<wpwrak>
(no git hist) are you talking about eetest -H . or ceene's problem ? eetest -H . (when in neo900's ee repo) should work for yuo
<Joerg-Neo900>
ceene prob
<wpwrak>
well, let's see what the diagnostics i've added will show. if they're inconclusive, it's time for eestest -C HEAD:neo900.pro that one is also easier to trace, if needed, since it does much less than eeshow does. could also be that the problem is elsewhere, e.g., somewhere in path name generation. or maybe it's a memory corruption. we'll see.
<Joerg-Neo900>
strange enough both show a history, I had expected them to both fail to notice there's a git at all
<wpwrak>
naw. eeshow is very persistent in finding repos. it's usually the opposite: it's hard to make it not find one ;-)
Pali has joined #neo900
<Joerg-Neo900>
aaah :-)
<wpwrak>
ah. interesting. there are indeed weird things happening when using hw/neo900.pro
<wpwrak>
ee/hw/neo900.pro and neo900.pro both do exactly the same. but hw/neo900.pro doesn't. prolly something going wrong with some the path name construction. let's see ...
<Joerg-Neo900>
well, there's a .git in /ee and one in ee/hw/
<Joerg-Neo900>
umm
<Joerg-Neo900>
nope
<wpwrak>
that's a little odd, but could be due to the age of your system (the handling of submodules changed and they rearranged something of that sort, forgot the details)
<Joerg-Neo900>
nah, I messed up my own stack of factoids
<Joerg-Neo900>
there's no ee/hw/.git
cybiko123 has joined #neo900
<wpwrak>
in any case, i can reproduce the problem here. it's caused by pathname construction not taking into account the prefix of the "parent" file.
<Joerg-Neo900>
line two gave it away my factoid was wrong
<wpwrak>
(parent) eeshow has a concept of "related" files, e.g., neo900.pro is related to the files we obtain from it (either by parsing it or by just changing file extensions). these files then inherit revision and path
<Joerg-Neo900>
voodoo :-)
Pali has quit [Ping timeout: 246 seconds]
ecloud has quit [Ping timeout: 268 seconds]
<wpwrak>
yeah. quite a bit of voodoo there, unfortunately. restructuring that jungle is one of my long-term to do items for eeshow. right now, it handles about half the possible situations it could encounter right, and may get very confused in the rest. fortunately, the half it handles well corresponds to what most people use :)
<wpwrak>
(but already most qi-hw projects are a bit outside this range. fortunately, the differences don't cause too much trouble if no going too deep into the history)
knttl has quit [Ping timeout: 250 seconds]
knttl has joined #neo900
joerg_ has joined #neo900
<joerg_>
Web Chat via Android, ein Bier Rebell. ...LOL so much for speech input
chomwitt has quit [Ping timeout: 256 seconds]
_whitelogger has joined #neo900
cybiko123 has quit [Quit: Leaving.]
_whitelogger has joined #neo900
DocScrutinizer05 has quit [Disconnected by services]
<wpwrak>
ceene: ah, this error is interesting. which version of libgit2 are you using ? and can you tell me where it segfaults ? that seems new. you can run eeshow under gdb by adding "gdb" as first argument, i.e.,
<wpwrak>
seems that libgit2 changed the rules for git_repository_open_ext a little between 0.22 and 0.24
<ceene>
oh, by the way, I've seen you've cited me as "ceene" on your commits :)
<ceene>
you can use my real name and email address if you want
<ceene>
i usually commit as Alvaro G. M. <alvaro.gamez at hazent.com>
<wpwrak>
ah, didn't have your real name anywhere :) so i use "Alvaro Gamez" then ? or "Alvaro Gamez M." ? btw, gamers must be dripping with envy when they learn of your name ;-)
<ceene>
lol
<ceene>
I just use Alvaro G. M., as my surname is already on the email address
<ceene>
also, since I like both of my surnames Álvaro Gámez Machado i hate having to choose one over the other!
<wpwrak>
i normally don't use the email address in this kind of attributions. no point in having people bother you about my bugs. so "Alvaro Gamez Machado" ?
<ceene>
sure :)
<wpwrak>
which distribution are you using ?
<ceene>
debian
<ceene>
sid
<ceene>
but hasn't been upgraded lately
<ceene>
so maybe it's sid from 6 months ago
<ceene>
but hmmm on the computer it's debian testing, and the same thing happens
<ceene>
on the other computer, the one at home, i meant
<wpwrak>
okay. that may explain why i don't see this issue with the libgit2-24 on ubuntu. (just upgraded from -22, but it still works fine here)
<wpwrak>
let's try a "blind" fix then ...
