-04:00 jah jah@jahboite.co.uk:

...

I understand your frustration, Normand,

You do not know me, jah, therefore you cannot possibly understand how *infuriated* I am. To say that I am still seething with cold fury would be an understatement.

normand, please don't feel that way: it's not worth the harm that it causes. i know because i've had standing-alpha-waves occur many times and they literally made me dizzy and sick to my stomach.

Of course I know that Github is not the "ideal" channel to share files (to say the least). I do not need this list's continued condescending lessons on the matter.

please, take it as absolute that in no way is anyone here intending to be condescending towards *anyone*. if however there is a misunderstanding as to the priorities, it would be completely remiss to leave it for those people *who do not* understand, to not explain it. this is a list with now around 900 people, there are 2,500 backers many of whom will *not* understand the importance of sticking to ethical principles, to whom it *will* be necessary to explain them from first principles.

it is therefore implicitly understood -i trust - that such explanations are there *for archive and reference purposes*, they're *not* there to "teach you how to suck eggs", or for the purposes of condescention, or other denigrating purpose which is directly against the whole damn point of this exercise.

But I spent a good part of that night whipping out these files, and it was well past midnight, I was tired and I still had to publish these files; I naively thought it was more important to get the files out there for discussion than to delay one more day to find an alternate hosting solution

... yehhh, sadly, if a project's based on libre principles and there are backers that have _specifically_ funded the project because of that, ignoring those principles - for *any* contributor - really isn't an option, normand. the irrevocable damage to the project's reputation *and* that of the FSF who are trusting me with *their* reputation... the consequences just don't bear thinking about.

• as I already had a Github account. With the idea to transfer

everything to a more acceptable server soon after.

norman i've said i'm sorry that i was too exhausted and too ill at the time to mention the offer to provide you with access to my server, what more am i supposed to do?

Only to read lkcl's message at awakening after a shortened night of sleep.

ah... now i understand.

That will teach me to try to help.

2017-03-31 7:52 GMT-04:00 Luke Kenneth Casson Leighton lkcl@lkcl.net:

...

i don't _enjoy_ asking people not to use proprietary services,

You didn't ask, you *demanded* in no uncertain terms, such as ***unacceptable*** and ***immediately***.

yes. i did apologise. i can't apologise again enough for that, and for not mentioning that i have had available libre hosting services for 10 years, now.

I would have understood if you wrote "I would really appreciate it if you took it down at your earliest convenience". But what makes me most mad is your intolerable accusation that through my action I was bringing "disrepute" to your project.

unfortunately it really does risk bringing the project into disrepute [and i was too exhausted at the time to write anything but the briefest and most immediate of response].

if the backers who expect this project to be developed using libre hosting were to find out that they're expected to download files from github then i'd be completely fucked. i explained (in a message that i believe from your response you haven't yet seen) in an extremely damaging hypothetical scenario as to why.

I find this assertion not only asinine but *utterly unacceptable*, it is an *attack* on my own reputation that I will not soon forget nor forgive.

we misunderstood, we cleared it up, no harm has been done to the project (because you took the publicly-published files offline from github immediately, despite being angry for having been asked to do so), so the envisaged hypothetical future scenario i described is not going to happen.

so there are no attacks, no harm (intended, implicit or explicit) to your reputation, okay?

a bit of reflection when you're not so white-hot angry and thus highly likely to mis-read what i've written will, i trust, allow you to appreciate that.

To make it worse, you're now saying that had you known about it months after the fact, you'd have considered it "nothing to stress about". That's plain adding insult to injury.

eh? what?? you must be misunderstanding: let me try to work out why.

* files were published on github (publicly, not private paid-for) * link sent publicly to list (which is accessible to backers and also permanently archived) * link sent at the time where it would result in the highest amount of attention * thus would result in huge numbers of (libre) backers noticing that proprietary services had been utilised. * probability that someone would complain: unacceptably high

vs HYPOTHETICAL scenario

* files published on github (publicly, not private paid-for) * *PRIVATE* emails sent to interested parties * link *NOT* published on list (which is accessible to backers and also permanently archived) * casework developed, finalised and backers received their casework. * total number of backers noticing that proprietary services had been utilised: zero. * total number of backers noticing over the next 6 months: zero * total number of backers noticing over the next 18 months: probably zero * probability that any given backer would complain: negligeable

does that help clarify the difference between the scenario that actually occurred and a hypothetical scenario where i would "grumble but not make anything of it"? in that *hypothetical* scenario i would be assuming that it is *months* after the files had been developed and the casework sent out to them.

