jonwil has quit [Remote host closed the connection]
<silviof>
Hey DocScrutinizer05 is the project still at "risk-buy" state? (referring: http://neo900.org/#donate)
<DocScrutinizer05>
errr, I don't know how that state is defined, but yes, I'm about to start a decent webshop where you can buy hard-to-source components that will allow Neo900 UG to build a device for you
<silviof>
Its a very often asked question?
<DocScrutinizer05>
e.g. we found 1GB RAM chips and we hope for a first 70 samples to arrive on monday so we could test them. After they been verified to work as expected, users could buy them and we'll use them in their Neo900 PCB
<freemangordon>
DocScrutinizer05: ok, what about htose that donated more than 100 euros? do they have to buy those risky parts as well?
<DocScrutinizer05>
no, the qiestion is actually new to me, that's why I got a bit clueless about how to answer it
<DocScrutinizer05>
freemangordon: the projected procedure is: wpwrak and me prepare a list of risk parts and a list of mech kit (N900) parts, and we estimate how much each set will cost. Then customers can opt in for each on their customer account page in webshop, where they also will see their balance
<silviof>
okay cool procedere
<freemangordon>
iiuc, as long as your donation covers what you've opted in so far, there is no need for a new donation/purchase, correct?
<DocScrutinizer05>
freemangordon: but it's actually a very "interesting" question since we so far counted the "excess" "donations" in for R&D and we're still short on funds
<DocScrutinizer05>
freemangordon: when the donation would cover the risk parts purchase, what would we use to pay R&D from?
<freemangordon>
ah, I see
<DocScrutinizer05>
I probably will open a regular preorder, with a minimum down payment of several 100 eur
<DocScrutinizer05>
everything else gets too confusing
<freemangordon>
but still, if one has donated 500 euros lets say, does she needs to make purchase of those risky parts?
<DocScrutinizer05>
when I decide let's say 500 EUR is the minimum down payment, then those who donated more than 500 will not need to
<freemangordon>
ok
<freemangordon>
that is what I wanted to know :)
<DocScrutinizer05>
as a general rule, not a single of our supporters will suffer from giving early support
<DocScrutinizer05>
you'll always benefit
<DocScrutinizer05>
the devices will not get cheaper, and you will not have to pay twice for anything
<DocScrutinizer05>
late joiners are the ones who will pay more
<freemangordon>
see, my mentality have some problems with filling forms, thats why I tend to try to avoid such :D
<freemangordon>
so my questions are rather - "is it possible to optimize the number of future payments I have to do"
<DocScrutinizer05>
I understand the "filling forms" issue, I suffer from that disease as well. But what you get on yozur customer page is a multiple choice checkmark thing to select if you want a NeoN board or a complete device, which case color, which keymat and what flavor of modem. And there are reasonable defaults for each option. And you can see your account balance and if you have credit or debit
<freemangordon>
ok
<DocScrutinizer05>
((optimize)) see what Oksana did ;-)
<freemangordon>
seems sane
<freemangordon>
where?
<DocScrutinizer05>
she did two payments and already owns 6 devices ;-D
<freemangordon>
oh :D
<silviof>
^^
<DocScrutinizer05>
generally your payment never is lost, worst case we transfer it back to you and you don't lose a cent
<freemangordon>
ok, ok, got it
<DocScrutinizer05>
(unless I get hit by a bus and werner and nik don't feel like accepting the heritage)
<DocScrutinizer05>
for now you bought some fine studies and papers on those studies, plus schematics and block diagram, plus a thoroughly evaluated hw concept
<DocScrutinizer05>
aka R&D
<DocScrutinizer05>
now we enter the stage of project where you buy material goods (components etc)
<DocScrutinizer05>
you own them even when they stay at factory where the devices get built. You can request shipping to your location of all the components and immaterial goods you already bought
<DocScrutinizer05>
any time - but beware! we cannot use components for production that left the factory
<silviof>
In the webshop is there a overview what I Have bought?
