Humpelstilzchen has quit [Read error: Connection reset by peer]
<DocScrutinizer05>
the page name (large topmost subbox) could use bold as well, and a 5 times larger fontsize
<DocScrutinizer05>
I love the page rulers <3
<DocScrutinizer05>
though row "F" looks truncated
<DocScrutinizer05>
I don't think exact squares are a requirement for page rulers/coords
Defiant has joined #neo900
<DocScrutinizer05>
we need to agree on a semantic regarding sheet counting now, with that stuff on index page. Or could we make the pdf content and page numbering start at 0 rather than 1?
<DocScrutinizer05>
for the record: OTG (sheet 2 (page 3)) and Cahrger / OTG-Booster (sheet 3) should probably get merged now
fling has quit [Ping timeout: 244 seconds]
herpderphurr has joined #neo900
fling has joined #neo900
announ has quit [Quit: announ]
<wpwrak>
the sheets are something i'm uncertain about. my concern is about "annotations", i.e., assigning component references (R1234, etc.). kicad can do this for you, and it takes the information from the sheet. so once we start rearranging sheets, problems will arise.
<wpwrak>
and the inconvenient bit is that kicad quite often complains about not having all the "annotations" it expects. most of the time it's trivial or even invisible stuff. and in a simple project, it's easy to check that it's indeed nothing major.
<wpwrak>
however, in something as big as neo900, such checks can get harder. so i need to investigate a bit more how to control annotations, and - once we start changing the sheet structure - how to override the sheet number.
<wpwrak>
(which may mean: identify what needs a real component reference (there are things that only need an internal unique number, such as all the GND symbols), and provide it manually)
<wpwrak>
given that this needs a bit more research, i'm putting this at the end of the transition, and so far i'm avoiding any changes to the sheet structure, even if they make a lot of sense
<wpwrak>
also, we have numerous sheets that bear very little resemblance to what should actually be on them. e.g., sim switch, ir, bob, even hackerbus. so before optimizing the sheet structure, these things should be cleaned up, since they're also affect what sheets we actually need
DocScrutinizer05 has quit [Disconnected by services]
DocScrutinizer05 has joined #neo900
xman has quit [Quit: Leaving.]
pagurus` has joined #neo900
pagurus` has quit [Remote host closed the connection]
pagurus` has joined #neo900
pagurus has quit [Ping timeout: 276 seconds]
dal has joined #neo900
herpderphurr has quit [Ping timeout: 240 seconds]
Xiaoman has quit [Ping timeout: 258 seconds]
Xiaoman has joined #neo900
louisdk has joined #neo900
maddagaska has quit [Ping timeout: 244 seconds]
maddagaska has joined #neo900
louisdk has quit [Ping timeout: 264 seconds]
louisdk has joined #neo900
Oksana has quit [Read error: Connection reset by peer]
louisdk has quit [Ping timeout: 272 seconds]
R0b0t1_ has quit [Ping timeout: 244 seconds]
louisdk has joined #neo900
paulk-collins has joined #neo900
louisdk has quit [Ping timeout: 260 seconds]
ecloud is now known as ecloud_wfh
Xiaoman has quit [Read error: Connection reset by peer]
Xiaoman has joined #neo900
Kabouik has joined #neo900
qwazix has quit [Ping timeout: 240 seconds]
louisdk has joined #neo900
louisdk has quit [Ping timeout: 260 seconds]
louisdk has joined #neo900
SylvieLorxu has joined #neo900
Kabouik has quit [Read error: Connection reset by peer]
qwazix has joined #neo900
qwazix has quit [Ping timeout: 240 seconds]
qwazix has joined #neo900
louisdk has quit [Ping timeout: 240 seconds]
<DocScrutinizer05>
yes, that's why I said "for the record"
fling has quit [Ping timeout: 264 seconds]
<wpwrak>
by the way, speaking of power, do we still want that "plan B" for the BQ27200DRK in the form of "bq27421? bq27441? bq27621?" ? (sheet 5)
<DocScrutinizer05>
it's a plan A2, in that the BQ27200 is not really smart for this usecase, so we might want to add a smart battery monitor
<wpwrak>
does anything depend on the 27200 for compatibility ?