I've done my share to promote free software as my skills allow me, in the past 7 years spending _thousands of hours_ helping quite possibly hundreds of people to learn FreeCAD, translating it, writing up documentation, packaging it and providing an experienced CAD user's point of view to its developers.

I am not dedicating most of my free time to be treated that way.

yyehhhh it's a bitch, isn't it? i met someone around 1998/1999 who said, when redhat did their IPO, "oh, i wrote a mouse howto for the linux kernel and i got enough money from the redhat IPO to be able to pay off my mortgage in full. you wrote over a hundred thousand lines of reverse-engineering code for samba and dedicated four years of your life to that, you *must* have got more than me, right?"

and he was terribly embarrassed to learn that i had received absolutely fuck-all.

i was particularly pissed off, a year later, to learn that jeremy allison had managed to get his brother (who's not written a single line of free software source code in his life) onto the VA Linux IPO - one of the highest recorded single-day jumps in the history of NASDAQ from its private IPO of $14 to a whopping $330 per share - but they neglected to include me in that, as well.

there's this terribly strange cognitive disconnect between people like us providing a *HUGE* service to others... and then not being financially rewarded for doing so because the people we're helping think "oh, it's free, there's nothing to stop me from taking, taking, taking.... so that's what i'll do. huh... nobody stopped me so... i'll just carry on".

proprietary software, there's the "sale of contract" - the opportunity is presented and *requires* that people hand over money which helps (we assume) to support the continued development of that software. or at least rewards the developers for having worked on it up to that point.

point is: i know how you feel. it's shitty, isn't it?

... now imagine having gone through that shitty treatment for OVER TWENTY FUCKING YEARS, normand, where time and time again you get slammed and taken advantage of not just by the users but BY FREE SOFTWARE DEVELOPMENT TEAMS AS WELL.

i have a long memory: there is a list of people whom have helped me out, i cannot begin to describe how grateful to them i am. i maintain that list because when there is a financially-stable self-sustaining business that will allow me to fund them, i will be contacting them and will just simply... send them a cheque with no conditions attached.

but i have learned - through people fucking well taking advantage time and time again - that setting up a *software* company alone based on software libre principles - is totally pointless.

that's why i've got into *hardware* because it is *hardware* that people, quite obviously, have to pay for. they get a physical item; that physical item has to have been manufactured. there therefore *has* to be an exchange / reward: there is no "disconnect" as outlined *and required* by software libre licenses.

but if that hardware goes out with any kind of taint from proprietary business practices... if i in any way copy those people whom i despise for them having sucked software libre dry and taken advantage of so many of us... i can't bring myself to do that.

i don't tell people very often these days of the consequences that the unethical decisions that various proprietary businesses have had for me, when over decades they've made hundreds of millions of dollars based on my work AND HAVEN'T PAID ME A CENT. having to work on a building site, sleeping on-site in a garage for three months, ending up homeless for two years at one point, and living in dangerous properties, putting my family at risk, because they're within the available budget (literally a tenth of what most people would deem to be an acceptable income). these aren't fun things to talk about, and the long-term effects they've had on my health... well, you know from the evidence of how i fail to respond in time to stop things like what's happened....

so i'll say it again. *i'm sorry* normand. please understand that i've been through more shit over the past two decades thanks to software libre than most people would be able to withstand. or understand. or believe. those experiences underscore what i'm doing, now, motivating me to make fucking sure no-one else ever has to go through that kind of hell, motivating me like you simply would not believe to undo as much of the harm that proprietary businesses have done as i can possibly manage.

but *thanks* to that damage i'm operating from a position where you don't know me, because my name's not "the successful and highly respected linus torvalds", my name's not "jeremy allison or andrew tridgell the highly respected developers of samba (two of the people who took advantage of my work, to their personal benefit, btw)", my name's not "bruce perens the highly respected software libre advocate", or any other person whom you would respect for their success and their reputation which came from that success.

anyway. can we please move on. if you don't want to help i'll understand, you have that right. but it would be very sad if you left based on a cascading series of misunderstandings because both of us are just... tired. i leave it with you, ok?

