infobot has quit [Ping timeout: 248 seconds]
infobot has joined #neo900
_whitelogger has joined #neo900
wpwrak has quit [Read error: Connection reset by peer]
wpwrak has joined #neo900
Pali has quit [Remote host closed the connection]
Kabouik has quit [Ping timeout: 276 seconds]
kungfumike has joined #neo900
<sn0wmonster> DocScrutinizer05, when dealing with FOSS projects and working for free, it's a bit disrespectful to those who contribute to not provide the tools necessary to do their work, nor any sense of streamline, that the work will actually be used
<sn0wmonster> i for one wouldn't type a single letter until i know it won't just be sitting in a txt file on some server waiting 6 months for someone to get around to using it.
<sn0wmonster> my time is too valuable for that.
pagurus has joined #neo900
<sn0wmonster> and judging by the result of my inquiries, no one is available in #neo900 project to actually post the news even if it's being written. i recommend that problem be solved first, even if a new solution need be devised.
pagurus` has quit [Ping timeout: 240 seconds]
<sn0wmonster> its like asking a cook to come to your house and make something for you, without even being sure if you can unlock the house or not.
<sn0wmonster> its just not how things are done when respect and trust are the fuel for progress
<Joerg-Neo900> you're making up problems, while my time is too scarce to deal with this type of unfounded concerns. What makes you think there would be a problem posting someting on a webserver?
<sn0wmonster> Joerg-Neo900, because everyone you've asked me to contact a month ago who supposedly has access to the site has not responded to a single inquiry?
<sn0wmonster> i don't call that making up problems, i call that recognizing an issue.
<Joerg-Neo900> the issue is I'm the only one dealing with this and you ask me to add more explaining and management of basically non-issues to my full plate
<sn0wmonster> Joerg-Neo900, i haven't asked anything, i was responding to DocScrutinizer05's claim that there is no issue in posting news. the issue is in fact that there is no way to post news.
<Joerg-Neo900> 95% of my time are already eaten by this type of managing side fringe issues that shouldn't even be there
<sn0wmonster> Joerg-Neo900, i agree. you need a streamlined process for it.
<sn0wmonster> there is an obvious breakdown if the people who are intended to handle such things on your behalf haven't responded to a message in a month.
<sn0wmonster> i'm looking for a solution, not complaining. i don't mind helping build said solution, so long as it is recognized that it's a problem.
<sn0wmonster> someone writes the updates and news -- okay now what? who do they contact? you? same problem. you're doing work you shouldn't have to be doing.
<Joerg-Neo900> there's no problem, except for the time it takes to "build said solution" zo a non-existing problem
<sn0wmonster> perhaps i'll try asking this question from the beginning. who is in charge of adding content to the site who has the access necessary? that's the person that will be needed to work with to get news and updates to the site.
<Joerg-Neo900> there's a git repo that has *all* the source texts of the website, I can find somebody to check in new versions to that git repo, and that's the whole magic
<sn0wmonster> so the news/updates can be pushed directly to the repo?
<Joerg-Neo900> sure
<sn0wmonster> who would be doing that?
<sn0wmonster> (since i don't have commit access, who would be the one approving/sorting the repo pull request for news/updates?)
<Joerg-Neo900> why do you want to know?
<sn0wmonster> because that's the person i'd need to have contact with
<Joerg-Neo900> no, you need contact to me
<sn0wmonster> great
<Joerg-Neo900> since without my consent nothing gets pushed there
kungfumike has quit [Quit: Leaving.]
<Joerg-Neo900> there are around 4 people who have push permissions to the git repo, and it's not a problem to add a 5th person. I think I linked to the readme in said git repo before, that has some deatils. I mentioned Pelican markup which is used. I explained that the two persons who used to do the newsletters and website management are not available on a regular basis but most likely will help out when we have any particular requests. I don't know what else
<Joerg-Neo900> to provide to demonstrate my respect for your work
<Joerg-Neo900> if you got questions and they are not answered by dos1 and/or how900, please ask me instead, and I either will know to answer them or will forward them to the aforementioned persons via email, since they are not following IRC very closely
<sn0wmonster> Joerg-Neo900, thank you for the detailed response. i have no issue with you personally, my comments were an attempt to explain to DocScrutinizer05 how FOSS projects tend to operate. you've made it clear that you are the go-to so that answers all the questions. :)
<Joerg-Neo900> DocScrutinizer05 == me
<sn0wmonster> okay. still, question answered, crisis averted. i appreciate the attention to it.
knttl has quit [Ping timeout: 248 seconds]
knttl has joined #neo900
<sn0wmonster> Joerg-Neo900, okay to PM?
<Joerg-Neo900> sure
kungfumike has joined #neo900
kungfumike has quit [Quit: Leaving.]