<DocScrutinizer05>
yes
<wpwrak>
if not, it may be a smart move to get rid of it anyway, given that it's EOL
<wpwrak>
ah, darn :(
<wpwrak>
could 27200 and 27421 etc. co-exist ? the comment in the schematics suggests that it's either one or the other
fling has joined #neo900
<DocScrutinizer05>
electrically there shouldn't be a reason why they can't coexist
<DocScrutinizer05>
those critters are pretty much "black magic" iirc, some claim to do extremely smart stuff like testing spectrum impedance of the cell, and I don't see *how* they do that. When I test a cell I need more input variables than those chips acquire (basically just voltage)
<wpwrak>
sounds scary :)
<atk>
sounds interesting
<DocScrutinizer05>
indeed. I *could* see how a chip creates a (AC) load current for the cell and checks the response in voltage domain. But... you know, chip discharging battery??
<DocScrutinizer05>
it's a while since I digested the data sheets
<atk>
It looks like things are taking some form of shape.
<DocScrutinizer05>
anyway see battery gauge in N9, they seem to use one of those magic critters
<wpwrak>
fun fact. seems that "BQ27421G1YZF" does not exist :) (but "BQ27421YZFR-G1A" and a few more variations do) i did find that "G1" a little suspicious ...
<atk>
It's quite cool seeing all the schematics, it all looks so simple on the local level when everything is spread over 38 pages.
<DocScrutinizer05>
atk: for project? we finally took the hard but obviously inevitable leap to move from eagle to kicad. That's basically all that's changing
<DocScrutinizer05>
the schematics were around since 8 months already
<wpwrak>
DocScrutinizer05: the much more important change is that we're now really in control of the schems, and get to change them :)
<DocScrutinizer05>
and 6 of those 8 months we spent 50% of effort with trying to find ways ahead that are less drastical than this switch of toolchain
<DocScrutinizer05>
yes, wpwrak
<DocScrutinizer05>
this is a really great though somewhat unexpoected result form such move
<DocScrutinizer05>
or even the cause
<DocScrutinizer05>
for the move
<wpwrak>
had a look at the BQ27421 data sheet. the way things are done in the schems, U502 is just a big no-op, since BAT ans SRX are shorted.
* wpwrak
adds a TODO ...
* DocScrutinizer05
idly wonders if there are kbd drivers that actively _forbid_ resp filter out N-key rollover, so a second printing key can't produce output as long as a first key isn't released yet
* wpwrak
idly wonders if we don't want to get rid of the keyboard controller and just use "GAIA"
<wpwrak>
especially considering that they're both on UPPER
<wpwrak>
okay, one drawback would be lack of testability in gaia-less v2
<DocScrutinizer05>
that's a big "why don't we" since beginning. Nik's rationale been "it simplifies breakout under GAIA" and you can hardly argue against that
<wpwrak>
okay, if he ran into problems there ...
<DocScrutinizer05>
My arguing if I had done would have been that we need way more intricate breakout for SoC anyway, so how hard coulkd it be to breakout the traces for kbd matric under the twl4030
<wpwrak>
yes, that's what i'd expect, too
<DocScrutinizer05>
I suspect he simply loved the GTA04 design and didn't want to change that detail. In GTA04 it been a logical consequence of original device not having any kbd at all
Xiaoman has quit [Read error: Connection reset by peer]
<DocScrutinizer05>
so yes, using twl4030 kbd matrix controller in final design is clearly an option we want to keep and reconsider, but for the reasons you mentioned above, we should keep the controller chip for proto_v2
<wpwrak>
grmbl. we have an invisible #FRAME component everywhere (and ERC really hates it). so just deleting the eagle legend apparently wasn't enough :(
<DocScrutinizer05>
the above little rant been written by me, so yeah ;-)
<wpwrak>
so it's good that it's almost unreadable in the schematics. that way, the flaws are harder to notice ;-)
<wpwrak>
aah ! DINA4_L has an "A" and a "B" part. "B" is the box, which i deleted. "A" is the invisible element. sneaky.
<DocScrutinizer05>
yeah, noticed that once I tried to augment the frame in eagle
<DocScrutinizer05>
also when writing my auto-indexer
<wpwrak>
well, it ain't no more :)
<DocScrutinizer05>
:-)
<DocScrutinizer05>
could you take care about the F row in kicad frame?
<wpwrak>
kewl. some of the weird component references make kicad think they are actually not component references
<DocScrutinizer05>
sounds nasty
<wpwrak>
(F) looks as if this is roughly what's intended (?)