<DocScrutinizer05>
there will be, yes
<silviof>
okay -fine.
<DocScrutinizer05>
actually internally we so far handled the finances even differently, in a sense of our customers giving us credit which we pay back to them (incl interest) the very moment they order some hw at neo900
<DocScrutinizer05>
this concept will change for the hardware parts, probably. It's all very tricky regarding german tax obligations etc
<DocScrutinizer05>
when we don't send out goods in parcels, german tax authorities assume it been an inner-german purchase and you need to pay 19% VAT, even when you're sitting in USA or anywhere else outside EU
<DocScrutinizer05>
a nightmare
<silviof>
hehe - jup!
<DocScrutinizer05>
for credits nobody pays VAT, that's the beauty of this concept that we deal with VAT only when we ship the final devices
<DocScrutinizer05>
and thus you all gave Neo900 UG a credit for now, payback date being the moment you order your device
<silviof>
I hope you have discussed this with a tax-manager? :-P
<DocScrutinizer05>
so if device costs X and you gave 500 credit, your invoice will be "X - 500"
<silviof>
... or lawyer ^^
<DocScrutinizer05>
with 4 of them already
<silviof>
haah - okay :-)
<DocScrutinizer05>
and it's finally approved
<DocScrutinizer05>
see "fineprint" on neo900.org/donation
<DocScrutinizer05>
in case of fatalities we push a button and commit all the parts pruchases
<DocScrutinizer05>
so the parts are yours in that moment.
<silviof>
another thing is, I have dropped 105 Euro to diliciouseokay - i got it :-)
<silviof>
sorry - damn buffer :-)
<DocScrutinizer05>
you already paid a 13(?)% of that amount to R&D, but that shouldn't bother you, Neo900 UG counts this as a credit of 105EUR to Neo900 UG and a payment to GDC we did for R&D
<silviof>
another question, I have buyed 110 euro at deliciouse, but they have transfered only 92.44 to neo900UG, what is the sum I had on my konto now?
<silviof>
ah
<silviof>
okay
<silviof>
thx
<silviof>
your answered faster than I have writen this.
<DocScrutinizer05>
hehe
<DocScrutinizer05>
that#s because we needed the answers before we came up with the stuff causing the questions at your side ;-)
<DocScrutinizer05>
we're thinking about how to handle all this very very carefully, before we actually do anything
<silviof>
I know, thats a open(s/h) project - with some kickstarter influences
<DocScrutinizer05>
since, even when all our customers are perfectly fine and content with the way we handle stuff, we still could do some time in jail thanks to german laws not sufficiently considered in our business model
<DocScrutinizer05>
so we need all this to be 110% legit and correct
<DocScrutinizer05>
in USA you can sue some company for not warning about "don't dry your cat in this oven", in Germany you can sue a company only for having a typo in impressum
<DocScrutinizer05>
and despite what some fools blamed me for, I do NOT do any tax fraud or whatever. This all is 110% checked and safe
<silviof>
DocScrutinizer05: bad
<DocScrutinizer05>
tax authorities will probably love my company ;-)
<silviof>
^^
<DocScrutinizer05>
on the bright side: looking forward to holding 1GB chips in our hands on monday
<silviof>
yeeaa!
<DocScrutinizer05>
added humidity sensor
<DocScrutinizer05>
changed accel, gyro, magneto to a single smart BNO055 chip
<DocScrutinizer05>
we will have 2 BB-xM reworked to employ the 1GB RAM PoP, to test the chips and also already start checking foss-fremantle if it can deal with such amount of RAM
<DocScrutinizer05>
(if some of the FPTF guys are willing to give it a try)
<silviof>
FPTF?