ArturShaik has joined #neo900
_whitelogger has joined #neo900
louisdk has joined #neo900
Pali has joined #neo900
jonsger has joined #neo900
cc___ has joined #neo900
_whitelogger has joined #neo900
_whitelogger has joined #neo900
_whitelogger has joined #neo900
_whitelogger has joined #neo900
_whitelogger has joined #neo900
jonsger has quit [Ping timeout: 260 seconds]
jonsger has joined #neo900
<ravelo> looks like posting news is possible soon...
err0r3o3_ has joined #neo900
<sn0wmonster> writing at atm
<sn0wmonster> it*
<sn0wmonster> waiting on Joerg to get back to answer some questions
err0r3o3 has quit [Ping timeout: 268 seconds]
Kabouik has joined #neo900
<Joerg-Neo900> just ask please. Most questions can get answered by a lot of users in this channel as well
<sn0wmonster> can you describe the current position in the production timeline?
<sn0wmonster> I gather it's still in the "Research and Development" phase, that there's a final board being designed and almost ready to print,
<sn0wmonster> and that will take us to the Purchases section of the timeline where people can properly pre-order their Neo900, to which it will then move to mass production?
<sn0wmonster> please correct otherwise
<Joerg-Neo900> currently layout of proto_v2, which is Neo900 with 'external brain' on BeagleBoard_xM (identical CPU). This is to take complexity and thus possibility of failure out of the OMAP integration, while the proto_v2 will provide a full featured testbed and development board for our community software developers like e.g. freemangordon, Pali and parazyd. We plan to build a maybe one dozen of those, then test them, then start kickstarter to reach the
<Joerg-Neo900> needed production numbers and build prot_v3 which - with luck of no bugs in it found - is identical to the mass production version
chainsawbike has quit [Ping timeout: 240 seconds]
chainsawbike has joined #neo900
<Joerg-Neo900> 2017-01-29 Neo900 v2 Prototype (UPDATED)
<Joerg-Neo900> start of this week I finally contracted Esa as layouter for proto_v2, after ages of searching for, and negotiations and discussions with several layouters - they are damn hard to find
<Joerg-Neo900> Esa of LNAFIN been found and contacted by ravelo, and he's sympathizing with the Neo900 project and gave us a good deal since thanks to the Open Source character of this project his company can make use for their own advertising (reference project) from the work they do for us
<Joerg-Neo900> they support KiCad! Pretty hard to find a company that does
<Joerg-Neo900> another cost reduction factor been "relaxed time schedule" meaning there's no hard deadline for when the task gets accomplished
<Joerg-Neo900> a huge problem was the tool fped which is needed to generate the kicad foorprints we provide for the layout to Esa (many thanks to ceene for doing the footprints), and there's no up-to-date version of fped in _any_ distro since fped has no release versions that distro maintainers could pick up. Esa, like any commercial layouter, is not willing to build fped from source
<Joerg-Neo900> I tried to make an up-to-date fped package on OBS, but I failed - I'm no packager
<sn0wmonster> what is "Esa" in "Esa of LNAFIN"?
<Joerg-Neo900> the first name of the guy
<sn0wmonster> ah
<Joerg-Neo900> see the pdf link
<Joerg-Neo900> CEO of LNAFIN
<sn0wmonster> What was the solution to the fped problem? is it still holding the project back or was it eventually solved?
<sn0wmonster> also, how are funds holding up for the project? has it spent what it said it was going to spend? less? more? is it in need of more funds to continue on or is it okay until prototype v3 is ready to begin?
<Joerg-Neo900> the (interim) solution is me sending Esa tarballs instead of him using git like everybody else in the project
<Joerg-Neo900> the funds are sufficient to build proto_v2. We need to generate more orders after that. Then N900 refurbished are basically sounced for everybody who preordered a full Neo900 device (with case, display etc)
<Joerg-Neo900> I fired myself as paid CEO from my own company, to cut down on recurring expenses
<sn0wmonster> and was that the same as was predicted at first receipt of funds? e.g. do pre-order/donators understand that there'd need to be a kickstarter at prototype v3 or will that be a surprise to them?
<Joerg-Neo900> there been always the need for more preorders than we have up to now, and that been announced. so no surprise
<Joerg-Neo900> admittedly the business plan was to get to proto_v3 finished with the preorders we have so far
<sn0wmonster> to clarify: there was a funding round at one point where supporters could purchase a refurbished Neo900, and those funds were used and the sufficient amount of stock of Neo900's is already secured?
