00:03 <houkime> ok. wpwrak pomted me a seful utility to measure 3d scans. This should help.

00:04 <houkime> *pointed me to a useful utility

01:06 <houkime> although spread of target position provided by different anchors is rather big to just average blatantly... Gonna sleep and see what can be done about this tomorrow.

01:07 <houkime> bye

05:46 <Joerg-Neo900> houkime: I told you the old connector is the reference. It been placed in proto_v1 for both LOWER and BOB, thus unambiguously defining X and Y postion of BOB in reference to LOWER. The camera position is not even exactly defined in Neo900 since we use a camera module with flex cable

05:49 <Joerg-Neo900> v1 was for evaluating exactly those mechanical things

05:50 <Joerg-Neo900> v1 BOB been adjusted to make SD card fit into case, iirc. Placement of other components like LED, HB connector etc consequently is in reference to the already defined in v1 SD holder

06:04 <houkime> hmm. ok, I guess finally found files you mention. However they are not marked as v1

06:06 <houkime> ok, I know what's going on here

06:06 <houkime> you didn't publish final v1 files

06:06 <houkime> and in public diles in cvt the board is marked as unfinished v2.

06:07 <houkime> however in ee-full cvt contains final v1 files

06:07 <Joerg-Neo900> yeah, sorry for the mess

06:08 <Joerg-Neo900> you have it to the point correct

06:09 <Joerg-Neo900> cvt is indeed the final v1, regarding layout

06:10 <Joerg-Neo900> however noe that v1 had a very simplistic edge cut which we improved then

06:10 <Joerg-Neo900> note*

06:11 <Joerg-Neo900> so position of BOB B2Bconn in v1 is in reference to other fixed part positions mainly

06:12 <Joerg-Neo900> it is not necessarily correct in absolute XY values in reference to origin[0,0]

06:13 <Joerg-Neo900> though I hope we made sure the origins match in v1 and v2

06:16 <houkime> you sure you machanicaly tested exactly that layout? If the connector markings and pins are correct, bob in this layout should stick out through the case.

06:17 <Joerg-Neo900> no, I'm not sure

06:17 <Joerg-Neo900> I try to keep all those details in my head since a few years now, I start to fail

06:19 <Juesto> why

06:19 <Juesto> how old are you

06:19 <Joerg-Neo900> old enough to forget stuff over the years

06:19 <Juesto> have you considered writing everything down?

06:20 <Joerg-Neo900> there's probably too much written down in emails and wiki "tickets" and stuff

06:21 <Joerg-Neo900> http://projects.goldelico.com/p/neo900/issues/ might help

06:25 <Joerg-Neo900> particularly http://projects.goldelico.com/p/neo900/issues/518/

06:34 <Joerg-Neo900> bur yeah, like every open source community project we suck at documentation of work been done

06:39 <houkime> You might need think more about this. You know, more powers towards transparency means more preorders.

06:41 <houkime> It's hard to preorder a thing a work on which has irregular timeline AND is not well-documented.

06:44 <Joerg-Neo900> mhm

06:45 <Joerg-Neo900> please explain "not well documented"

06:47 <Joerg-Neo900> anyway sorry, I'm afk for the next 3 weeks

07:15 <houkime> You have tons of documents like 3d scans and specifications, but your working procedures and timeline (which are exactly the things customer is looking for) are not. From the practical perspective, this issue you pointed at has no info on whether bob was tested or not.

07:17 <houkime> Also I can't know what exactly version of the layout was tested

07:19 <houkime> and whether all issues there were adressed or it is me who needs to adress all of them or some.

07:24 <houkime> usually you create a ticket, and then close tickets with git commits with commit text containing issue number.

07:25 <houkime> and on the issue tracker page if it is not done automatically you mention this commit's sha.

07:35 <Joerg-Neo900> you're talking about sw development, right?

07:38 <houkime> Yes, but it can be done with any development. Eases things. Joerg, you're pioneering OSHW here, it's ok that HW development doesn't have yet all the bonuses of OS employed as a routine.

07:38 <Joerg-Neo900> n oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo, it's not *ME* right now

07:39 <Joerg-Neo900> damn VNC

18:19 <houkime> tried to calculate needed BoB position based on cvt-v1. However even if you flip the bob on cvt (for it to be in a legacy N900 direction and not protruding through the case or going around the camera the wrong way around) and match the connectors while placing bob over the LOWER, it still seems misaligned even on cvt-v1.

18:20 <Joerg-Neo900> then nevermind

18:21 <Joerg-Neo900> maybe use highres scans of N900 PCB and PCBA, and see the fixation holes which *might* help

18:21 <Joerg-Neo900> prolly won't though

18:58 <houkime> i will try to do a multi-mounthole-anchored approach again I guess and do a transfrom from stl to v2.

19:01 <houkime> however hb.pdf uses a little bit modified bob geometry, so the priority here is to mainain clearances rather so everything fit.

19:03 <houkime> What I foresee is that connectors or entire bob should be moved a little for connectors not to land directly at usb port.

19:04 <houkime> it was not a problem with v1 one-connector design but seems to be with 2 of them. Calculating and checking.

22:32 <houkime> created a linear transform from stl to v2 and drew a projection of legacy n900 on the drawings layer.

22:32 <houkime> Looks neat

22:34 <houkime> *legacy n900 BOB

22:38 <houkime> And yes, the nearest to the edge connector lands on usb

22:40 <houkime> (not on the logic, but the usb itself, which is immovable )

22:44 <houkime> cleaning up, commiting and pushing drawing and transform, then thinking how to adjust stuff.

23:20 <houkime> pushed.

23:21 <houkime> calculating needed vs possible bob layout changes.