<wpwrak>
the commit i just pushed should make the segfault go away
<ceene>
correct
<ceene>
but :)
<ceene>
pressing 'd' key also makes a sigsev
<ceene>
Program received signal SIGSEGV, Segmentation fault.
<ceene>
0x0000000000420138 in vcs_git_summary (h=0x0) at file/git-hist.c:333
<ceene>
333if (!h->commit)
<ceene>
i guess you'll have to check every access to h all around
<ahycka>
wpwrak, thanks for fixing all "segmentation fault" :D:D:D
<wpwrak>
yeah, the problem with finding the git repo put eeshow into a configuration i almost never have. and thus my sloppiness got revealed ... :)
* Joerg-Neo900
would check each malloc with an assert if it would make any sense
<Joerg-Neo900>
;-)
<wpwrak>
ceene: ah, you may find that, then you invoke something that overlays other things (e.g., the index), the underlying controls still work. you can also produce interesting crashes that way :) i haven't quite decided if there are actually valid use cases where this feature would be useful, or whether i should rather make the whole system modal
<Joerg-Neo900>
but yeah, that's the requirement to myself, the reality often looks different :-)
<Joerg-Neo900>
wpwrak: did you change -N default to 20 or whatever?
<wpwrak>
Joerg-Neo900: no, -N should work exactly as it always did
<Joerg-Neo900>
ticking bomb :-S
<Joerg-Neo900>
it works like it always did, just the default is different
<ahycka>
wpwrak, my real problem was a cat walking along the keyboard xD
<Joerg-Neo900>
so it doesn't kill systems
<wpwrak>
some day i'll add a "stop" button. or maybe some [click here for more] feature.
<wpwrak>
ahycka: aah, cats make the best testers ! :)
<Joerg-Neo900>
HAHA stop button, my RAM got eaten before the thing got a chance to draw a GUI in X11
<ahycka>
yeap! she is sleeping till i switch on the laptop
<Joerg-Neo900>
a stop button would assume I'm faster than my PC
<Joerg-Neo900>
ahycka: they are envious little masters. Why you dare to pet that nasty plastic thing, pet *HER* ;-D
<wpwrak>
Joerg-Neo900: that's a little extreme. well, maybe there are some resource leaks or particularly inefficient constallations, too. all those optimizations will need some time for analysis of what's really going on.
<Joerg-Neo900>
if you don't want to change default for -N, maybe add a new option -M byte<UNIT> for the max RAM the process may allocate
<Joerg-Neo900>
and set that per deafilt to 50% of `free`
<wpwrak>
Joerg-Neo900: that may be tricky. chances are that these allocations happen in libgit2, especially if you run out of RAM even before the GUI is up.
<Joerg-Neo900>
that's why I strongly suggest to simply change the -N default
<wpwrak>
naw, you normally just use "make view" anyway. no point in spending a lot of time optimizing such rare issues.
<wpwrak>
(default) "make view" does that for you :)
<Joerg-Neo900>
I normally don't expect a command to nuke my systenm!!!
<wpwrak>
anyway, time to do some shopping. fridge is empty after the long weekend.
<Joerg-Neo900>
I really hate to add a safety wrapper around eeshow, but I'll do and commit that
<Joerg-Neo900>
it's a nogo that even the savvy but not exactly alerted (almost) *kill* their system by a simple commandline
<Joerg-Neo900>
yes, the system is fsckdup with lack of proper ulimits and whatnot, but...
<Joerg-Neo900>
to elaborate: my system /Aactualy all my linux systems I had so far, during the last 20 or 30 years) went into a swap hell in such situation and it literally never comes back
<Joerg-Neo900>
s./A.(.