* sn0wmonster doesn't want to get the details wrong
<Joerg-Neo900> basically yes. We have a webshop (was a pita to set up and maintain) where customers can do preorders with a lesst than 50% down payment upfront. All that have ordered a Neo900 complete device there have a N900 sourced by us for them
<Joerg-Neo900> a less than*
<sn0wmonster> for those who ordered a complete unit, what is it they need to wait on?
<Joerg-Neo900> sorry?
<sn0wmonster> does the PCB need to be finished for them to get their unit as well?
<sn0wmonster> or was their unit supposed to be modified directly using existing hardware only?
<Joerg-Neo900> I guess you're assuming something I can't tell what exactly
<sn0wmonster> so you made a distinction that confused me given my limited knowledge, i just need to clarify:
<sn0wmonster> "We need to generate more orders after that. Then N900 refurbished are basically sounced for everybody who preordered a full Neo900 device (with case, display etc)"
<Joerg-Neo900> the project offers to design and manufacture a NeoN board (bare PCB sandwich) that users can retrofit into their N900. Users can also order a complete device Neo900 in which case we get a N900 for them and do the hardware upgrade
<sn0wmonster> okay. and the N900's necessary for the existined preorders are already purhcased and in your possession at the moment?
<Joerg-Neo900> ooh those two sentences have no obvious relation
<sn0wmonster> oh okay
<Joerg-Neo900> yes
<sn0wmonster> great, so the N900's are there, but the PCB still needs to be done to deliver it
<sn0wmonster> got it
<Joerg-Neo900> yes
<Joerg-Neo900> and to reach abreak even point on the promised sales price, we need more preorders
<Joerg-Neo900> thus the kickstarter we're planning for
<sn0wmonster> take a look at the link and let me know what you think
<cc___> a crowdfunding campaign again ? o_O
<sn0wmonster> ^ this is what i was concerned about. it doesn't seem clear on the site from what i've read that there would be multiple stages of crowdfunding.
<sn0wmonster> but if you frame it as the "purchasing" period, then it makes total sense
<sn0wmonster> so it's not a kickstarter so much as it is a proper pre-sale.
<sn0wmonster> (with a specific goal of course, so yea, a kickstarter :D)
<Joerg-Neo900> yes, basically
<sn0wmonster> great. could you elaborate on the hold-ups of the project so far?
<Joerg-Neo900> I see you're relatively new to the project, so let me elaborate on what happened (and caused delays): first the plan was that Nikolaus of GolDeliCo does the webshop, etc, and particularly the layout, using Eagle EDA CAD. Eagle is up to that task only with Nikolaus' own router extension plugins written for MacOS only afaik. When PayPal stopped us by freezing our account, Nikolaus postponed work for Neo900 and did Dragonbox Pyra which left no
<Joerg-Neo900> timeslots for us and took quite a bit longer than expected. Then finally Nikolaus told us he's not available for the layout task anymore, so we had to consider alternatives. The alternative (due to Eagle's shortcomings) was to migrate the whole project to KiCad, which we accomplished successfully though it took quite a while again. Werner developed a very nice (actually already legendary) tool called eeshow to facilitate cooperative concurrent
<Joerg-Neo900> schematics design under kicad. Then we re-did a lot of footprints for layout since basically we didn't inherit them from Nikolaus, and finally we had to find a new layouter who is willing to use KiCad to do the layout for us
<Joerg-Neo900> ~eeshow
<Joerg-Neo900> migrating a project of that size from one toolchain to another is like... let me compare it to migrating Apache from c++ to c#
<sn0wmonster> yea, i went through that with #Taskhive when we started in Java for Android and ended up switching to Python for desktop
<Joerg-Neo900> but now we have 100% FOSS EDA toolchain
<sn0wmonster> and that was just for a simple application. i can't imagine what you went through
<Joerg-Neo900> alas FOSS EDA isn't very commonly used in commercial layout
<Joerg-Neo900> so finding a layouter wasn't as easy a task either
<Joerg-Neo900> I'm very happy ravelo found LNAFIN
<sn0wmonster> its like the eternal struggle of designers.
<Joerg-Neo900> the fped issue is only a sidenote
<Joerg-Neo900> though an annoying one
<sn0wmonster> everyone in the industry uses Adobe on Windows/Mac, and even if you are a master of GIMP, krita and inkscape, you won't find any company that can work with you, so it's not just a software issue. it's a social issue.
