libera/#neo900/ Friday, 2018-08-17

ravelohoukime, Joerg-Neo900: to discuss these topics is imho very Important. I might join tomorrow as i am on vacation right now00:18
houkime1) The goal is clearly indicated in the heading on my file "improve contributors workflow". I can say it also in an another way: to minimize delays and effort needed for contributor to successfully contribute. I think it is clear enough. I think the delays that are stemming out of a closed model are very significant. They consist of 1) waiting for you to show up in this chat and respond 2) writing and sending a signed email to you with nda 3) waiting00:57
houkimefor atk to jump in 4) waiting for atk to create an account and manage rights 5) generating ssh keypair for a server 6) waiting for atk to be there and to receive and register your open key 7) cloning and making your branch 8) waiting for atk to be there to make you a remote branch to push to and manage your write access to it. 9) making changes and pushing.00:57
houkimeformalities alone can take several days, yes, and in fact oksana still doesn't have even read rights even given she was talking about it yesterday and already gone as far as to install kicad.00:59
houkimemetacollin didn't have so much delay with eeshow because eeshow is completely open and takes literally cloning, changing and sending out a patch01:00
houkimecomplex scripts are a problem because atk for example had to work several days to support open git logs for non-master branches whilst that wouldn't be an issue if not for a system of repo filtration.01:03
houkimebecause just any files and logs would be open by default01:04
houkime*would have been01:04
Joerg-Neo900houkime: I also got a goal: stop this discussion.01:06
Joerg-Neo900you either tell me what particular procedure you have trouble with, and I'm pretty confident we already have a solution in place for it that I can help you to understand and use so you achieve to complete the task at your hands. Or I will ignore any further request asking for abstract goals like "streamlining contribution" etc01:15
Joerg-Neo900you never indicate any real task you want to achive (aka goal), you always talk about properties of the project (ease of contribution) you want to change by modifying the tools used (git etc) but frequently you fail to solve a task by all that (like "multiple users can concurrently edit the FOOBAR file")01:20
Joerg-Neo900>>complex scripts are a problem because atk for example had to work several days to support open git logs for non-master branches whilst that wouldn't be an issue if not for a system of repo filtration.<<  sorry to put it that plain: you got no idea what you're talking about01:23
Joerg-Neo900the changes you ask for in project, I suggest you better implement in your own project. You're welcome to clone Neo900 and start it your own way01:24
houkimeso you're opensourcing it and clearing ndas?01:25
houkimebecause otherwise I can't really clone01:25
houkimepublicly at least01:26
Joerg-Neo900sorry, that's *again* nonsense and shows you have a completely wrong idea about so many things01:27
houkime*clone-> fork01:28
wpwraki think it would be hilarious if a copycat actually did try to steal the design ;-) i mean, first, they're likely to work with very different numbers. second, in order to copycat the design, they need n900 parts. now, who owns all those that could still be found in the market ? and will joerg really want to sell them to the copycat ? ;-)01:28
Joerg-Neo900you can't clone the footprints and layout you inherited from metacollin. And I think rhis statement already shows me that you failed to even understand that and I need to rethink the access to project I granted you based on the NDA you sent01:29
Joerg-Neo900everything else is under CC and you're free to start your own NON-COMMERCIAL project from it01:29
Joerg-Neo900and U'd be very happy if you rather do that than continue to pester me about changes in toolchain01:30
Joerg-Neo900obviously you're not using any of the existing toolchain anyway, so why not start your own project right away? AGAON: NOTHING but the layout stuff metacolling and you did is under any "NDA"01:33
Joerg-Neo900so if you hate that much the tools wpwrak and I got set up and provided for contributors, why do you try to talk us unto changing them?01:34
houkimeno, I just thought for a second that you want to get rid of me so much that you release layout. Probably not though. Sad because this would be 1) from my complaint list auto-completed.01:37
Joerg-Neo900wpwrak: the project is basically dead since sourcing of further N900 parts is getting impossible but to do production we need 200 or 300 more orders for which we can't source the N900. A copycat OTOH could produce a 5ß boards and offer them to those users who already have a N900, "for cheap". I don't know if I care one way or the other01:37
Joerg-Neo900houkime: if there's any layout not done by you that you need me to release so you could continue with the project on your own, then I think you could NOT continue with the project on your own, due to lack of skills. We got no layout to publish yet01:39
houkimeI can reiterate everything myself to get rid of nda-protected stuff of metacollin and my own, but it will take time and i'm a bit lazy to redo.01:42
Joerg-Neo900yes, more dilligentce in productive work and less bitching about tools would help a lot01:45
Joerg-Neo900I just can say mrtacollin got his stuff sorted and didn't ask for changes in toolchain, rtaher contributed to fixes in the existing one, within 2 days. This is not a pony park for devels to drop by, pick an interesting little weekend duty and then walk on. I had a basically two time 18h nonstop chat with metacollin where I helped him with sharing hints like pointers to documents and whitepapers, setting up his git account and whatnot. With you01:49
Joerg-Neo900I had a month of discussion if gitolite was a good or a bad tool for this purpose01:49
Joerg-Neo900and after that month (and them several more) you still have a >>Instead people will just have their own forks and no write access to the main repo is needed. forks a reviewed upon merge but that's it.<< in your list of issues. Which is exactly how we manage things on our server01:54