<Joerg-Neo900>
you can't enter any commands to stop the runaway process either
<Joerg-Neo900>
X11 doesn't process the event queue anymore
<Joerg-Neo900>
dreeeeeeze
<Joerg-Neo900>
f*
<Joerg-Neo900>
(no X11 events -> no ctrl-alt-F1 for console :-S )
Pali has joined #neo900
<ceene>
Joerg-Neo900: sysrq+r, will make keyboard events pass to kernel directly, so x11 doesn't intercept them and you can go to console
<ceene>
alt+sysrq+r
<ceene>
it'll still take at least a minute or so to obey
<DocScrutinizer05>
if only friggin systemd hadn't nuked sysreq keys on my system :-((
<Joerg-Neo900>
meh, wrong client
<ceene>
i hate that shit too
<ceene>
i've found that having a swap too big produces that
<ceene>
in fact, i've found that no swap at all gives bests results
* Joerg-Neo900
just looks at `help ulimit`
<Joerg-Neo900>
cmd: total used free shared buffers cached
<ceene>
if there's no swap available, either malloc fails or kernel oom kills the offending process pretty soon
<Joerg-Neo900>
oom alas doesn't (always) kill the *offending* process
<ceene>
usually does
<ceene>
but it can get t wrong, yes
<Joerg-Neo900>
often it kills innocent processes that happen to be the unlucky one to malloc the last 10bytes plus one
<ceene>
you may configure limits under /etc/security
<ceene>
at least on debian
<ceene>
/etc/security/limits.conf and /etc/secirity/limits.d
<Joerg-Neo900>
I *think* I upgraded from suse5.x in 1999 to now opensuse13.2, in an uninterrupted chain, so possibly it's time to install an all fresh system :-/
<ceene>
i've always found linux to be very easily reinstallable
<ceene>
windows takes hours to upgrade and dozens of reboots
<ceene>
anyway, things under /etc/security are pam modules configuration files
<ceene>
so if your system uses pam, you'll have somewhere those same files
<Joerg-Neo900>
it's not about installing linux, that's ultrasimple. It's about customizing the system so you feel 'at home' which takes weeks or even months
<Joerg-Neo900>
and never works to 100% reproduce all the little tweaks you had applied and then forgot about them
<Joerg-Neo900>
suddenly keymappings in shell are not as you want them, kmail doesn't display the mails in the way you want, hotkeys on desktop are missing (some of which you only use twice a year but THEN you NEED them) aso aso
<ceene>
well, that kind of thing should go whenever is possible under home
<ceene>
so in theory you could even change distros and have your environment as before
<Joerg-Neo900>
...or you notice "damn, sysreq keys don't work! what was that? ctrl-shift-print? alt-print? ctrl-alt-break?"
<Joerg-Neo900>
"did I change the ssh port from 22 to 223344 in router port forward or on the PC, and if on PC where the heck and what was the correct config file/line?"
<Joerg-Neo900>
stuff like that, again and again, every day, for weeks
<Joerg-Neo900>
you can't keep all that in $HOME
<MonkeyofDoom>
that really seems like a failing of the distro imo; package managers should be able to diff configuration files against defaults
<Joerg-Neo900>
;-)
<MonkeyofDoom>
or user modifications to configs should be stored as diffs
<Joerg-Neo900>
MonkeyofDoom: I'm talking about a plain fresh new install, not an upgrade which usually *mostly* works
<MonkeyofDoom>
right, but you should have a way to extract your customizations from the old corpse ;)
<Joerg-Neo900>
hmm, didn't find any
<MonkeyofDoom>
it's an underserved use-case, sadly
<ceene>
the solution most probably lies on making a simple list annotating what are you changing when installing a new system
<Joerg-Neo900>
yes. There's not even a script (or whatever tool) to migrate a system to a new iron - well maybe apart from autoyast, for fresh systems without data and customizations
<ceene>
i see there's a HPF with cut frequency 10Hz
<ceene>
isn't that too low?
<ceene>
i'd use at least 100Hz, so you cut out 50/60Hz from the power grid
<Joerg-Neo900>
that is sort of obsolete
<ceene>
and flicker from TV screens
<ceene>
well, it's dated 01-Jul-2016
<ceene>
not so obsolete, although i haven't looked at the schematics
<Joerg-Neo900>
alas we have IR-UART which has up to 10 symbols (aka bits) with no state change
<Joerg-Neo900>
anyway the RX is designed to be very insensitive and supposed to only work up to maybe 5cm distance to IR-TX
<ceene>
in any case, frequencies under 100Hz are very common and could be problematic
<Joerg-Neo900>
it's only for debugging by connecting to console, and for learning CIR, so the usecases are such that you can control the amount of noise
<ahycka>
do you know that bq27200 chip is not recommended for new designs?
<Joerg-Neo900>
yes
<Joerg-Neo900>
of course :-)
<Joerg-Neo900>
we want it nevertheless for N900 legacy
<ahycka>
ok!
<Joerg-Neo900>
great, somebody reviewing our schematics. VERY welcome and appreciated
<Joerg-Neo900>
:-))
<ahycka>
i just wanted to be sure :)
<bencoh>
Joerg-Neo900: I dont really understand the legacy aspect here
<Joerg-Neo900>
of course, many thanks. Please go on with review
<Joerg-Neo900>
bencoh: there are several apps that use the bq27200 directly
<bencoh>
sure but ....