<Joerg-Neo900> yep
<sn0wmonster> Joerg-Neo900, i see now. thank you for clarifying, i am interested in learning more and will obviously need to if i want to write anything worth reading :)
<Joerg-Neo900> :-D
<Joerg-Neo900> many thanks for asking :-)
<Joerg-Neo900> it's not exactly simple to communicate the whole story
<sn0wmonster> of course
<sn0wmonster> its not necessary to anyway because the people just want to know "where's my f***ing Neo900" usually
<sn0wmonster> :)
<sn0wmonster> need clarification on this statement:
<sn0wmonster> "Nikolaus postponed work for Neo900 and did Dragonbox Pyra which left no timeslots for us and took quite a bit longer than expected"
<ceene> Joerg-Neo900: i've compiled fped for debian several times
<ceene> if you need some help on that, let me know
<sn0wmonster> what took longer than expected? for him to finish Dragonbox?
<Joerg-Neo900> ceene: I'd need a public repo providing the binary, or at very least a binary for Esa's system - alas he doesn't even disclose which system he's using
<Joerg-Neo900> sn0wmonster: yes
<sn0wmonster> ty
<Joerg-Neo900> I'm not quite sure he alrady finshed that job
<ceene> there's been no update to fped since May
<Joerg-Neo900> anyway he suffered maybe a burnout or whatever, not taking any further layout jobs since
<Joerg-Neo900> ceene: sure, just all distros have a fped binary several years old
jonsger has quit [Ping timeout: 248 seconds]
<Joerg-Neo900> Esa installed such obsolete fped binary and the make pull job failed miserably
<ceene> make works fine on my system
<ceene> i've just built latest fped
<Joerg-Neo900> our git doesn't hold the kicad modules output of fped, so... :-/
<ceene> that's right, it holds only fped source files
<Joerg-Neo900> I'd love to add the kicad modules to our internal git
<ceene> i usually don't like having output files onto git, only source files
<ceene> but if that's going to make the work faster, maybe it's better for you or me, or whomever, to include fped output files into kicad repo
<Joerg-Neo900> sure, that's the usual objection to this plan, however it would solve the fped issue
jonsger has joined #neo900
<ceene> in fact, maybe Esa needs to modify some footprint and he may want to use only kicad editor, if he is proficient with it
<Joerg-Neo900> exactly
<Joerg-Neo900> he for sure won't use fped
<ceene> given that fped output files aren't binary files, but are also source files for another footprint editor...
<Joerg-Neo900> yep
<ceene> i guess there can even be a few things that aren't doable with fped and must be modified directly on kicad
<Joerg-Neo900> possibly
<Joerg-Neo900> now add on top my expertise with hit which is that of a bucket full of sand
<Joerg-Neo900> git*
<Joerg-Neo900> I *guess* it's just a few simple edits of a few .gitignore that is needed...
<Joerg-Neo900> **PLUS** copying all fped footprints into a directory that's not in the git submodule(?), or whatever the right term
<sn0wmonster> Joerg-Neo900, recheck
<sn0wmonster> specifically "What held us back so far?"
<Joerg-Neo900> when I ran into that submodule thing, I gave up
<Joerg-Neo900> sn0wmonster: since when?
<Joerg-Neo900> and for how long?
<sn0wmonster> since last update mostly
<sn0wmonster> but i was asking you to check the content of the pad for clarity
<Joerg-Neo900> ok
<sn0wmonster> specifically the "What held us back so far?" section
<Joerg-Neo900> takes an hour though, since I'm just receiving an interrupt request from RL
<sn0wmonster> i have no idea what that means :D
<Joerg-Neo900> real live calling
<Joerg-Neo900> I'll be back in maybe an hour
<Joerg-Neo900> ceene: could you please ponder what to do about the fped thing and git?
<enyc> w
<ceene> there are two distinct directories for fped source files
<ceene> those directly on our repo, under hw/modules, and those on kicad-libs/modules
<ceene> the idea is being able to provide Esa with .mod output files, and also letting him commit .mod files if he chooses to modify them
<ceene> i talked to ahycka about this before, and she thought of fped not exactly as a footprint builder itself, but more like altium's wizard footprint creator
<ceene> because in fact, it's basically that: you have a .fped files, but most of them aren't defining a single footprint, but a variety of them using a specific template
<Joerg-Neo900> :nod:
<ceene> once she has used this wizard to create a footprint, it's sometimes needed to be fixed by hand (maybe pin numbering or ordering, small differences not accounted for on the wizard, etc)
chainsawbike has quit [Ping timeout: 248 seconds]
cc___ has quit [Ping timeout: 240 seconds]
<ceene> so, we could think of fped exactly like this: not a footprint creator, but a wizard for footprint creation, and those .fped files i did, aren't footprints but wizards to draw some type of footprints
<Joerg-Neo900> makes absolute sense
<ceene> so, even if I usually am against including output files on git repos, i think we should just go for it, and include .mod files on the repo
<Joerg-Neo900> my thinking
<ceene> in this case, fped and .fped files aren't themselves files used by kicad, but tools used to create files that *are* kicad files
<ceene> so, what about we just add .mod files to the repo, (both the output of hw/modules and kicad-libs/modules) to a directory under our repo
<ceene> we can provide initial .mod files generated by fped
<ceene> and let Esa modify them as needed
<ceene> if he thinks fped source files can be faster modified to suit his needs, he can just us know and we'll do the quick fixing for him
<ceene> so as to keep this .fped files as useful as possible
<Joerg-Neo900> yes, please evaluate that. what's needed to implement it for internal git?