<wpwrak>
btw, that frame is part the page layout stuff (pl_editor)
<DocScrutinizer05>
I suggest reducing width of all rows so to give F row full width
<wpwrak>
why don't you give it a try ? i don't really know pl_editor anyway (okay, i managed to revert some messiness for anelok, when the default was rudely changed, but that was mainly guesswork)
<DocScrutinizer05>
agrid made of one tiny dot every 2cm would also possibly come in handy, for the sheet coords. Dunno, never seen it before
<DocScrutinizer05>
ok, I can give it a try
<wpwrak>
while at it, you could also fix the legend box. ah, i put the sheet name in Comment1, but it really ought to go into Title. so Title should be the one that sticks out prominently.
<DocScrutinizer05>
I hope the frame is a component which is "linked dynamically"
<wpwrak>
yes, so we won't have merge conflicts :)
<DocScrutinizer05>
could you provide the name of that frame component so I could find it in kicad?
* wpwrak
glares at sheet 24. so much wrong there ...
<wpwrak>
i think you just start with the default layout and tweak it. at least that's what i did. you save it to a *.wks file, which can then be loaded into kicad
<wpwrak>
(in "Page setting" > "Page layout description file")
<DocScrutinizer05>
default is A4 here
<wpwrak>
yuo can change the size in Page Settings
<wpwrak>
(in pl_editor)
<DocScrutinizer05>
*sigh* I want to edit the frame you're using
<DocScrutinizer05>
there seems to be no *.wks in all git
<wpwrak>
i just use the default frame. so that should be the one you get when setting A3
<wpwrak>
just start pl_editor and you get a default frame
<DocScrutinizer05>
yes, that default frame is A4!
<DocScrutinizer05>
unless I load the complete project
<wpwrak>
go to page settings in pl_editor, change to A3
<DocScrutinizer05>
I don't think we want to change the kicad wide page default
<wpwrak>
you're not. you're changing the thing you're editing
<DocScrutinizer05>
this would get nuked by next app update
<wpwrak>
when done you save your .wks, and we'll then use this file for neo900
<wpwrak>
the rest of the world doesn't change :)
<DocScrutinizer05>
ok
<DocScrutinizer05>
so you didn't define a Neo900 specific frame yet
<DocScrutinizer05>
this explains why no .wks is found in ~/Dokumente/Neo900/kicad-et-al/kicad-git/ee
<wpwrak>
now .. how to convince eeschema to remember to apply it ...
<DocScrutinizer05>
the >> 2016-07-20T16:20:58.619669+02:00 localhost kernel: [4691105.683554] type=1400 audit(1469024458.613:627041): apparmor="DENIED" operation="file_mmap" info="Failed name lookup - disconnected path" error=-13 parent=23737 profile="/usr/share/gitweb/gitweb.cgi" name="var/run/nscd/passwd" pid=23738 comm="sh" requested_mask="r" denied_mask="r" fsuid=30 ouid=0 << are as old as gitweb, I complained about them a million times but nobody
<DocScrutinizer05>
really takes care
<DocScrutinizer05>
I thought I fixed them once, but it ressurected
<wpwrak>
can you just turn off apparmor for gitweb ?
<DocScrutinizer05>
I don't think we can "turn off apparnmor for $x"
<DocScrutinizer05>
the problem is with gitweb trying to access name="var/run/nscd/passwd
<DocScrutinizer05>
which is ....
<DocScrutinizer05>
but that was always no problem (except for the admin looking into syslog) when you say misc git works
<DocScrutinizer05>
ee git not working has some other reason
<DocScrutinizer05>
NB I *never* before created a git repo
<wpwrak>
i got also an error with "misc" now. but only after a while, so it worked mostly
<wpwrak>
lemme see if i can clone it over ssh ...