<Joerg-Neo900>
bencoh: eventually we simply may not pupulate the chip, we have a concurrent better battery gas gauge already
<Joerg-Neo900>
bencoh: they are both populated concurrently, so no probably they are not pin compatible but that doesn't matter
<Joerg-Neo900>
one 'problem' we have with bq25421 is: we need to pick the chip variant for the battery Vmax (4.2, 4.3, or 4.4V)
ahycka has quit [Ping timeout: 256 seconds]
<Joerg-Neo900>
bq27421*
<Joerg-Neo900>
I couldn't find a nice chip that allows programming that parameter
ahycka has joined #neo900
xman1 has joined #neo900
xman has quit [Ping timeout: 250 seconds]
jonsger has quit [Ping timeout: 256 seconds]
Pali has quit [Ping timeout: 258 seconds]
<wpwrak>
(ir.pdf) hmm yes, our current circuit has the filter operating at about 30 Hz. so the figure on pg 6 is slightly off. well, for that matter is section 7.1. who needs config gpios if you have a glorious silego to play with ;-)
paulk-collins has joined #neo900
<ceene>
in any case i still think 100hz or even 200hz would be better
<ceene>
tvs work at 100hz easily nowadays
<Joerg-Neo900>
LED lamps and FL with electronic shunt work at 40kHz
<ceene>
yes, they're now switched
<wpwrak>
what we really need is a DSP ;-)
<Joerg-Neo900>
TVs are no longer CRT so the frame rate is not related to the "stroboscope" which those TV don't have anymore - unles the FL/LED backlight is PWMed at 40kHz or more
<Joerg-Neo900>
anyway see rationale above
<ceene>
uhm
<ceene>
that's right
<ceene>
a typical monitor screen refresh rate nowadays is 75Hz, isn't it?
<ahycka>
or 60
<Joerg-Neo900>
the design has changed a bit in that it's now a schnitt trigger with autoadjusting threshold, so no filtering at all basically and as long as IR signal level higher than noise signal levem all is just fine
<Joerg-Neo900>
we have no frequency selective receiver
<ceene>
could a high level of noise trigger an interrupt or something like that?
<Joerg-Neo900>
no
<ceene>
well, i'm happy then
<Joerg-Neo900>
you *could* encounter "take over" of console during boot, *IF* you enabled the IR-UART in bq27200 (yes!)
<ceene>
that would be hellish to debug :)
<ceene>
"this thing works when on the living room, but not on the kitchen, where i have fluorescent lights"
<Joerg-Neo900>
hehe, in normal operation UART is off
<ceene>
you've given me an idea
louisdk has quit [Ping timeout: 260 seconds]
<ceene>
a circuit with an light dependent resistor somewhere which makes it only work if inside its cover
<ceene>
but once you open it to see how it's done, it stops working
<ceene>
placing it in the feedback loop of a power supply would be fun
<ceene>
intentional heisenbugs
<ahycka>
stops working??? nooo, a different behaviour xD
<ahycka>
we should do that!
<ahycka>
not in neo, of course :D
<wpwrak>
ahycka: could be a nice feature for some future samsung phone, though
<ceene>
instead of exploding?
<ceene>
that'd be a real breakthrough
<ahycka>
they have already!
<ahycka>
if you open the cover it doesnt explode!
<wpwrak>
Joerg-Neo900: (no filter) we have R5/C1 (ir.pdf page 19) aka R1403/C1403 (schematics), though
<wpwrak>
ahycka: how to prevent your note 7 from exploding on its own ? detonate it from the outside ! :)
<Joerg-Neo900>
>>I still have that switch in my basement. Maybe I'm silly, but I usually keep it set on ‘more magic’.<<
<Joerg-Neo900>
that's how true EE think ;-)
<wpwrak>
funny that they didn't notice the explanation at the end earlier. well, i'm no true EE, maybe that's why this was about the first thing i thought of :)
<Joerg-Neo900>
no, *they* are no true EE
<ceene>
if in 1994 when he did the edit he still had the switch, he could have measured it and test the hypothesis
<Joerg-Neo900>
a true EE instantly had thought of this
<Joerg-Neo900>
ceene: of course
<ahycka>
wpwrak, Joerg-Neo900, i can't find pull up resistor for i2c bus, where are they?
<Joerg-Neo900>
HAH!!!