<ceene> if it's some small tinkering on one .mod file that doesn't apply to the global template, then let him modify this .mod and commit it directly
chainsawbike has joined #neo900
<ceene> currently, .mod files are provided to kicad by just one file: fp-lib-table, which contains the path to .mod output files
<ceene> so my proposal would be
<ceene> a) create new directory "footprints" on our repo, including there all .mod files we have (hw/modules/*.mod, kicad-libs/modules/*.mod)
<Joerg-Neo900> great!
<ceene> b) Modify fp-lib-table so it links to mod files under that directory, instead of under the original path of .fped files
<Joerg-Neo900> s/we have/we use in layout/
<ceene> that's right
<ceene> there's no need to populate that directory with all footprints there exist under kicad-libs
<Joerg-Neo900> :nod:
<ceene> c) We should annotate on some file, maybe the same FOOTPRINTS.txt we have right now, which is the source of each footprint we use. Is it a kicad drawn footprint? Is it direct output of a .fped file (and which file)? Is it the output of a .fped file, but was hand modified later on? If show, note which file and which were the modifications, and maybe a little note on why this couldn't or shouldn't be done on
<ceene> the original fped file
<ceene> this is important so no one overwrites a .mod file which has been manually tinkered with
<Joerg-Neo900> yes, absolutely
<Joerg-Neo900> can you take care of that?
<ceene> sure
<ceene> neo900-legacy.mod contains footprints imported from eagle, is that right?
<Joerg-Neo900> yep
<Joerg-Neo900> aiui
Venemo has joined #neo900
<Joerg-Neo900> actually **As I Understand It** *-legacy.mod is where kicad *creates* footprints that are in *.pcb layout but not in any lib, so kicad on startup of pcbnew opens a requester if you want to save those embedded footprints and if you answer yes then it creates *-legacy.mod. Note please *AIUI*!
<Joerg-Neo900> we didn't explicitly import any footprints from eagle, but the *.pcb is kicad compatible, so...
<Joerg-Neo900> again IIRC and AIUI
<Joerg-Neo900> wpwrak: ^^^ nay comments?
<Joerg-Neo900> any*
<ceene> this would be commit summary
<ceene> and this would be the begining of FOOTPRINTS.txt file
<Joerg-Neo900> sounds excellent
<ceene> i'll push it then
<Joerg-Neo900> you tested it?
<ceene> yep
<Joerg-Neo900> :-)
<ceene> i think i haven't broken anything
<Joerg-Neo900> go ahead
<ceene> kicad still opens the project, pcbnew still opens the pcb and everything seems in order
<ceene> there it is, i've done something useful today :)
<Joerg-Neo900> what needs to get done is changing pcbnew footprint reference to the new "infra" aka dir, and also providing the footprint.mod files there
<Joerg-Neo900> aiui
<Joerg-Neo900> sed job?
<ceene> i don't thin it is needed
<ceene> i think fp-lib-table is the only file that contains that information
<Joerg-Neo900> aah wait, yes. kicad pcbnew just has a list of dirs to look at, like $PATH for shell, and it takes first matching foorprint it finds
<Joerg-Neo900> right?