2017-04-02 16:55 GMT-04:00 Luke Kenneth Casson Leighton lkcl@lkcl.net:

...

norman, i haven't checked if you're still subscribed to the list

You will address me with my proper name please. That's Normand with a D.

sorry normand, for mistyping.

Obviously I have not unsubscribed *yet*.

I've been mulling over the matter over the weekend. I'm not sure I want to be associated with this project anymore. I've already modified my Crowd Supply account so I would no longer be listed as a backer. I've procrastinated about it, but I am leaning toward scrubbing my HDD from all my files relating to this project.

The irony of all of this is, the use I was planning for the microdesktop was to cut my dependency on proprietary cloud services for email, file hosting etc.

I will end with this: I was willing to provide an actual tangible contribution to this project. I ask all of those who would lecture me: apart from endless (and quite often fruitless) talk on this list, what are *you* doing to help this project, exactly?

very good point. well, of the people who write regularly, they're helping keep me sane and keeping me from making too many mistakes, if nothing else. for that alone i have to be absolutely grateful.

l.

On Fri, Mar 31, 2017 at 11:34 AM, Normand Chamberland gemnoc@gmail.com wrote:

2017-03-31 5:47 GMT-04:00 Luke Kenneth Casson Leighton lkcl@lkcl.net:

...

normand, can you please not use non-free proprietary git services for this project, it sends completely the wrong message. please delete that github repository immediately, it brings the EOMA68 project's reputation into disrepute and that's completely unacceptable.

FYI My repo had a README file for which content specifically said

...

This repository hosts my personal design files for the EOMA68 project. This is not the official project repository.

I fail to see how they could bring "disrepute" to it.since these files were clearly *PERSONAL* and *NON-ASSOCIATED TO THE OFFICIAL PROJECT*.

norman, i haven't checked if you're still subscribed to the list, but if you'd not published that link to the list, and had only discussed it in private email directly with other people, i would not have known and, on a random-chance discovery six to twelve months from now would probably have grumbled a bit, privately, but would have gone, "rhhm.... ok nothing to stress about, no impact on the reputation of the project, time to move along switfly".

fact is: we have something like 50% of the 2,500 backers supporting this project because they believe in libre ethics, and i am also answerable to the FSF for the RYF Certification (josh gay explained it to me that RYF Certification effectively means placing responsibility for the FSF's reputation into the hands of other people, which is not something they do lightly).

in the next update i was going to say, "here you can see where the files are being developed, normand has been kindly assisting, you can help him out as well if you like, by joining the list, he's operating from a libre-hosted repository, thus demonstrating a continued committment to the principles that you backed the project for with your money".

now, if any one of those libre backers had noticed it was a github repository, far from wishing to actively pitch in they would instead be sending me (or worse PUBLICLY writing) strongly-worded enquiries asking, in effect, "what the FUCK do you think you're playing at, lkcl?? i thought this was an entirely libre-based project, and here you are advocating the use of github which has an F on the GNU project's hosting watchlist?? i gave you my money because i believed in you, now you've completely destroyed my trust in you and that can never be restored or regained"

this is the sort of thing i'm up against. it has fuck-all to do with "zeal". i *made a promise* to people, they believe in me, and i simply cannot let them down else absolutely nobody will trust me ever again. the internet is *forever*. they'll see the (hypothetical) conversation in internet archives - forever - and the entirety of the last 6 years of work will have been for nothing.

so please, people, for goodness sake: if you need to host any files please don't use github or any cloud services: it's me (and the project) that get into deep shit, not you. i've got full control over a server which i've been running for 10 years, it has 4 separately-allocated static IP addresses, and i have two friends who also run servers, both of whom make their living from software libre support. we *really* don't need to use proprietary non-free services. if you'd like to host files or repositories send me an ssh key and i can sort something out.

thanks for understanding.

l.

On Sunday 2. April 2017 23.05.25 Luke Kenneth Casson Leighton wrote:

ok so i've created corner-pieces which will need to be 3d-printed, they contain slots for the PCB as well as for upright 1.5mm 3-ply sides. the far corner near the PCMCIA eject button needs some work / rethinking because there's a cut-out in the PCB.

the cut-out is basically too far back, and that's lack of iterations / planning on my part - i'm not doing another revision of the microdesktop PCB now, not at this stage, so i'll work something out.

http://hands.com/~lkcl/eoma/kde_tablet/microdesktop_3.png

This should be the following, I think:

http://hands.com/~lkcl/eoma/kde_tablet/3dcase/microdesktop_3.png

:-)

Paul

On Sun, Apr 2, 2017 at 10:27 PM, Paul Boddie paul@boddie.org.uk wrote:

This should be the following, I think:

http://hands.com/~lkcl/eoma/kde_tablet/3dcase/microdesktop_3.png

thanks paul (and marqueteur), i'm typing this in by hand so keep getting it wrong...