<Joerg-Neo900>
ahycka: great spotting
* ceene
hides
<Joerg-Neo900>
they *ought* to be next to master which is SoC
<Joerg-Neo900>
we have no SoC yet, so they most likely are on BB-xM
<ahycka>
we can always put as "DO NOT MOUNT"
<Joerg-Neo900>
TODO: check that hypothesis
<Joerg-Neo900>
ahycka: ACK!
<ahycka>
in every bus xD
<Joerg-Neo900>
wpwrak: please add PUs to all I2C
<ahycka>
\o/
<Joerg-Neo900>
add note "TODO: check if already present in BB-xM"
goiken has quit [Ping timeout: 268 seconds]
<Joerg-Neo900>
this been a minor*) but very good find :-) - *) we could add those PUs as wired components next to BB-xM B2B conns to fix that oopsie, if we had missed on finding it now
* Joerg-Neo900
also wonders if the OAMP IO could apply some (too weak, prolly) internal PUs to the I2C pins
<Joerg-Neo900>
OMAP*
<wpwrak>
bb-xm provides PUs on i2c#3 (inside the level shifter, u12), and it has DNP locations for i2c#2 PUs as well
<Joerg-Neo900>
:-D
<Joerg-Neo900>
best parctice to have DNP footprints for all of them on proto_v2
<wpwrak>
for some reason, the omap doesn't seem to be able to PU i2c (sez the manual)
<Joerg-Neo900>
LOL
<Joerg-Neo900>
well, they would be way too weak anyway
<wpwrak>
well, we have room on the extension ...
<Joerg-Neo900>
BB-xM has a LS for I2C???
<Joerg-Neo900>
wow, how weird
<wpwrak>
don't get too excited. internal use ;-)
<Joerg-Neo900>
HAH
<wpwrak>
well, DVI. so you could get it. costs you a dvi connector :)
<Joerg-Neo900>
eeew
<Joerg-Neo900>
nope
<Joerg-Neo900>
~ ahycka++
<Joerg-Neo900>
:-)
<Joerg-Neo900>
ahycka++
<Joerg-Neo900>
~karma ahycka
<infobot>
ahycka has neutral karma
<Joerg-Neo900>
MEH!
<ceene>
lol
<Joerg-Neo900>
~karma ahycka++
<infobot>
ahycka++ has neutral karma
<Joerg-Neo900>
nevermind :-)
<wpwrak>
Joerg-Neo900: suggested nominal value 10 kOhm ? (besides flagging them as DNP. well, we can place the ones for i2c#2, so that we don't need to rework the beagle for this)
<Joerg-Neo900>
sure, why not
<Joerg-Neo900>
changing the value is easy :-)
<wpwrak>
yup :) 10 k is also what the LS on i2c#3 has
* Joerg-Neo900
hands ahycka a huge can of sangria
<ceene>
lol
<ceene>
she enjoys more cerveza
<Joerg-Neo900>
oooh, that's even better, we got the best beer of the world here
<wpwrak>
especially when venturing a little further to the east ;-)
<Joerg-Neo900>
WUT???
<Joerg-Neo900>
nah!
<Joerg-Neo900>
Pilsener is OK, but nothing beats Frankonian beer
<ceene>
it's not like it's the best beer in the world or anything like that
<ceene>
but it's just so nice...
<ceene>
you can drink liters of it :D
<Joerg-Neo900>
hard to find any decent beer in Spain
<Joerg-Neo900>
well, not easy in Germany either, so a tad simpler than in Spain still
<ceene>
i wouldn't say that
<ceene>
but it's true that our beer culture is pretty basic
<Joerg-Neo900>
I wouldn't say there don't exist good beers made in Spain, but ti's even harder to get them than it's hard to get a good beer here in any pub
<ceene>
we enjoy the beer with a few tapas
<ceene>
yes, that's true
* Joerg-Neo900
suddenly feels terribly hungry
<wpwrak>
ahycka: what attribution shall i use in the commit ? (Suggested by ...)
<ahycka>
Ana Martin Ayuso
<wpwrak>
thanks ! updated.
<ceene>
something strange with make view
<ceene>
for some reason, if i do make view
<ceene>
eeshow gui doesn't appear
<ceene>
but instead it's downloading all datasheets
<ahycka>
another! in modem power sheet, chip INA231, i miss all 100nF Cbypass (Vs pin- B1 ball)
<Joerg-Neo900>
ceene: expected
<Joerg-Neo900>
ahycka: :nod:
<ceene>
Joerg-Neo900: first time it didn't do that, it only downloaded things when i clicked on one
<Joerg-Neo900>
it's supposed to build up a cache
<ceene>
but... it's a new behaviour somehow, on first execution it didn't do that
<Joerg-Neo900>
it won't do the complete download when cache exists
<wpwrak>
ceene: normally it should populate the entire cache on the first run. not sure how you avoided that ;-)
<ceene>
strange thing, i didn't do anything to avoid it
<Joerg-Neo900>
wpwrak: (INA231 bypass C) one for all 3 should suffice
<ahycka>
1 por all???
