**** BEGIN LOGGING AT Fri Jan 09 02:59:58 2015
Jan 09 15:02:51 <uhhimhere> <uhhimhere> hi can someone port openmoko to the Samsung GT-S7580
Jan 09 15:02:52 <uhhimhere> <uhhimhere> it has an opensource friendly GPU the Broadcom Videocore4
Jan 09 15:02:52 <uhhimhere> <uhhimhere> its as oss friendly as a SoC gpu can get
Jan 09 19:41:11 <DocScrutinizer05> starseeker: I think Openmoko Inc is more or less dead
Jan 09 19:42:08 <DocScrutinizer05> uhhimhere: I don't think that's gonna happen
Jan 09 19:43:20 <DocScrutinizer05> there's more to porting a OS than just having a "OSS friendly GPU"
Jan 09 19:44:44 <DocScrutinizer05> I guess the device you quoted is a phone? Then you got a huge problem about modem interface and integration into the linux APE environment. You also need to deal with possibly tivoized bootloader
Jan 09 19:45:29 <DocScrutinizer05> s/APE/AP/ since E=environment
Jan 09 19:49:05 <DocScrutinizer05> and generally many recent SoCs that come with modem integrated actually run whatever flavor of "linux" (== Android usually) as a subprocess under a supervisor that shares the one CPU between modem and linux world. Usually such supervisor (think of it as DOM0) is integrated into the modem OS
Jan 09 19:50:52 <DocScrutinizer05> and thus you can't tamper with the supervisor since modem OS is invariably locked vy cryptographic signature measures nowadays, since that seems to be a legal requirement
Jan 09 19:51:21 <DocScrutinizer05> s/ vy/ by/
Jan 09 19:51:21 <apt> DocScrutinizer05 meant: and thus you can't tamper with the supervisor since modem OS is invariably locked by cryptographic signature measures nowadays, since that seems to be a legal requirement
Jan 09 19:52:52 <DocScrutinizer05> the above also usually implies that the bootloader as well is part of such protected OS resp firmware
Jan 09 19:59:09 <starseeker> DocScrutinizer05: OK, thanks - not surprised.
Jan 09 20:03:19 <DocScrutinizer05> starseeker: any particular reason you asked?
Jan 09 20:06:21 <starseeker> CC-BY-SA is problematic for journal papers - I was hoping they could either given an explicit public OK for people to use the CAD models for computer graphics papers or (now that the CAD models are for a phone design that's rather old by phone standards) loosen the license itself
Jan 09 20:07:02 <starseeker> the openmoko CAD files are the most complex open source CAD model created using NURBS geometry that I am aware of
Jan 09 20:07:47 <starseeker> they'd make a great "standard" test case for research work that needs real-world NURBS based geometric models as inputs
Jan 09 20:08:12 <starseeker> but if no one can publish renderings of 'em without having to make the whole paper CC-BY-SA, it's a no-go
Jan 09 20:09:43 <starseeker> (they're even more attractive in that they're not merely "conceptual" CAD models but phones have actually been manufactured for real using them)
Jan 09 20:21:21 <DocScrutinizer05> well, I dunno much about the implications, but I'd dare to grant use of such files in the way you suggested, in my position as ex-OM-employee who been involved in copyright and publishing
Jan 09 20:21:49 <DocScrutinizer05> I'm pretty sure nobody will sue you
Jan 09 20:22:42 <DocScrutinizer05> particularly when you show best intentions by mentioning the origin in some appropriate and feasible way
Jan 09 20:24:01 <DocScrutinizer05> the intention in using CC-BY-SA was to have a licensing as liberal and permissive as possible
Jan 09 20:25:09 <starseeker> DocScrutinizer05: what about the parent company of OM Inc.?  Would it be worth contacting them?  At the very least it would be a chance for them to milk a little more positive PR out of the files
Jan 09 20:25:34 <DocScrutinizer05> I guess simply naming them the "Openmoko files" will already suffice as attribution
Jan 09 20:25:50 <DocScrutinizer05> starseeker: there's no parent company of OM
Jan 09 20:26:05 <starseeker> huh.  Wikipedia got it wrong?
Jan 09 20:26:12 <DocScrutinizer05> you may try to contact Sean MossPultz
Jan 09 20:26:45 <DocScrutinizer05> ((wikipedia)) well, kinda they got that detail wrong, yes
Jan 09 20:27:39 <DocScrutinizer05> initially OM been associated with FIC, but that changed eventually, since the FIC heritage was ... unfortunate
Jan 09 20:27:57 <starseeker> ah.
Jan 09 20:28:21 <starseeker> so the copyrights actually do rest with OpenMoko Inc. itself?
Jan 09 20:28:27 <DocScrutinizer05> yes
Jan 09 20:28:58 <DocScrutinizer05> so the ultimate decision about it would be with Sean
Jan 09 20:29:09 <starseeker> cool - what's the best way to get ahold of him?
Jan 09 20:29:09 <DocScrutinizer05> (OM CEO)
Jan 09 20:29:19 <DocScrutinizer05> sorry, no idea
Jan 09 20:29:27 <starseeker> heh - no problem :-)
Jan 09 20:29:32 <DocScrutinizer05> I think he had a blog and stuff
Jan 09 20:29:39 <starseeker> is he still active in the openmoko community?