<DocScrutinizer05>
schematics for proto_v2 (to connect to those (1GB) beagleboards) schould be finalized this year already
* infobot
grabs a pen, screams like she's possessed, and begins chasing silviof
<silviof>
boaaaaahhh
* silviof
ducks and run
<DocScrutinizer05>
~gender
<infobot>
I'm hermaphroditic
<silviof>
^^
<DocScrutinizer05>
she's also schizoid
<DocScrutinizer05>
~gender
<infobot>
I'm pregnant
<DocScrutinizer05>
in 140k factoids there's a goodie for everyone ;-)
<DocScrutinizer05>
~+status
<infobot>
Since Wed Dec 10 13:01:39 2014, there have been 6 modifications, 62 questions, 0 dunnos, 0 morons and 16 commands. I have been awake for 2d 19h 46m 38s this session, and currently reference 119367 factoids. I'm using about 85644 kB of memory. With 0 active forks. Process time user/system 486.61/8.66 child 0/0.01
<DocScrutinizer05>
oops 119k
<silviof>
ah that fptf thing let me ask some other thing (Maybe wroing channel) what is the best way to get a debian on my n900?
<DocScrutinizer05>
40 times the number of factoids than the average person has active vocabulary
<DocScrutinizer05>
and that's without ~dict aka miriamwebster
<silviof>
^^
<silviof>
good bo--- ehm goodgirl!
<DocScrutinizer05>
yes, that's the right question in wrong channel, but let me give a terse answer: fremantle *is* debian, mostly. For 100% pure debian you prolly want to use easydeb
<silviof>
DocScrutinizer05: thx. the bsp of neo900 is debian or freemantle?
<silviof>
s/is/would be/
<DocScrutinizer05>
which is a changeroot so you can keep the device's full functionality while running your pure debian userland inside the chroot
<DocScrutinizer05>
the bsp is debian
<DocScrutinizer05>
based
<DocScrutinizer05>
a subset of
<silviof>
okay.
<kerio>
what's a bsp
<DocScrutinizer05>
~bsp
<infobot>
[bsp] bug squashing party, in #debian-bugs
<DocScrutinizer05>
blaergh
<DocScrutinizer05>
~wiki bsp
<kerio>
lol
<infobot>
At http://en.wikipedia.org/wiki/Bsp (URL), Wikipedia explains: "{{Wiktionary|BSP}} 'BSP' may refer to: {{TOC right}} * Bell System Practices, technical documentation series published internally by the AT&T Bell System * Business service provider, a company offering business applications over the Web * Billing and Settlement Plan, an accounting system for airlines * Binary space partitioning ** BSP (file format), used in games such as the Quake series * ...
<kerio>
board support package
<DocScrutinizer05>
g'ammit
<kerio>
i assume it's going to be debian stable
<DocScrutinizer05>
what kerio said
<kerio>
cuz sid has systemd
<DocScrutinizer05>
you prolly can bet on that
<silviof>
oh no - please no systemd blaming (or whoreshipping) here
<DocScrutinizer05>
though stuff like system init actually isn't part of a BSP. Actually a BSP only consists of the drivers and example code needed to make special aspects on the hw platform work
<silviof>
and mybe a shell?
<DocScrutinizer05>
should be pretty much portable. and when it comes in a working system so you actually can run the example code right away, then that system can be anything that's compatible to the examples
<DocScrutinizer05>
wich translates to "debian" but doesn't assume a special init
<kerio>
isn't it basically just a bare-bones OS to test if the hardware's working
<DocScrutinizer05>
I cn promise that the BSP parts will run with whatever init you choose to use
<kerio>
so the manifacturer can go "here's the thing, we tested this, there's the drivers, everything works, bye!"