<ahycka>
+for
<ahycka>
*for
<ahycka>
new laptop -.-
<wpwrak>
Joerg-Neo900: so the one at the output of U401 is enough ? (C402)
<ahycka>
ill put 1 for each component
<Joerg-Neo900>
ahycka: sure, they are very close to each other, and not particularly 'surgy' on their power supply
<ahycka>
well...
<ahycka>
let me put if i have space
<Joerg-Neo900>
of course
<Joerg-Neo900>
:-)
<ceene>
she'd rather fight against space than be sorry something doesn't work later
<Joerg-Neo900>
wpwrak: U402 is silego, that one needs one for its own
<wpwrak>
Joerg-Neo900: U401-C402
<ceene>
however, when i want her to put a death star on any board she's always like "it won't fit! i won't put that!"
<ceene>
so meh
<Joerg-Neo900>
haha
<ahycka>
but i put a beer!!!
<ceene>
we're gonna fund Beer Circuits, Inc.
<wpwrak>
we could also add some cat pictures ...
<Joerg-Neo900>
:-D
<ceene>
well, in Spain it's gotta be S.L. or S.A.
<wpwrak>
SL sounds like a lot less bureaucracy :)
<ceene>
a hell lot less of money
<ceene>
SL requires 3K€, SA 60K€
<wpwrak>
don''t SA also have nasty reporting requirements ?
Pali has joined #neo900
<ceene>
yes, they are intended for bigger business
<Joerg-Neo900>
UG needs 1€ ;-D
<Joerg-Neo900>
GmbH needs iirc 50k
<wpwrak>
Joerg-Neo900: so .. is C402 is enough for { U401, U403-5 } ? (U402 already has C403)
<ceene>
they can later on go public in the stock market and all that
<ceene>
an SL cannot, but you basically can do with it whatever you want
<ceene>
just report earnings and expenses
<ceene>
and pay lots of taxes
<Joerg-Neo900>
wow, my brain is totally opposing me (as silly as it sounds) - I thought "25k" (since that's the actually needed 50% of the nominal 50k) but TYPED 50k
<ceene>
lol
<Joerg-Neo900>
wpwrak: I answered that
<ceene>
homer simpson usually has conversations with his own brain, so i guess it's a reasonable thing to happen
<Joerg-Neo900>
[2016-11-29 Tue 20:35:25] <Joerg-Neo900> wpwrak: U402 is silego, that one needs one for its own
<wpwrak>
Joerg-Neo900: but we're talking about U401 :)
<Joerg-Neo900>
wpwrak: ahycka: ceene: we have a power plane I suppose, which is supposed to be very 'hard' so bypass caps are not *that* vital usually. But we for sure need them on 'high power' logical circuits like Silego
<ceene>
we've got bad experience with very simple chips, even with a power plane
<wpwrak>
ceene: and isn't SAS even easier than SL ? afaik SAS is like the uk ltd. quick and cheap to set up.
<Joerg-Neo900>
generally you can 'share' nearby buffer caps unless the datasheet of a chip is explicit about placing the buffer as close to the chip footprint as possible
<ceene>
wpwrak: that doesn't exist in Spain
<Joerg-Neo900>
it doesn't hurt to have proper bypass everywhere and just successively ommit them when we run out of psace. subject to layouter's discretion, in coordination with EE
<Joerg-Neo900>
since that's very clearly a layout issue, not a schematics design one
<ceene>
sounds right, if space is limited
<ahycka>
IINA datasheet- page 10: The figure on the front page shows a typical application circuit for the INA231. For power-supply bypassing, use
<ahycka>
a 0.1-μF ceramic capacitor placed as close as possible to the supply and ground pins.
<ceene>
one has to sacrifice something
<Joerg-Neo900>
ahycka: so you have earned a second point :-D
<ahycka>
so, pliiiiiis....
<ahycka>
xD
<ahycka>
2 beers!
<ceene>
lol
<ahycka>
perfect!