<wpwrak>
yup. works perfectly
<DocScrutinizer05>
I rather tend to blame webserver
<wpwrak>
could be
<DocScrutinizer05>
shopping pending, bbl
<DocScrutinizer05>
maybe hellekin or dos1 can help us out
<DocScrutinizer05>
the apparmor loglines clearly state wich file access got blocked. Please check this is really the reason for the problem before messing with apparmor policies
<DocScrutinizer05>
and even then, rather fix gitweb
<wpwrak>
i'll leave it to { DocScrutinizer05, dos1, hellekin } ;-)
<DocScrutinizer05>
I don't see why gitweb e.g. needs access to var/run/nscd/passwd or /proc/loadavg
<wpwrak>
DocScrutinizer05: you may want to enjoy a comparison between the eagle version and the current kicad for the right side of sheet 24, the digital microphones :)
<wpwrak>
(git clone problem) cloning misc via http fails with a 408. so the real issue could be some network problem.
paulk-collins has quit [Ping timeout: 244 seconds]
<wpwrak>
but for ee, the problem is different: neo900.org 181.171.122.65 - - [20/Jul/2016:16:59:01 +0200] "GET /git/ee.git/info/refs?service=git-upload-pack HTTP/1.1" 404 966 "-" "git/2.6.2"
<DocScrutinizer05>
btw told ya it's way too easy to forget checkout before editing a file, and then you enter git hell
<DocScrutinizer05>
so this git workflow doesn't work for me
<wpwrak>
you just have to get used to pulling in things before editing. and if it's something you share with someone else, tell them to avoid making changes while you edit it
<DocScrutinizer05>
damn, I pulled a 30 min before editing
<wpwrak>
git just tells you when something went wrong. with scp/rsync, you'd just overwrite each other's changes without even knowing what happened
<wpwrak>
i was working on sheet 24. didn't know you'd change it, too
<DocScrutinizer05>
I didn't
<wpwrak>
hmm. in any case, it seems that sheet 24 was the problem
<DocScrutinizer05>
I didn't touch anything
<wpwrak>
try to commit and push the wks now
<DocScrutinizer05>
I copied the neo900pageframe_A3.kicad_wks into the git repo and then did a commit
<DocScrutinizer05>
I won't do anything except c&p commandline snippets
<DocScrutinizer05>
ooh that worked? looks like mega fail
<DocScrutinizer05>
I'd not expect half a screenfull of educational output on a succeeded command
<DocScrutinizer05>
NB the sting "OK" OWTTE didn't show up a single time
paulk-collins has joined #neo900
<DocScrutinizer05>
(([2016-07-20 Wed 17:05:37] <wpwrak> i was working on sheet 24. didn't know you'd change it, too)) even *if* I had done (which I didn't), I for sure didn't commit
<wpwrak>
it tells you that there is a config choice you should make at some point
<DocScrutinizer05>
anyway, shopping
<DocScrutinizer05>
what I did was copy the A3frame file into the dir (==edit), then
<wpwrak>
btw, when working on a shared repo, you should always git pull immediately before git push
<DocScrutinizer05>
so you need to make the script that way ;-)
<wpwrak>
by the time you're familiar enough with things to get scripts right, you don't need scripts anymore :)
<DocScrutinizer05>
I won't go there
Pali has joined #neo900
<wpwrak>
stack exchange is still useful, though, when hitting more exotic problems
<DocScrutinizer05>
I'm explicitly refusing to take a git crashcourse lesson now or any time of the near future
<wpwrak>
btw, congratulations on your first commit ! :)
<DocScrutinizer05>
meh
<DocScrutinizer05>
rather condolences to my fist branch merge or whatever
<wpwrak>
that wasn't a branch :)
<DocScrutinizer05>
first clean simple commit still pending
<wpwrak>
eeschema picked up the wks change automatically. perfect.
xman has joined #neo900
<DocScrutinizer05>
I guess I need sth roughly along the line of "cc=`pwd`; trap "cd $cc" EXIT; cd $EE; git pull; mv $1 ${1]_AWAY; r=$?; git checkout $1; cp $cc/$1 .;[ 0 -ne $r ] && git add $1; git commit $1; git push "
<DocScrutinizer05>
s/-ne/-eq/
<DocScrutinizer05>
err nope
<DocScrutinizer05>
I guess I need sth roughly along the line of "cc=`pwd`; trap "cd $cc" EXIT; cd $EE; git pull; git checkout $1||:; mv $1 ${1]_AWAY; r=$?; cp $cc/$1 .;[ 0 -ne $r ] && git add $1; git commit $1; git push "
<wpwrak>
naw, you don't need anything so complicated. the moving away was just to isolate the conflict, since i don't know what exactly you've been doing. normally, you just pull and you're good
<DocScrutinizer05>
dang, I did
<Pali>
git status
<wpwrak>
just pull before pushing, to be up to date with the repo
<DocScrutinizer05>
that would overwrite my edited file, no?