<ceene> more than so, fp-lib-table file contains the full path to all footprint files
<ceene> (lib (name neo900-legacy)(type Legacy)(uri footprints/neo900-legacy.mod)(options "")(descr ""))
<ceene> (lib (name neo900)(type KiCad)(uri "footprints/neo900.pretty")(options "")(descr ""))
<ceene> (lib (name lga)(type Legacy)(uri "footprints/lga.mod")(options "")(descr ""))
<ceene> (lib (name stqfn20)(type Legacy)(uri "footprints/stqfn20.mod")(options "")(descr ""))
<Joerg-Neo900> zhough there's still that timestamps concept in pcbnew
<ceene> etc
<ceene> oh, that i don't know about
<Joerg-Neo900> me neither, only heard about it
<Joerg-Neo900> absolute noob here
<Joerg-Neo900> I'm just asking about it in #kicad
<Joerg-Neo900> reply <nickoe> But I don't think it should be a problem. In pcbnew the footprints are embedded in the kicad_pcb file anyways. So you may need to try it out and see what happens when you update a footprint from the modified nlib
<Joerg-Neo900> (I don't really understand) reply <xzcvczx> if you weren't talking about legacy and were talking about changing folder names for "footprints/MyCoolLibrary.pretty/<footprints>.kicad_mod" then if you changed MyCoolLibrary.pretty folder here you would in fact be changing the lib the footprints were in
jonsger has quit [Read error: Connection reset by peer]
<Joerg-Neo900> <xzcvczx> is the timestamping crap not mainly just used for re-referencing? rather than footprint association
<Joerg-Neo900> ceene: ^^^
jonsger has joined #neo900
<ceene> (module neo900:TMD2671 (layer F.Cu) (tedit 5862B9AA) (tstamp 58BB5D2E)
<ceene> (module neo900-legacy:SOT23 (layer F.Cu) (tedit 200000) (tstamp 58BB5D57)
<ceene> (module uDFN-9:uDFN-9 (layer F.Cu) (tedit 58BA9FB6) (tstamp 58BB5D3F)
<ceene> that's the only information i can see on .kicad_pcb file
<ceene> so i think it just looks for the
<ceene> sorry
<Joerg-Neo900> yep, see the (tstamp 58BB5D2E) part
<ceene> so i think it just looks for the module/footprint name and uses whatever there is
<Joerg-Neo900> no idea here, sorry
<ceene> and timestamp i think is only used when you ask pcbnew to update footprints/schematics from .net file
<ceene> just so it doesn't update things that are already up to date
<Joerg-Neo900> ;nod:
<Joerg-Neo900> quite possible
kungfumike has joined #neo900
<ceene> i can't really see how this could go wrong, given that paths are only stored on one single file
<ceene> which is our own fp-lib-table
<Joerg-Neo900> :nod:
<Joerg-Neo900> just making sure
<Joerg-Neo900> you know the less I understand the more I'm worried ;-)
ArturShaik has quit [Ping timeout: 248 seconds]
<ceene> that's usually the case... well, at least for those that know that they don't know; for those that don't know that they don't know, everything is simple, easy and unworrysome :)
<Joerg-Neo900> now for Esa to easily get the *.mod via git pull we need to move all used footprints to the new folder, right?
<Joerg-Neo900> or did you already do that?
<ceene> that's already done in that commit i just pushed
<Joerg-Neo900> \o/ :-D
<ceene> there's nothing for him to do save git pull
<Joerg-Neo900> I hope he'll love that :-)
<ceene> he can modify .mod files as he wish, and commit them
<ceene> although if the modification he wants to do is something that could be done on fped and is useful in a wider scope, it'd be better to modify the .fpd file
<Joerg-Neo900> yes, I already negotiated that part with him
<Joerg-Neo900> he will drop us a note and we'll do it then
<Joerg-Neo900> *we* are responsible for providing correct footprints
<Joerg-Neo900> according to the contract
<ceene> seems reasonable
<Joerg-Neo900> if anything is missing or looks fishy, the recommended operation procedure is he's informing us and we take care
<Joerg-Neo900> he asked for "all the footprints are in the layout already"
<Joerg-Neo900> I think that's warranted
<Joerg-Neo900> ceene: do you agree?
<ceene> the only thing that I didn't do and that is missing, is the RF couplers
<ceene> because that falls in a grey area between footprints-routing
<Joerg-Neo900> pph, ok
<Joerg-Neo900> ooh*
<ceene> and I didn't want to draw them, as they are dependent on layer structure
<Joerg-Neo900> hmm how so?
<Joerg-Neo900> we use discrete thin film couplers
<Joerg-Neo900> not layouted ones
<ceene> i didn't understand that from the project file
<ceene> at least i don't remember seeing any part number
<Joerg-Neo900> though Esa would be competent enough to do couplers in pcb
<ceene> but i may have been wrong
<Joerg-Neo900> no problem
<ceene> error-report.txt:Error: Cannot add new component "J501:/578AFEFA/57F2CBF3" due to missing footprint "OPT-MINI-RF-TAP".
<ceene> that's the one missing
<Joerg-Neo900> CPL-WBF-00D3
<ceene> that footprint exists
<ceene> but there is also J501 J502 J701 J801 J802 whose footprint is called OPT-MINI-RF-TAP
<Joerg-Neo900> "OPT-MINI-RF-TAP" sounds like a receptacle
<Joerg-Neo900> we don't have those in BD
<ceene> ah, i think you're right
<Joerg-Neo900> Hirose or whatever
<Joerg-Neo900> :nod:
<Joerg-Neo900> the 'calibration receptacles'
<Joerg-Neo900> worst case Esa kicks them out and replaces them by a trace ;-)
<ceene> seems good enough
<Joerg-Neo900> but we should find a nice conponent and according footprint, can't be hard. Or did we already?