On 04/02/2017 04:55 PM, Luke Kenneth Casson Leighton wrote:

On Fri, Mar 31, 2017 at 11:34 AM, Normand Chamberland gemnoc@gmail.com wrote:

...

2017-03-31 5:47 GMT-04:00 Luke Kenneth Casson Leighton lkcl@lkcl.net:

...

normand, can you please not use non-free proprietary git services for this project, it sends completely the wrong message. please delete that github repository immediately, it brings the EOMA68 project's reputation into disrepute and that's completely unacceptable.

FYI My repo had a README file for which content specifically said

...

This repository hosts my personal design files for the EOMA68 project. This is not the official project repository.

I fail to see how they could bring "disrepute" to it.since these files were clearly *PERSONAL* and *NON-ASSOCIATED TO THE OFFICIAL PROJECT*.

norman, i haven't checked if you're still subscribed to the list, but if you'd not published that link to the list, and had only discussed it in private email directly with other people, i would not have known and, on a random-chance discovery six to twelve months from now would probably have grumbled a bit, privately, but would have gone, "rhhm.... ok nothing to stress about, no impact on the reputation of the project, time to move along switfly".

fact is: we have something like 50% of the 2,500 backers supporting this project because they believe in libre ethics, and i am also answerable to the FSF for the RYF Certification (josh gay explained it to me that RYF Certification effectively means placing responsibility for the FSF's reputation into the hands of other people, which is not something they do lightly).

in the next update i was going to say, "here you can see where the files are being developed, normand has been kindly assisting, you can help him out as well if you like, by joining the list, he's operating from a libre-hosted repository, thus demonstrating a continued committment to the principles that you backed the project for with your money".

now, if any one of those libre backers had noticed it was a github repository, far from wishing to actively pitch in they would instead be sending me (or worse PUBLICLY writing) strongly-worded enquiries asking, in effect, "what the FUCK do you think you're playing at, lkcl?? i thought this was an entirely libre-based project, and here you are advocating the use of github which has an F on the GNU project's hosting watchlist?? i gave you my money because i believed in you, now you've completely destroyed my trust in you and that can never be restored or regained"

this is the sort of thing i'm up against. it has fuck-all to do with "zeal". i *made a promise* to people, they believe in me, and i simply cannot let them down else absolutely nobody will trust me ever again. the internet is *forever*. they'll see the (hypothetical) conversation in internet archives - forever - and the entirety of the last 6 years of work will have been for nothing.

so please, people, for goodness sake: if you need to host any files please don't use github or any cloud services: it's me (and the project) that get into deep shit, not you. i've got full control over a server which i've been running for 10 years, it has 4 separately-allocated static IP addresses, and i have two friends who also run servers, both of whom make their living from software libre support. we *really* don't need to use proprietary non-free services. if you'd like to host files or repositories send me an ssh key and i can sort something out.

thanks for understanding.

I agree with your view point, github is not trustworthy. heck, gitlab which is a few steps better is still crap.

I am not sure about gogs though...

https://gogs.io/

I am really though glad you take this sort of thing seriously. We should all thank you for such zeal towards keeping things "Libre"

I hope you are successful Luke. And really, if normand sees this post, I hope he has a better understanding of the consequences of such actions.

Anyways though, I look forward to when you get the rockchip processor working to your specifications.

One last thing though, you said the a20 doesn't support 4.9 kernel... but what about the rk3288 processor?

l.

---

arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk

On Sun, Apr 2, 2017 at 10:51 PM, zap zapper@openmailbox.org wrote:

I am really though glad you take this sort of thing seriously. We should all thank you for such zeal towards keeping things "Libre"

mmm... zeal's a dirty word, now :)

I hope you are successful Luke. And really, if normand sees this post, I hope he has a better understanding of the consequences of such actions.

let's hope so

Anyways though, I look forward to when you get the rockchip processor working to your specifications.

i'll be happy it works _at all_!