<DocScrutinizer05>
honestly, please *test* the procedure
<Pali>
git pull is "alias" for git fetch && git merge
<wpwrak>
DocScrutinizer05: your files are safe, especially if we make sure we don't work on the same file at the same time
<Pali>
lit a bit compilicated, but git pull is written in shell: /usr/lib/git-core/git-pull
<DocScrutinizer05>
I pulled a 30 min before I edited just one file and then did a commit and push. The commit worked fine, the push failed epically. If there's a race condition in git pull then what is this tool worth for?
<Pali>
provide error message from git push
<Pali>
I can decode it for you :-)
<DocScrutinizer05>
I pastebinned everything above
<wpwrak>
just pull again. i committed unrelated stuff during those 30 minutes. the pull makes sure the unrelated stuff gets incorporated into your repo. (and since it's unrelated, that's not a big deal)
<Pali>
going to look at chanlog
<DocScrutinizer05>
when I need to do git pull *after* editing, you need to tell me so
<DocScrutinizer05>
how hard can it be tpo provide a failsafe procedure to push one file via git?
<wpwrak>
yeah, sorry. it's a bit like walking, once you know how to do it, you don't think of all the simple movements that make up a step :)
<DocScrutinizer05>
exactly my point
<wpwrak>
when git complains about something, that's usually a situation where you need to decide that nothing bad has happened
<wpwrak>
git tries to be smart, but it can't quite guess all the things people might be doing ;-)
<wpwrak>
btw2, didn't you want to make the title bigger ? or is it big enough now ?
<wpwrak>
btw2, while i'm wrestling with sheet 24, we have TSC2007IPW or CRTOUCH. does it really make sense to have TSC2007IPW still sitting there ?
<DocScrutinizer05>
I never touched sheet24. definitely. I wanted the sheet name to be bigger, size of title. You said title _will_ be sheet name, something went to wrong field, right? btw2.2 will have to wait until I got shopping finished
<DocScrutinizer05>
a quick answer: no
<DocScrutinizer05>
legacy cruft
<wpwrak>
i already moved the sheet names to the title
<wpwrak>
(TSC2007) thought so. scheduled for deletion :)
<DocScrutinizer05>
I still wait for the generic instructions how to commit a file regardless of concurrent git push changes to the repo by other persons
<wpwrak>
the solution was, literally, "have you tried turning it off, then on again ?"
<ceene>
lolç
<ceene>
so
<ceene>
i see
<wpwrak>
(git-daemon needed restarting to fully become aware of the new repo)
<ceene>
cvt, eagle and hw directories
<ceene>
what's all that? :P
<wpwrak>
all the new stuff is in hw/
<wpwrak>
eagle/ has the original design files
<wpwrak>
cvt/ are the untouched conversion results (the conversion takes a good while, with various interactions, so i kept the output for reference, even if it's not strictly "source")
<wpwrak>
also, some things under cvt/ haven't been incorporated into hw/, e.g., the footprints
<wpwrak>
(haven't been) yet :)
<ceene>
ok
<ceene>
looks good
<ceene>
now i need to learn how to navigate this thing
<ceene>
i clicked on a sheet
<ceene>
and now i don't know how to back to the main one
<wpwrak>
right-click, leave sheet
<wpwrak>
or click on the "Navigate schematic hierarchy" icon
<ceene>
oh, nice
<ceene>
love my tiling window manager
<ceene>
don't love that the navigate schematic hierarchy window gets shut off when i click on some other sheet
<ceene>
i want to see them all at the same time!
<ceene>
a bit arcane
<ceene>
but looks good
<ceene>
i think the pcb router is better than the schematic editor
<DocScrutinizer05>
((<ceene> oh, nice)) not really
<DocScrutinizer05>
those subsheets are kicad's major brainfuck
<DocScrutinizer05>
And the worst: you can't simply browse through pages, or flip swiftly between sheet N and N+1
<DocScrutinizer05>
I tried to create a kbd shortcut, which possibly even would have worked, if the VM wouldm't reject the hotkey manager's "Alt" in "Alt+V"
<DocScrutinizer05>
obviously the hotkey manager sends x11 key events, while the VM reads raw keycodes
<ceene>
why are you running kicad under a VM?