<ceene> not that i know of
<Joerg-Neo900> wpwrak: ^^^ ?
<ceene> at work, we tipically use MMCX connectors for RF signals
<Joerg-Neo900> wpwrak: do you know about those damn jacks?
<ceene> but maybe they're too big for a mobile device
<Joerg-Neo900> yep
<Joerg-Neo900> wpwrak has a list of all non-trivial components he uses to check sourcing at DK etc. I wonder if the jacks are in there
<Joerg-Neo900> we had datasheets so we even know those jacks only support a maybe 50 mating cycles
<Joerg-Neo900> got lost *somewhere* during the process, it seems
<Joerg-Neo900> where's the dang particular component on DK, for the datasheet?
<Joerg-Neo900> 1.7 mm x 1.7 mm x 0.85 mm; sounds good (AMMC)
<Joerg-Neo900> aah down there's a link
<Joerg-Neo900> then we also need a jack with in and out pin and embedded switch to break the two when plig inserted
<Joerg-Neo900> plug*
<ceene> or placing two jacks and closing the circuit with a small cable when non testing
<Joerg-Neo900> ouch
<Joerg-Neo900> we could plave a footprint over the 50 Ohm trace, and cut the trace when assembing the jack, but that's not reversible
<Joerg-Neo900> not easily at least
<Joerg-Neo900> and since those jacks are on LOWER, we already want final layout for that if any feasible
<Joerg-Neo900> for production we need the jacks with switch
<Joerg-Neo900> "production" == series product. eevrything before are just prototypes, they are assembled, not produced
<Joerg-Neo900> in real industry, there would be Production Validation runs of a few 100 or 1000 items. We don't have that :-)
<ceene> oh, yes
<ceene> if lower ends up being the definitive, that'd be a great advance
<Joerg-Neo900> that's the plan
<Joerg-Neo900> after proto_v2 we should have final LOWER and electrically verified UPPER that needs integration of SoC etc, massive layout re-design
<Joerg-Neo900> all the really delicate stuff like switch / jack placement etc is on LOWER
<Joerg-Neo900> UPPER actually only has keyboard
<Joerg-Neo900> as a "delicate" item
<Joerg-Neo900> and that should be verified and tested in proto_v2 as well
<Joerg-Neo900> the rest is the nasty OMAP FPGA etc
<Joerg-Neo900> which we left out on proto_v2 on purpose, to reduce complexity and to make sure we can test the rest
<ceene> yep
<Joerg-Neo900> reduced complexity also means reduced cost, in that stage
<ceene> everytime we design a board based around an fpga we're always a little worried something will go wrong with the fpga and there will be no way test anything at all
<Joerg-Neo900> exactly
<Joerg-Neo900> thus the BB-xM brainboard
<Joerg-Neo900> we got a dozen BB-xM now
<ceene> i'm liking Arty boards very much lately
<ceene> very useful for a lab
<Joerg-Neo900> nice
<ceene> we've even built a project around that
<ceene> by designing a board that fits in the arduino headers, with lots of RF switches
<ceene> so you save a lot on designing the FPGA part, which is always a pain in the ass
<ceene> and in the end it's cheaper buying that thing fully assembled for 100$ than building one from scratch
<Joerg-Neo900> yep, sire
<Joerg-Neo900> sure*
<atk> a dozen
<Joerg-Neo900> sure, sire
<atk> rare as diamonds
<Joerg-Neo900> ;-)
<atk> a dozen
<atk> must be worth a billion dollars
<atk> so wait
<atk> they ARRIVED?
<Joerg-Neo900> hehe, yes it was a bit tedious to get them
<Joerg-Neo900> atk: no, mouser never delivered
<Joerg-Neo900> got them second hand
<Joerg-Neo900> thanks to ravelo, our best guy
<Joerg-Neo900> and xes
<Joerg-Neo900> sn0wmonster: that was just another "funny" little story, getting those BeagleBoard_xM. Mouser had them on hold for over two years in the end, and kept the prepaid amount for that time, until finally they discontinued them a few days after I got tthem elsewhere finally, and paid back our money
<Joerg-Neo900> there's a lot of those lottle stories that all accounted to "why the delay", though often they were just another roadbock not ghetting effective since we had other closer roadblocks already
<Joerg-Neo900> nevertheless they added to the workload and nerve killing
<ravelo> Joerg-Neo900, thanks :)
<ravelo> btw, there is a rumour someone started producing them again
<ravelo> (Beagleboards)
<Joerg-Neo900> yeah, I heard that but we don't need more of them now
<Joerg-Neo900> do they still come with 'only' 512MB RAM? ;-P
<Joerg-Neo900> if they had upgraded to 1GB finally, I would possibly be interested in getting one or two nevertheless
cc___ has joined #neo900
<ravelo> they are shipped at mouser, only 512 MB RAM
<ravelo> 144 EUR
<Joerg-Neo900> WOW!!!! on stock: 133
<Joerg-Neo900> however "on order: 300" don't get that, how can they have stock og 133 when they got orders on hold for 300?