One last thing though, you said the a20 doesn't support 4.9 kernel...

correction: it does... but not with 100% hardware functionality supported (which is no good for what i promised and demonstrated in the updates)

but what about the rk3288 processor?

mainline for everything. u-boot, kernel - everything. it's been that way for some considerable time.

i'm still struggling with reverse-engineering the proprietary DDR3 200mhz initialisation binary, working out what the discrepancies are between that and u-boot. it doesn't help that i don't know if i soldered the processor on incorrectly or any one of the four 96-pin DDR3 RAM ICs...

i've bought a stack of cheaper RAM ICs with which to experiment: they're the same ones as on the RK3288-Firefly so in theory i should be able to use exactly the same dtb file for the dram settings... except the damn impedance is only (appx) 50 ohms not 100 like it is for a 6-layer 1.6mm PCB...

basically there's too many unknowns in order to triangulate exactly where the problem might be when any one of them could be the source of the failure.

at some point i'm just going to say "fuck it", work out how to load the proprietary DRAM initialisation, then u-boot with DRAM initialisation *disabled*, work out how to get the RK3288 *out* of DRAM initialisation mode and to continue running u-boot.

apparently the internal ROM USB-loader can be asked to "shift" whatever's loaded down into a different address in memory, but how it's then executed i don't know.

l.

On Sun, Apr 2, 2017 at 11:26 PM, Luke Kenneth Casson Leighton lkcl@lkcl.net wrote:

http://hands.com/~lkcl/eoma/kde_tablet/3dcase/microdesktop_5.png

so i've added in curved sides but the DXF files (microdesktopcase_side.dxf) are generated based on the radius of the arc (and then not displayed in the image above),

i found a random description online of how to fit an arc between two points, it works well.

the next thing to do is to redo the top and bottom layers, this time not as simple rectangles but with proper arcs including on the corners.

l.

On 04/02/2017 06:17 PM, Luke Kenneth Casson Leighton wrote:

On Sun, Apr 2, 2017 at 10:51 PM, zap zapper@openmailbox.org wrote:

...

I am really though glad you take this sort of thing seriously. We should all thank you for such zeal towards keeping things "Libre"

mmm... zeal's a dirty word, now :)

My bad...

...

I hope you are successful Luke. And really, if normand sees this post, I hope he has a better understanding of the consequences of such actions.

let's hope so

...

Anyways though, I look forward to when you get the rockchip processor working to your specifications.

i'll be happy it works _at all_!

I am sure you will figure it out sooner or later.

...

One last thing though, you said the a20 doesn't support 4.9 kernel...

correction: it does... but not with 100% hardware functionality supported (which is no good for what i promised and demonstrated in the updates)

...

but what about the rk3288 processor?

mainline for everything. u-boot, kernel - everything. it's been that way for some considerable time.

i'm still struggling with reverse-engineering the proprietary DDR3 200mhz initialisation binary, working out what the discrepancies are between that and u-boot. it doesn't help that i don't know if i soldered the processor on incorrectly or any one of the four 96-pin DDR3 RAM ICs...

i've bought a stack of cheaper RAM ICs with which to experiment: they're the same ones as on the RK3288-Firefly so in theory i should be able to use exactly the same dtb file for the dram settings... except the damn impedance is only (appx) 50 ohms not 100 like it is for a 6-layer 1.6mm PCB...

basically there's too many unknowns in order to triangulate exactly where the problem might be when any one of them could be the source of the failure. at some point i'm just going to say "fuck it", work out how to load the proprietary DRAM initialisation, then u-boot with DRAM initialisation *disabled*, work out how to get the RK3288 *out* of DRAM initialisation mode and to continue running u-boot.

I am hopeful you will succeed in reverse enginnering the rk3288.

apparently the internal ROM USB-loader can be asked to "shift" whatever's loaded down into a different address in memory, but how it's then executed i don't know.

l.

---

arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk

Hello Hendrik,

2017-03-30 23:58 GMT+02:00 Hendrik Boom hendrik@topoi.pooq.com:

On Thu, Mar 30, 2017 at 10:39:01AM -1000, Tor, the Marqueteur wrote:

...

Thickness is where we lose all plywood moisture stability with the stacked option. 8 layers, 3mm thick = ~24mm thick, or a little less than 1". As I posted earlier, between someone living in a desert, and someone living in a wet tropical area, the latter may have a case almost 1/16" or 1.5mm thicker than the desert person.

Is there another material we could use that can be laser-cut but doesn't have problems with dimensional stability?