<kerio>
and then the graphics drivers are for linux 2.1
<DocScrutinizer05>
actually not exactly, it's a set of info and example code and drivers that are needed to make the hardware work
<DocScrutinizer05>
the check thing is a factory test system
<DocScrutinizer05>
we will probably unify/merge the two though
<DocScrutinizer05>
kerio: yes, basically that would be a valid approach, though you could already prepare for shitstorm
<DocScrutinizer05>
the drivers have to be in sourcecode that's sufficiently commented to enable porting to 3.20
<DocScrutinizer05>
but it's not mandatory that they are already ported
<DocScrutinizer05>
actually the drivers not even need to be fully functional, they just need to provide a PoC
<DocScrutinizer05>
worst case a shellscript demonstrating the correct communication to a chip would suffice
<DocScrutinizer05>
sort of charge12.sh
<DocScrutinizer05>
however when hw manuf claims charge12.sh is a comprehensive BSP package for bq24150 and battery maintenance, then the manuf has to be damn sure it#s not lacking any safety measures like checks for overtemperature etc. Otherwise users can sue the manuf for any damage caused by implementations based on the flawed PoC
<DocScrutinizer05>
luckily Neo900 will have a much smarter batt maintenance chip that doesn't need *any* driver basically
<DocScrutinizer05>
so complete "BME for Neo900" will only deal with info gathering, since the chip doesn't need any control by software
<DocScrutinizer05>
(except for one particular detail)
<DocScrutinizer05>
and that detail could result in increased aging of battery cell when it's getting ignored completely, but we will take care about and not ignore it
<kerio>
will neo900 have fully hardware charging?
paulk-collins has joined #neo900
<DocScrutinizer05>
to be a tad less obscure, lemme elaborate: the charger chip on power-on-reset defaults to max battery charging current ~2A which is too much for gentle treatment of battery. However we will limit this indirectly by limiting max USB current to 1A and reset that chip config to a sane 850mA during bootloader startup. It stays sticky in chip until no-power aka battery removed
<DocScrutinizer05>
yes, Neo900 resp bq24297 can do completely hw-controlled charging
<DocScrutinizer05>
NB that USB curent limit is set by an easily changed resistor, and you could charge 3000mA mugen DIY batt with decent 2A then
<DocScrutinizer05>
I think for the hacker friendly device that's Neo900 this is an excellent solution
<kerio>
is 1A that bad for a standard bl-5j?
<kerio>
also, this is the current when the battery is near empty, right
<DocScrutinizer05>
1A is @5V, @3V that's 1*5/3 * efficiency then
<DocScrutinizer05>
so roundabout 1.5A for empty battery, which is a tad on the high side
<kerio>
mh k
<DocScrutinizer05>
for standard BL-5J that's around 1.2C while it should get charged at 0.7C
<kerio>
i still don't like it that much
<kerio>
software can and will fuck up, eventually
<kerio>
especially software that can be changed by the user
<DocScrutinizer05>
but it won't explode when gettinmg charged by 1.2C, particularly it will not suffer at all from an occasional 5s of 1.2C charge current
<DocScrutinizer05>
kerio: you can overvolt the core to insane ampounts, just via software
<kerio>
feature request: liquid cooling
<DocScrutinizer05>
there's stuff in every embedded device or even every system at large that is much more worrisome than a charging battery by 1.2C instead of the recommended 0.7C
<DocScrutinizer05>
actually Neo900 is much less prone to sw induced hw defects than N900
<DocScrutinizer05>
and actually users may deliberately decide to quickqharge their battery with 1.2C and not care about the expected livespan getting reduced from 1000 cycles to 200 cycles by that
<DocScrutinizer05>
not our call to stop them from doing so
<DocScrutinizer05>
you probably could even crank up to >2A charging current when you change the USB current limit by replacing that resistor
<DocScrutinizer05>
the cell usually isn't made for quickcharging and will suffer massively, but there might be valid usecases. E.G. when you opted for a better battery
<DocScrutinizer05>
there *are* LiIon cells that allow 5C or even 10C for charging current
<DocScrutinizer05>
ask the RC freaks
<DocScrutinizer05>
they got awesome cells in their helicopters and racing cars
<DocScrutinizer05>
they charge a completely depleted battery pack in 5 minutes
<DocScrutinizer05>
well, maybe 10, dunno
<DocScrutinizer05>
there been tiny electric RC cars like 30mm size, which got charged in 60s to run for a 5 min on one charge
<kerio>
feature request: helicopter rotors
<kerio>
in the stylus slot
<kerio>
would you expect batteries that can charge faster to be bigger as well?