* wpwrak
adds more bypass caps
<ahycka>
\o/
<Joerg-Neo900>
ahycka: kudos for your approach to 'layout', I wish all layouters I had to deal with in my life would have that expertise and notion about what means 'layout'
<Joerg-Neo900>
:-))
<ahycka>
Joerg-Neo900, im a full EE
<Joerg-Neo900>
hard to think otherwise :-)
<ahycka>
i do: design, schematics, layout and component choice of my boards
<ahycka>
but i have no time to do this for new
<Joerg-Neo900>
I don't want to start rant about OpenMoko here, but...
<wpwrak>
ahycka: you know nothing of the horrors we've seen (at openmoko) ;-)
<ahycka>
layout is the fastest part (for me)
<ceene>
and i critize her work and from time to time, also provide insightful comments
<ahycka>
openmoko?
<Joerg-Neo900>
~openmoko
<infobot>
it has been said that openmoko is a joint venture of FIC and OSC, attempting to create open source phone platform stack. see http://openmoko.com It's known for failing miserably, due to overpriced devices: http://www.handheld-linux.com/wiki.php?page=GTA04 and zealot'ishly following ideology, without respect for practial usage (like re-using gta01/gta02 keyboard-less form factor, poor/small screen, etc).
<ceene>
ahycka: that was a previous phone making attempt
<Joerg-Neo900>
wtf?
<Joerg-Neo900>
~factinfo openmoko
<infobot>
error: you do not have enough flags for that. (o required)
<infobot>
openmoko -- last modified at Sun Jun 3 21:37:48 2012 by Estel_!~Estel@dgf46.neoplus.adsl.tpnet.pl; it has been requested 15 times, last by Joerg-Neo900, 39s ago.
<Joerg-Neo900>
ESTEL!!!!
<Joerg-Neo900>
sucker
<Joerg-Neo900>
worst troll ever
<ceene>
moco in spanish means booger
<Joerg-Neo900>
I know :-P
<ceene>
lol
<Joerg-Neo900>
we have more funny stuff: Nokia Lumia
<ahycka>
Lumia = bitch :P
<ahycka>
we know that :D
<ceene>
there was also mitsubishi pajero
<Joerg-Neo900>
and maemo NOLO - which means sth like "mean" in Borwegian(?)
<ceene>
which means wanker
<Joerg-Neo900>
Norwegian*
<Joerg-Neo900>
the OM bootloader was "dirtymoko" X-P
<Joerg-Neo900>
originally Sean Moss-Pultz (and Harald Welte?) minted OpenMoko as a chimera english-german of "Open" "Mobile" and "german "Kommunikation"
<Joerg-Neo900>
afaik
<Joerg-Neo900>
I wasn't even aware of the project at that time
<ceene>
harald has been remembering lately the anniversay of openmoko on his blog
<Joerg-Neo900>
yep. I think I posted that in here a few days ago
<Joerg-Neo900>
:-)
<ceene>
i have it on the rss, so it sometimes pops up
* Joerg-Neo900
still searches the cruise missile with target preprogrammed "@Estel"
<ceene>
lol
<ceene>
what happened?
<ceene>
wasn't he doing an aluminum case for n900?
<ceene>
or something like that
<Joerg-Neo900>
<infobot> openmoko -- last modified at Sun Jun 3 21:37:48 2012 by Estel_!~Estel@dgf46.neoplus.adsl.tpnet.pl; it has been requested 15 times, last by Joerg-Neo900, 39s ago.
<Joerg-Neo900>
yes
<Joerg-Neo900>
at least he claimed to
<Joerg-Neo900>
prolly led nowhere, since it fscks up all RF design
<Joerg-Neo900>
also the N900 case is pretty complex
<ceene>
i'd like a bigger version of the neo900, more like a tablet
<Joerg-Neo900>
he also never understood the difference between Openmoko Inc (GTA01/02), Golden Delicious (Nikolaus, GTA04) and Neo900
<Joerg-Neo900>
the plan how to get there is: make Neo900 a success
<ceene>
yep, without that, no tablet, no step2 and no nothing
<Joerg-Neo900>
exactly
<ceene>
lets see if we can make this :)
<Joerg-Neo900>
:-))
<ceene>
well, time to have dinner!
<ahycka>
move the cat!
<ceene>
cat over table
<ceene>
i see myself eating on the floor
<ceene>
gonna risk my hands
<ahycka>
or eating by a cat xD
<ceene>
if i can't type tomorrow, ahycka will tell you so
<ceene>
cya!