Acrylic is the obvious answer to this question. Though I'm quite sure the plywood's potential z dimensional stability issues can be worked around...

Kind regards,

Mark

-- hendrik

---

arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk

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

On Thu, Mar 30, 2017 at 4:34 AM, Normand Chamberland gemnoc@gmail.com wrote:

Hello,

Thanks Mark. Just to give some context to these files:

• This is a first draft. Even though Luke wrote he'd take care of it, I had

already started the process with Mark last week.

not a problem, let's use (keep going with) both approaches, each of us will spot different things, by using different tools.

I didn't spend more than a couple of hours on this last night and I didn't calculate precise clearances, the PCB as per Luke's STL model will fit but there may be too much or not enough clearance along X and Y axes (X and Y being parallel to the ground).

what i'll do is 3D print all the parts from DXF-extrusion using pyopenscad, and double-check them that way.

• It requires 8 layers of 3mm wood panels, overall dimensions of casing

(without screws) being 111mm wide by 115mm deep by 24mm high. Consequently the PCB is near flush to the front and rear of the casing.

• For 3 layers, the panels are separated into a left side and a right side.

4th and 5th layer (from bottom) are identical.

see my notes from overlapped-message.

• If it was judged acceptable for the card reader to be exposed underneath

the casing, the bottom layer could be removed to reduce the total number of layers to 7.

6 if they're shifted... nehh let's not do that.

• The PCB board sits on the 2nd layer, not having seen any picture of this

board I don't now if components protrude underneath it (I know nothing about electronics). It may not be viable to have it sit on a flat surface that way.

it's not. the top edge (along the top side of the TOP layer) has no components so can be used to hold down the PCB, likewise the bottom edge (along the bottom side of the TOP layer). but it's pretty tight: about 1.5mm as you can see from where you've done an overlap running alongside the DCjack. you can't make it much more than that as you end up running into the 20-pin header and 3.3v and 5v PMIC components. looking at component placement you've actually probably got about 1mm along the bottom edge, might have to make a C-shape (cut-out along most of that bottom edge)

• There are 11 lasercut wood pieces. That's a lot. Is the layered design

definitive

not really. i do like it however. i did consider a hybrid 3d-printed and (thinner) plywood technique, similar to for the laptop.

(I guess that's what the backers will expect since that's what's shown on Crowd Supply), or could this be discarded in favour of a new design optimizing fabrication? A more classic box with dovetail joints may be less expensive, and allow more control on the fitting of the board. Something akin to https://images-na.ssl-images-amazon.com/images/I/91H9TOVowCL._SL1500_.jpg (but hopefully with better looks!)

honest reaction: blegh. you can't do curves, you can't ever hide the dovetail joints no matter how hard you try.

Luke, I know you prefer working with OpenSCAD,

pyopenscad. can't stand the scad "language". i know people love it who use it but for goodness sake writing your *own language* in a specialist (small) field is never going to end well. pyopenscad treats scad as "not-for-reading machine code" which is as it should be.

I was able to export my FreeCAD file to .csg which is a format recognized by OpenSCAD, but the generated text file is over 1MB and contains a series of multmatrix and polyhedrons with an unmanageable number of coordinates. Let me know if you want to have a look at it anyway.

i have freecad installed so can view it - i just can't get my head round editing. too obtuse!

btw are you having difficulties with rendering at all? every 30-60 seconds of continuous rotating i get a "hang" - it's a known bug somewhere in opengl / x11 - which can only be "fixed" by doing ctrl-alt-f1 followed by ctrl-alt-f7.

vincent full link (corrected) http://hands.com/~lkcl/kde_tablet/3dcase

l.

Hello,

Just throwing in my 2 cts, as I know nothing of CAD design...

ha! cool! looks really good. i love the curves on the corners and the edges, if that shape can be retained i think it'll look absolutely stunning.

For the sake of CAD-illiterate people like me, would it be hard or take much time to create some rendered pictures of the thing ?So that we can see with our eyes, and maybe spot a thing or two, or have ideas...

• the PCB will fall out of the casework (both ways, forwards and backwards).
• it will also lift upwards by about... 1mm, rattling around.
• the middle layers also need to be split into separate halves.

right now, the components are in the way.