<DocScrutinizer05>
so "Alt+V" becomes "V"
<DocScrutinizer05>
ceene: I can't get the recent version on my host
<ceene>
debian has the latest it seems
<ceene>
debian sid, i mean
<DocScrutinizer05>
RPM system here
<DocScrutinizer05>
the VM has latest stable kicad, built locally under genuine devuan
<ceene>
which is devuan status?
<DocScrutinizer05>
beta
<DocScrutinizer05>
but rock solid for a beta
<DocScrutinizer05>
several of the core devels use it productive in their data centers
<DocScrutinizer05>
there are several hiccups, sure
<DocScrutinizer05>
but the core system just works, and it does so really nicely
<ceene>
how many people work on it?
<Sicelo>
it tracks debian .. so yes .. should be rock solid
<ceene>
ei, hi, Sicelo :)
<DocScrutinizer05>
ask hellekin
<ceene>
how is it going?
<Sicelo>
hey ya ceene. good thanks. yourself?
<ceene>
i feel like a normal person using whatsapp now :/
<Sicelo>
haha, i left it permanently, and happy about it
<DocScrutinizer05>
eew, condolences ;-)
<Sicelo>
had been wanting to quit for a looooong time
<DocScrutinizer05>
anyway kicad looks rather great
<Sicelo>
at least all members of my immediate family (except dad) were willing to move to Telegram for me .. so good enough. i use it in Pidgin, and web on my computer at work, home, and laptop
<Sicelo>
*pidgin fo n900
<Sicelo>
DocScrutinizer05: i seem to have read passively about some further hiccups for Neo?
<ceene>
good!
<ceene>
i hope i can write a telegram version of yappari on september
<Sicelo>
that'd be nice ceene. the pidgin plugin works really good (for me)
<DocScrutinizer05>
Sicelo: nope, there are no special hiccups regarding N900 or Neo900. Whet's needed is a CI building for armel
<DocScrutinizer05>
so far official support only for armhf
<DocScrutinizer05>
basically a negligence
<DocScrutinizer05>
and no issue at all to create a CI for armel, aiui
<DocScrutinizer05>
just somebody needs to do it
<Sicelo>
okay .. there's still lots of persons interested in it :) at DebConf it came up few times
<Sicelo>
and i saw TOH .. looks quite good tbh
<DocScrutinizer05>
kbd TOH?
<DocScrutinizer05>
well
<Sicelo>
yes, that one .. Jolla
<DocScrutinizer05>
I sort of stopped touching Jollaphone kjust like I did with N9, when I realized you need to send device in for a reflash, when you nuked the closed partitions
<wpwrak>
btw, i'm merging some multi-unit components into a single unit. this produces some growling from eeschema, with warnings that the library differs from the cache and so on. just tell it to shut up :)
<wpwrak>
phew. all component references check out now.
<wpwrak>
the bad news: ERC has 830 complaints
<DocScrutinizer05>
ouch
<wpwrak>
i especially like those: "Global label POWERON is not connected to any other global label."
<DocScrutinizer05>
but that's expected, I guess none of the components _really_ has sane pin class settings
<wpwrak>
oh, that comes later ...
<wpwrak>
and eagle actually seems to have produced halfway decent settings there
<DocScrutinizer05>
yeah, a lot of 'off-sheets' points to nowhere
<DocScrutinizer05>
thanks zo lack of GPIO from (non-existent) SoC
<DocScrutinizer05>
should get better drastically as soon as we throw in the (proto_v2 specific too) IO-extenders
<DocScrutinizer05>
often it's not even off-sheets but simply an unterminated wire like net$2345
<wpwrak>
yes, there are those, too ...
<DocScrutinizer05>
I'm actually surprised there are *only* 830
<DocScrutinizer05>
you know about error amplification, from c compilers. a single typo can cause 400 errors. Same with ERC
<wpwrak>
yes, some issues produce multiple errors
<DocScrutinizer05>
I suspect a _lot_ of ERC errs will be about GND
<DocScrutinizer05>
and power rails
<wpwrak>
anyway, what all this was really about was that i wanted to know what things ERC would find worth complaining about on the keypad sheet. turns out, that it's okay with most of it.
* DocScrutinizer05
ponders symlinking the ee dir into the VM exported dir, but... rather copies the stuff
<wpwrak>
you really ought to get / set up that new machine :)
<DocScrutinizer05>
and then?