houkimejoerg-neo900: sorry but you're wrong in this last one. Repo is the project folder altogether with all branches.01:56
houkimeFork is a similar WHOLE repo copied from the first repo at some point.01:57
Joerg-Neo900listen, I'm fed up with this01:58
houkimeright now you're managing your repo in a way that different branches are for different people01:58
houkimeI'm proposing a different thing which is much easier to do and which is usually done01:58
Joerg-Neo900get over it, what we have JUST WORKS and WONT GET CHANGED because "something new is simpler". Nothing is simpler than a working system01:59
Joerg-Neo900if you want simpler new stuff, do your own02:00
houkimeIt introduces delays and oksana is currently stuck in one.02:00
Joerg-Neo900oksana is stuck in nothing02:00
Joerg-Neo900except I missed any request by her inb the noise produced by you02:01
Joerg-Neo900he might be stuck with false info you provided, regarding the repo02:01
Joerg-Neo900I can't see any other stuff she's stuck with02:02
Joerg-Neo900if she wants to review your layout done inder the NDA, then she and you are stuck with me giving an OK to you to share the layout to her. Everything else there IS NOTHING anybody could get stuck02:04
Joerg-Neo900sprry I have a hard time considering all this presented by you as productive cooperation02:06
Joerg-Neo900actually I think it's detrimental to the minimal rest of life in the project, by trampling over the achievements made so far02:07
Oksanahoukime : 1) Don't turn me into example, I haven't even installed VirtualBox yet, and it's perfectly fine by me to wait a month for access even when I get around to opening to kicad - I have many other things to do, and I have better idea about python than about kicad (and I don't even like python);02:15
Oksana2) while it would be fun for contributors to be able to jump in easily, it could wreak havoc onto large and complicated project of Neo900 - SparkFun is for tiny things;02:15
Oksanaseveral days is not too long;02:15
OksanaAnd it's interesting to see that apparently board/layout is better done by one person, and schematics by another - but then, I don't know enough about the difference between board and schematics.02:17
houkimewith neo900 it's just a bit too big to be hadled alone from end to end esp if you're not a contracter.02:19
houkimeat least in reasonable time02:20
houkimeAnd because it is a conversion phone lots of extra research is needed and considerations.02:22
houkimethat's why i think more people involved is better.02:24
houkimeesp since stuff gets obsolete on the fly02:24
houkimeand that's why i want to ease contributions.02:24
OksanaIdeally, important [which cannot be replaced by a newer version] stuff gets bought before it gets obsolete (and then "obsolete" doesn't matter).02:26
houkimeIdeally yes. Not the case here though.(02:28
OksanaDon't whine. I can easily give example of a phone-making company which hasn't given any updates for the last four months... Despite claiming that they are all ready, will ship soon [just some more checks to do] And they had much more publicity in the news than Neo900.02:31
houkimeWas it an opensource company?02:33
OksanaNay. But at least, they promised multi-boot aka "you can boot anything you wish".02:34
houkimemy point was that we need more devspeed that I can possibly provide in order not to be late to the shinking N900s market.02:38
houkimeAnd the devspeed for opensource is contributors and their ease of doing stuff.02:39
houkimeright now we don't have the money to reserve all n900s project needs if i understand correctly02:40
houkimealso clientbase is getting smaller no matter of reserved n900s.02:41
Oksanahoukime : Far from shrinking, I would hope. But then, it's difficult to figure out proportion of people looking for something like Neo900, without doing aggressive marketing.02:42
OksanaAnd for aggressive marketing to work, it would be nice to have some good news. Such as, "layout good, making the final prototype" or something.02:43
OksanaAnd large number of contributors can make layout-ing difficult because of mis-communication.02:44
OksanaAnd I have installed geda, besides Kicad. Once I install VirtualBox, will try to take a look at with three different programs, I guess (KiCad4, KiCad5, and gEDA).02:48
* Oksana is looking at , and wonders how it meshes with VM - eeshow already installed, I guess, but initialization probably still needs to be done before viewing the schematics02:49
houkimeI don't really believe colayouting can't be organized ok, also I don't think we will achieve real large numbers on this one, at least soon.02:55
Joerg-Neo900you don't need the VM, kicad4 should be available genuinely for your system02:55
* Oksana nods - which version of kicad is in Debian Stretch?..02:58
* Oksana should switch to Devuan repositories, but with the way I am going, I gain experience with setting up the system by going through unnecessary obstacles. Such as, installing jessie, and then upgrading to stretch (instead of installing stretch from the start)03:03
OksanaYes, KiCAD 4.0.5 in Stretch/aka/Stable03:03
houkimeOksana: do you have a public key on a keyserver so I can send you layout stuff encrypted with it?03:30
* Oksana has no idea what a keyserver is, and hasn't ever got around to making a public key03:41
houkimeIf atk were around he could give you read access and then it could be a standard ssh procedure.03:47
houkimeI'm not sure I got his talent of guiding people through stuff.03:50
houkimerecovered my conversation with atk on this. It should work like this. You install gnu-pg, then run $ gpg --full-generate-key to generate a new gpg keypair04:05
houkimethen you want to export public key to an external keyserver $ gpg --keyserver --send-keys <your public key footprint here>04:06
Oksanahoukime : I am not at Linux computer right now, have got to wait till 20h later or so.04:07
houkimeok (although this procedure is not really linux-bound)04:09
houkimeI need to sleep04:09
ac_laptophello people _o/14:42
ac_laptophow is the project going ?14:42
houkimeyou can check the current state of the pcb here
houkimealthough probably it's not the best video ever produced. Kind of new to the genre, sorry.14:47
ac_laptop < Interesting!14:47

Generated by 2.17.0 by Marius Gedminas - find it at!