<Joerg-Neo900> >> Estimated Dispatch Date 300 20.07.2017<< that leaves them around two more weeks to push that date 3 or 6 months into the future, like they did for my order a two years long
<ravelo> maybe they can produce more immediately if more are needed
<Joerg-Neo900> poor mouser, it's not their fault, it's the manufacturer that is fsckng up stuff
<Joerg-Neo900> the bice folks at mouser hotline were pretty pissed about them
<Joerg-Neo900> nice*
<Joerg-Neo900> hmm seems manufacturer for that NN-xM changed, would make sense
<Joerg-Neo900> BB-xM. wasn't it originally built by global solutions or something like that?
<Joerg-Neo900> anyway, still 512MB RAM, so a) good since they don't compete on our precious unique really-rare-as-diamonds 1GB RAM PoP chips, and b) not interesting to get any of those "new" BB-xM
<ravelo> yes
<Joerg-Neo900> atk: could you, as our git expert, give the project repos a sanity check please?
<atk> What do you mean specifically? You want me to see if I cry when I see it?
<atk> Aside from accidentally losing some data when you type the wrong thing, there aren't many ways to screw up the repository itself.
<atk> But I can have a look.
<Joerg-Neo900> "You want me to see if I cry when I see it?" yes basically exactly that :-)
<Joerg-Neo900> our filters from internal to public git might have suffered from the recent commit, for example
galiven_ has joined #neo900
<atk> hmm
<Joerg-Neo900> the new dirs might propagate "too much" or "too little"
<atk> well
<Joerg-Neo900> the new commit might not propagate at all
<atk> someone didn a push --force at some point :P
<Joerg-Neo900> o.O
<atk> although last time I pulled this repository was a long time ago
<Joerg-Neo900> was that pre-internal ?
<Joerg-Neo900> then yes, at that time I guess there been a push --force
galiven has quit [Ping timeout: 260 seconds]
<atk> never mind, I was on an odd branch
<Joerg-Neo900> :-D
<atk> Well, the recent history doesn't look like a complete mess anymore
<atk> I can't view the hooks as that doesn't get cloned.
<atk> Where's the public git access?
<atk> any url I can pass to clone?
<atk> nm, that works
<Joerg-Neo900> atk: many thanks already! :-))
<atk> looks like everything works correctly
<Joerg-Neo900> \o/
<atk> as expected, the public copy has a flattened history
<Joerg-Neo900> :nod:
<atk> what does the filter filter out?
louisdk has quit [Ping timeout: 248 seconds]
louisdk has joined #neo900
<cc___> Joerg-Neo900: the pyra is pretty much in production stage now afaik
louisdk has quit [Ping timeout: 248 seconds]
<Joerg-Neo900> ooh finally? I ordered one, wonder when it will ship :-)
<Joerg-Neo900> didn't look into it since... ages. Actually since I reviewed a halfway final schematics for them
<Joerg-Neo900> atk: many many thanks for taking care about our git!
<atk> no problem
<Joerg-Neo900> I managed to add an account for Esa, took me as little as one full day until it worked like expected ;-)
<Joerg-Neo900> ...and then it turned out to be almost futile thanks to the fped problem
<Joerg-Neo900> now we finally might fix that all for good
xmn has joined #neo900
Pali has quit [Remote host closed the connection]
Venemo has quit [Ping timeout: 240 seconds]
jonsger has quit [Ping timeout: 240 seconds]
jonsger has joined #neo900
jonsger has quit [Ping timeout: 268 seconds]
cc___ has quit [Ping timeout: 255 seconds]
<Joerg-Neo900> atk: thanks for fixing that! and good night! :-)
<Joerg-Neo900> ok, now we need to find a decent tiny RF connector receptacle, like aforementioned AMMC, just with feed-thru and switch, so a real 3pin device (electrically)
<Joerg-Neo900> pv158=407&pv158=994&pv158=414&pv158=997&pv158=940&FV=ffe001b5&mnonly=0&ColumnSort=0&page=1&quantity=0&ptm=0&fid=0&pageSize=25
<Joerg-Neo900> ugh sorry
<Joerg-Neo900> ceene: I guess we'll want the Hirose MS-180.
<Joerg-Neo900> aah here we are, decent DK search:
<Joerg-Neo900> nevermind
<Joerg-Neo900> here's the search now: - I hope :-)