<wpwrak>
then you don't need a vm anymore
<DocScrutinizer05>
then I have a shiny kicad running not in a VM (which at least has a shared dir to host, and shares kbd and mouse and monitor) but a real machine completely detached from my workstation
<DocScrutinizer05>
I don't see any advantage other than maybe OpenGL being easier to get working
<DocScrutinizer05>
and the whole topic is *completely* orthogonal to git topic
<wpwrak>
well, for git, yuo could set up networking in your vm and just run it in there ...
<DocScrutinizer05>
IOW I would still copy stuff out of the git dir, even if I had git dir on a dedicated kicad iron
<wpwrak>
and the machine could be your migration target. i.e., once you're comfortable there, leave suse
<DocScrutinizer05>
yeah, when i'm terribly bored, eventually
<DocScrutinizer05>
please don't expect to migrate me to devuan now that I been touching git more than I planned to. This doesn't imply anything beyond the mere fact
<DocScrutinizer05>
s/devuan/ubuntu/ *freud*
<wpwrak>
;-)
<DocScrutinizer05>
I pretty much hate buntkuh and canonical, for all they do and are. And that won't change any time soon
<DocScrutinizer05>
ubuntu has a record of striving towards a windows-me-too and forcefeeding crap to users, way longer than debian / suse have with systemd
<DocScrutinizer05>
IOW I _never_ heard of any *really* professional company using ubuntu for a mission critical environment
<pigeons>
have to agree
<wpwrak>
ubuntu desktop gets a bit overloaded, yes
<DocScrutinizer05>
before I migrate to ubuntu, I'd rather go for CentOS or RHEL or SLE
<pigeons>
well they make lots of choices for you, that you have to figure out whats going on and undo
<wpwrak>
what you can do is install ubuntu server, then whatever user interface you really want
<pigeons>
i do know a shop using gentoo
<DocScrutinizer05>
or Devuan
<wpwrak>
note: before trying to boot, make absolutely sure the linux-image-extra matching your kernel is installed. else, you may find support for irrelevant features like the keyboard missing ...
<DocScrutinizer05>
in ubuntu? irrelevant
louisdk has joined #neo900
<wpwrak>
a server install even gets rid of consolekit ;-)
<DocScrutinizer05>
I'f call myslef douchebag if I decided to use ubuntu despite all my notions about what's wrong with it
<wpwrak>
as i said, there's a lot less wrong with ubuntu server than with ubuntu client. but you're still compatible with all the packages.
<DocScrutinizer05>
I don't care
<DocScrutinizer05>
sounds to me like "there's a lot less wrong with Junge-Union than with CDU"
<wpwrak>
btw, more stuff to pull
pagurus` has quit [Remote host closed the connection]
<DocScrutinizer05>
I'll pull before I look for any files
<wpwrak>
not my battle is at the keypad. that one has its set of mysteries, too. e.g., those six "N900" buttons.
<wpwrak>
s/not/now/
<wpwrak>
and of course all the component references are bogus ...
<DocScrutinizer05>
nik added a note that kbd is a construction site
<DocScrutinizer05>
so nothing unexpected there
* DocScrutinizer05
idly wonders how that will pan out when KiCAD layout editor, on reading in the netlist from schematics, will place a thousand and then some componently all on the same spot
<DocScrutinizer05>
-ly, where did that come from?
<wpwrak>
yeah, moving all the components to a reasonable place will be fun. this is one of the more annoying "features" of kicad
<DocScrutinizer05>
I ponder to add a pseudorandom XY offset while KiCAD does that placing
<DocScrutinizer05>
or rather, write a python scropt to spread that deck over the place
<announ>
The FAQ claim that the Neo900 can be used with a 100% Free Software stack; but will it ship that way, i.e. with no proprietary software pre‐installed?