Jan 09 20:29:42 <DocScrutinizer05> maybe just google
Jan 09 20:29:48 <DocScrutinizer05> hardly
Jan 09 20:30:22 <DocScrutinizer05> I dunno if he occasionally still shows up in #qi-hardware
Jan 09 20:30:43 <DocScrutinizer05> rumor has it he's peeking, usually "anonymous"
Jan 09 20:30:56 <DocScrutinizer05> but you're prolly better off googling for him
Jan 09 20:31:40 <DocScrutinizer05> https://www.google.de/search?q=Sean+Moss+Pultz
Jan 09 20:31:43 * starseeker will take a shot at it - thanks for the pointer.  I know it's a non-sequitor as far as the openmoko community proper is concerned, but you'd be amazed how hard it is to find good open source CAD models
Jan 09 20:35:34 <DocScrutinizer05> my advice: try to reach Sean. If no luck, just don't worry and go ahead
Jan 09 20:42:06 <starseeker> I'll see if I can reach him.  Journals are sticklers for making sure they have the right to publish stuff, so I need to do things "the right way"
Jan 09 20:49:24 <DocScrutinizer05> if you can't reach Sean, contact me again. I think when Sean isn't available, I might speak officially for Openmoko
Jan 09 20:49:50 <starseeker> DocScrutinizer05: awesome - thank you!  is the channel the best place?
Jan 09 20:50:16 <DocScrutinizer05> yes, just highlight me. Or send a mail to joerg at openmoko org
Jan 09 20:52:08 <starseeker> regardless of the CC-BY-SA questions, I want to thank you guys for making the files available - I'm with the BRL-CAD open source computer aided design project, and these files have been extremely useful test cases as we've been developing our importers for STEP and IGES
Jan 09 21:06:50 <DocScrutinizer05> hey, would you maybe interested in "giving back" by helping with 3D design of an object for Neo900?
Jan 09 21:07:22 <starseeker> DocScrutinizer05: possibly - what did you have in mind?
Jan 09 21:07:47 <starseeker> I'm more of a programmer than a 3D designer, fair warning...
Jan 09 21:08:20 <DocScrutinizer05> aah, well
Jan 09 21:08:32 <DocScrutinizer05> anyway, here's what it's about: http://neo900.org/stuff/joerg/spacerframe/sketch.pdf
Jan 09 21:09:23 <DocScrutinizer05> http://neo900.org/stuff/werner/scans/
Jan 09 21:09:48 <starseeker> so a 3D model of that case, or a modification of an existing 3D model along those lines?
Jan 09 21:10:57 <starseeker> ah - you've got STL scans, but not actual CAD models?
Jan 09 21:11:02 <DocScrutinizer05> we need a 3D model of that new spacerframe, based on the existing scans of case in which it needs to fit, and the old smaller kbdframe which gets replaced by the new spacerframe
Jan 09 21:11:11 <DocScrutinizer05> yes, exactly
Jan 09 21:12:00 <starseeker> what format(s) would you need?  STL, STEP, ...?
Jan 09 21:13:05 <DocScrutinizer05> sorry, no idea at all, I'm not into CAD
Jan 09 21:13:28 <DocScrutinizer05> sth that allows for 3D printing and later on maybe CNC milling or whatever
Jan 09 21:14:29 <DocScrutinizer05> if you're actually interested resp willing to help, I suggest /join #neo900
Jan 09 21:15:33 <starseeker> DocScrutinizer05: I'm trying to think what the best approach would be.  I'll take a closer look at the existing data sets
Jan 09 21:17:45 <starseeker> DocScrutinizer05: if we don't need to have a NURBS version of the model, that's one thing.  If we do, then basically someone needs to take those scans and create a parametric model
Jan 09 21:18:13 <starseeker> how much accuracy/precision is necessary?  i.e. what are the necessary tolerances?
Jan 09 21:21:16 <starseeker> DocScrutinizer05: I'll do a bit of research and get back to you - if it looks like it's something not within my modeling skill set, I can pass the word among the open source CAD communities
Jan 09 21:21:48 <starseeker> openscad in particular has a lot of following in the 3D printing community so we might be able to scare up an actual modeler there
Jan 09 21:22:57 <DocScrutinizer05> sounds fine :-)
Jan 09 21:23:39 <DocScrutinizer05> about precision: needs to fit on the case, so 0.1mm precision is what we need for sure, maybe better if possible
Jan 09 21:24:01 <DocScrutinizer05> but we can fine tune stuff that doesn't fit
Jan 09 21:24:35 * starseeker nods - the most direct approach would probably to be to try working with the mesh directly, but I don't know how practical that is
Jan 09 21:25:11 <DocScrutinizer05> please /join #neo900
Jan 09 21:25:40 <DocScrutinizer05> there are some guys who know more about this particular CAD issue than I do
Jan 09 21:27:24 <starseeker> done.  I've got to run, but I'll be back later on neo900.  I'll also try an email to Sean.  Thanks again for taking the time to respond!
Jan 09 21:28:54 <DocScrutinizer05> yw :-)
**** ENDING LOGGING AT Sat Jan 10 02:59:59 2015