<ahycka>
cya tomorrow!
ahycka has quit [Quit: Leaving]
<Joerg-Neo900>
hehe
<Joerg-Neo900>
time for dinner here too
<Joerg-Neo900>
~openmoko
<infobot>
i heard openmoko is see ~wiki openmoko
<Joerg-Neo900>
~wiki openmoko
<infobot>
At https://en.wikipedia.org/wiki/Openmoko (URL), Wikipedia explains: "{{For|the mobile telephone operating system|Openmoko Linux}} {{Merge from|Neo 1973|Neo FreeRunner|date=August 2015}} , an open source smartphone created by the Openmoko project.]] 'Openmoko' was a project to create a family of open source mobile phones, including the hardware specification, the operating system (Openmoko Linux), and actual smartphone development implementation like ...
<wpwrak>
herd instinct threatens to set in, dragging me towards food ...
jonsger has joined #neo900
<Joerg-Neo900>
wpwrak: please change U303:B1 BIN from BATID to to 10kR PD
<Joerg-Neo900>
wpwrak: rationale: we might not have a thermistor on battery BSI
<Joerg-Neo900>
even worse: we might have HDQ data on BSI/BATID
jonwil has joined #neo900
qws-user-1228 has quit [Read error: Connection reset by peer]
qws-user-1228 has joined #neo900
<ravelo>
ceene, awesome work
<ravelo>
sounds very nice to me at least :)
ds2 has joined #neo900
paulk-collins has quit [Quit: Leaving]
<Joerg-Neo900>
((difference between Openmoko Inc (GTA01/02), Golden Delicious (Nikolaus, GTA04) )) even wikipedia got that quite wrong
<Joerg-Neo900>
GTA04 (GDC) in no way whatsoever based on OM's GTA04, it simply reused the GTA03 case and display, and battery
<Joerg-Neo900>
err, s/OM's GTA04/OM's GTA03/
<Joerg-Neo900>
OM's GTA04/03/04/03 (chronological renaming history) been based on S3C64000
<Joerg-Neo900>
I'd be surprised to find a single identical chip between OM's GTA03 and GDC's GTA04 (except the embedded display controller)
<Joerg-Neo900>
and of course the BQ27000 in battery
<wpwrak>
do chip resistors count ? :)
<Joerg-Neo900>
hehe
<Joerg-Neo900>
I wonder if anybody wants to fix that nonsense on wikipedia
<wpwrak>
luckily, we all are protected by the commandments of wikipedia from people with first-hand knowledge messing with this ;-)
<Joerg-Neo900>
I ahve an account, so... but meh!
<Joerg-Neo900>
wpwrak: indeed, seems to be sadly true
<Joerg-Neo900>
I *never* wrapped my head around THAT nonsense
<Joerg-Neo900>
"sorry, Einstein. Where's the newspaper article explaining your theory? you MUST NOT write first and info into wikipedia!"
<Joerg-Neo900>
first Hand*
<wpwrak>
i can understand part of it. e.g., we'd get an endless flood of blatant advertizing otherwise. but they seem to have forgotten to define a channel for fixing things if you actually are the relevant authority.
<wpwrak>
of course, then people could even lie then. e.g., about their own biography. so .. :)
<Joerg-Neo900>
"as Joerg stated [1]: >>GTA04 (GDC) in no way whatsoever based on OM's GTA04, it simply reused the GTA03 case and display, and battery<< ... [1]https://irclog.whitequark.org/neo900/2016-11-29#18296710;<< -- edited by Joerg ;-P
<Joerg-Neo900>
s/GTA03/GTA02/
<Joerg-Neo900>
I also didn't know (according to wikipedia) that Openmoko first was about software and only much later came up with 1973 idea :-o
<Joerg-Neo900>
I would have tought, based on the gossip about Sean being room mate of a FIC nephew or whatever, that focus initially been on hardware
<Joerg-Neo900>
thought*
<Joerg-Neo900>
wpwrak: you were 'foundation member' of OM?
<Joerg-Neo900>
I knew of Michael Lauer, Sean Moss-Pultz and Harald Welte.
<Joerg-Neo900>
the former probably joining in some weeks later
<Joerg-Neo900>
>>The phone released was inconsistently-named Neo 1973 or Neo1973<< errrr
* luke-jr
grumbles at Pyra using a non-free license
<luke-jr>
yeah, Wikipedia is nonsense
<luke-jr>
"what the mass media propagandizes" rather than "what is actually true"
<Joerg-Neo900>
HAHA, you could refer to skunkworks GTA04 (the 'sekrit' project Sean assigned to me): ... :-/ _-S if only I could find where the heck I published it