<DocScrutinizer05>
yes
<DocScrutinizer05>
it actually ships with pretty mich no software preinstalled. Just like a plain (Asus) PC motherboard
<DocScrutinizer05>
except the firmware inside modem which is a blackbox in a sandbox to us
<DocScrutinizer05>
and we sandbox it since we consider all blackboxes rogue per se
<announ>
I though that Debian GNU/Linux would be bundled (thatʼs what the Specs page says)
<DocScrutinizer05>
yes, a Devuan Board Support Package ships with the system
<DocScrutinizer05>
you don't want to use that for anything but development and running the factory test software to verify correct function of all subsystems
<DocScrutinizer05>
we probably also will ship an install wizard that simplifies installing your choice of OS that community prepared for running on Neo900
<DocScrutinizer05>
our recommended OS is maemo neo-fremantle
<announ>
Thanks for all the clarification and information, very useful
<DocScrutinizer05>
yw :-)
<DocScrutinizer05>
wpwrak: pushed new frame
<DocScrutinizer05>
announ: please see /topic! useful links in there :-)
<DocScrutinizer05>
announ: what's not visible on our website and/or talk.maemo.org yet: we are inmidst of migrating from eagle EDA (proprietary tool) to KiCAD EDA (FOSS), so the project becomes way more "open hardware"
<wpwrak>
nice !
<wpwrak>
let's kill the (c) crap then ...
<DocScrutinizer05>
what about the concrats?
<DocScrutinizer05>
;-P
<DocScrutinizer05>
congrats even
<DocScrutinizer05>
first clean commit
<DocScrutinizer05>
actually second
<DocScrutinizer05>
you must be happy
<DocScrutinizer05>
Wizzup: thanks again, worked great
<wpwrak>
see, it's easy :)
<DocScrutinizer05>
yeah, with proper instructions (and an alias)
<wpwrak>
purge complete
<wpwrak>
(c) purge, to be exact
<Defiant>
ooh are the kicad files online?
trx has quit [Ping timeout: 240 seconds]
trx has joined #neo900
trx has quit [Changing host]
trx has joined #neo900
<Wizzup>
DocScrutinizer05: welcome
<atk>
DocScrutinizer05: so how is that router you were liking the look of working out for you?
<atk>
DocScrutinizer05: the routing features of kicad
<atk>
you finally found the settings window, does it all work as you expected?
<DocScrutinizer05>
aah, ok. The routing looks fine, in a few very basic tests
<DocScrutinizer05>
yes
<DocScrutinizer05>
I thought you asked about a router ;-)
<DocScrutinizer05>
as in Ubiquity ERpro8
<atk>
is that what you use? :P
<DocScrutinizer05>
yep
<atk>
cool
<atk>
I've heard good things about ubiqity routers
<DocScrutinizer05>
pretty much bang for the buck
<atk>
but isn't the ERpro8 a rack mount beast for professional use?
<DocScrutinizer05>
steep learning curve (unless you are a network engineer)
<atk>
Do you really need something that expensive/powerful? :P
<DocScrutinizer05>
well, it's a versatile device that doesn't throw sticks at your legs or doesn't allow *any* config
<DocScrutinizer05>
and yes, so far 5 of the 8 ports are already used, and this isn't yet the planned final network topology
<atk>
I've been considering getting a ERPoe-5 and getting one of their POE wifi things
<atk>
My network is set up so that the router is in another room (next to the modem for wifi reasons) but with a good enough wifi thing I could move it to the same room that I have everything else in.
<DocScrutinizer05>
time to get me a Pizza
<DocScrutinizer05>
bbl
<atk>
I just use a switch for all the machines connected to it.
<atk>
is that breakfast?
<DocScrutinizer05>
yep
<DocScrutinizer05>
I started frowning at switches when I realized nothing stops the proprietary stuff connected to my LAN from ARP-poisoning the living shit outa the whole LAN
<DocScrutinizer05>
I got printers, Home Automation devices, whatnot else
<zakx>
ERPros suck in production
<DocScrutinizer05>
amd at least the Brother and the HomeMatic MFC are 'calling home'
<zakx>
as soon as you want to do anything that they cannot offload to hardware, you're fucked
<DocScrutinizer05>
offload can get disabled
<zakx>
it should be ok for home use, but as soon as you have real traffic, spend some more money
<DocScrutinizer05>
yeah
<DocScrutinizer05>
there's a reason they don't cost the same as a Cisco or whatever
<zakx>
"anything" can even mean running an openvpn server on them for like 3 clients
louisdk has quit [Ping timeout: 272 seconds]
<DocScrutinizer05>
the hw accel is a tricky beast, but anyway even without it it's still mostly a fine device
<DocScrutinizer05>
this is a SOHO LAN, anything more beefy would be a real overkill
<zakx>
feature-wise it is unparalleled in the price segment
<DocScrutinizer05>
and afaik you can use offloading on all ports where you don't need "special stuff"