• the layer below the PCB (with a cutout for the micro-sd card) will

be in direct contact with the PCB, preventing heat dissipation

• also you notice the cut-away in the corner of the PCB underneath

the eject button for the PCMCIA holder? the idea there is to put a small recess (which we have to think how not to interfere with the aesthetics) that will make it easier to eject.

• layers 5 and 6 really should be hollow, so that there's a chance

for air to circulate

... because that is a bit hard to follow if you can't see anything.

also, a decision needs to be made whether to line up layer 2's "high edge" with the top of the PCB or whether to line it up with the *bottom* of the PCB, ass you've done. the way i've tried it, you end up with one less piece of plywood. but the way i've tried it, layer 2 will need to be 4mm thick. i'm still leaning towards that

From a build perspective if the layers can be made with identical

source materials, it would be better, as there will probably be lost cutouts, and some of them may be thrown away. It would go with the eco- consciousness of the project.

one other consideration, should there be ventilation holes drilled in the top and bottom, for air circulation?

I'd hazard that this would be a absolute yes, even in case of no OC'ing, nor tinkering. Always good, no real downsides.

people might wish to overclock these things and there should be room for them to put an ultra-slim fan inside, as well as electronics (access to the 20-pin header).

But you'll have to stop somewhere, not to overdesign, so the tinkerers will be able to modify the thing to suit their needs, as all will be libre.

on the crude technique i'm using, images microdesktop_1.png and microdesktop_2.png are here: http://hands.com/~lkcl/eoma/tablet/3dcase

Ah, that may be the answer to my first question above, but: 404, so I can't really tell.

Thanks

2017-03-30 6:47 GMT-04:00 Vincent Legoll vincent.legoll@gmail.com:

For the sake of CAD-illiterate people like me, would it be hard or take much time to create some rendered pictures of the thing ?So that we can see with our eyes, and maybe spot a thing or two, or have ideas...

Sorry Vincent, since it was my first draft I didn't bother with providing proper documentation... Will try to whip something up in about 12 hours.

Normand

Hello Hendrik, Luke already replied that he always forgets it is "kde_tablet":

http://hands.com/~lkcl/eoma/kde_tablet/3dcase/

Am 30.03.2017 2:26 nachm. schrieb "Hendrik Boom" hendrik@topoi.pooq.com:

On Thu, Mar 30, 2017 at 11:21:36AM +0100, Luke Kenneth Casson Leighton wrote: ... ...

...

on the crude technique i'm using, images microdesktop_1.png and microdesktop_2.png are here: http://hands.com/~lkcl/eoma/tablet/3dcase

It gives me a

Not Found

The requested URL /~lkcl/eoma/tablet/3dcase was not found on this server. Apache/2.4.10 (Debian) Server at hands.com Port 80

Perhaps not everyone has this problem, because the discussion continues as if others have seen it.

-- hendrik

---

arm-netbook mailing list arm-netbook@lists.phcomp.co.uk http://lists.phcomp.co.uk/mailman/listinfo/arm-netbook Send large attachments to arm-netbook@files.phcomp.co.uk

2017-03-30 6:21 GMT-04:00 Luke Kenneth Casson Leighton lkcl@lkcl.net:

• the PCB will fall out of the casework (both ways, forwards and

backwards).

Actually it won't. :-) It cannot slide frontward due to upper layers stopping the DC/VGA ports. As for backwards it's more of an issue, only the right back corner of the EOMA card receptacle (have no idea how it's called) will be stopped by a width of 3mm.

• it will also lift upwards by about... 1mm, rattling around.

Yes that bugs me too.

• the middle layers also need to be split into separate halves.

Yes, that's what I did... Not happy with it either as it increases the number of parts. I went with a case depth approx. equal to the PCB because the Crowd Supply page mentions the microdesktop casing to be 4.5 x 4.5 inches (114mm). But if we can make it deeper and have the separate halves connect, it would be much better.

right now, the components are in the way.

• the layer below the PCB (with a cutout for the micro-sd card) will

be in direct contact with the PCB, preventing heat dissipation

I naively thought that being a low powered CPU, it would not be needed. I did mention I know nothing about electronics. :-P

• also you notice the cut-away in the corner of the PCB underneath

the eject button for the PCMCIA holder? the idea there is to put a small recess (which we have to think how not to interfere with the aesthetics) that will make it easier to eject.

Was wondering about that cut-away. The eject button isn't modelled in the microdesktop_model2.stl file.

I need to go, I have an idea about holding the PCB but I need to think more about it...

Cheers

Normand