<DocScrutinizer05>
no, they usually trade off capacity for max current they can handle
<DocScrutinizer05>
the higher the current the thicker the electrodes
<DocScrutinizer05>
so yes, in volume they are bigger on same capacity
<DocScrutinizer05>
;-)
<DocScrutinizer05>
or on capaicty lower when same volume
<kerio>
k
<DocScrutinizer05>
((heli rotors)) somebody already had similar request for Jolla TOH
<DocScrutinizer05>
so I will wait for them doing that, I don't want to plagiate
<DocScrutinizer05>
kerio: ((will neo900 have fully hardware charging?)) nor only that, it even can power up without any battery inserted (though probably you will run into problems with GSM, due the the spikes in current consumption that uses particularly during sign-in resp each connection setup)
<kerio>
jeat
<kerio>
*neat
<DocScrutinizer05>
no battery also includes defect or completely depleted battery
<DocScrutinizer05>
so no more bootloop deadlock
<DocScrutinizer05>
(not that we would need system to be up and running to charge battery to full)
<DocScrutinizer05>
((problems with GSM)) the modem allows that it comes up in a offline status which will work for sure, and it also allows restricting used bands to e.g. UMTS only, which doesn't have such nasty surges in current consumption like GSM does
<DocScrutinizer05>
alas I only have a ¿ and the afaik more commonly used ¡
<DocScrutinizer05>
HEHE >>This article contains special characters. Without proper rendering support, you may see question marks, boxes, or other symbols.<<
<DocScrutinizer05>
ironically that one exists in my fontset, unlike the irony tag
<DocScrutinizer05>
err Irony_punctuation
<DocScrutinizer05>
and the correct spelling is interrobang
<Pali>
DocScrutinizer05: linux virtual mem != swap
<Pali>
in virtual mem is also included size of all mmaped files
<DocScrutinizer05>
yes, I know
<Pali>
so if you call mmap on big HD movie, then you will have big virtual mem usage
<Pali>
but mmap will not copy all file content to ram immediately
<Pali>
but but 5GB is really big for systemd!
<DocScrutinizer05>
anyway some process must eat those 10GB of ram and swap, and I'm pretty sure systemd_journald is one of them
<Pali>
it means you will not be able to run it on 32bit system
<DocScrutinizer05>
killing and restarting kded also helped a dang lot, it freed up a ~8GB as well when it segfaulted due to "out of memory"
<Pali>
kded4 had some bug in power management code
<DocScrutinizer05>
uhuh
<Pali>
it cause that for every acpi event there was memleak
<DocScrutinizer05>
well, whatever counts as acpi event
<Pali>
and if acpi send "battery % changed" too often, it can eat even 1GB
<DocScrutinizer05>
desktop PC here, no battery involved
<Pali>
it does not mean that there is no other memleak :D
<DocScrutinizer05>
maybe every other day it blanks the screen
<Pali>
which kde version do you have?
<DocScrutinizer05>
whole damn KDE4 is a single mem sink
<Pali>
$ kde4-config --version
<DocScrutinizer05>
4.11.5
<Pali>
I think that problem was with older kde versions...
<DocScrutinizer05>
well, system monitor (KDE) shows processes that are long vanished and forgotten.
<DocScrutinizer05>
tells a story about system monitor as well as whole KDE
<DocScrutinizer05>
zooming in in konqueror on http://acanohayluz.com.ar/ map caused sth in konqueror to allocate some 5GB, I.E all that been available, and hog all 4 cores at 100% system and thus freeze the complete KDE, until the konqueror process segfaulted and freed up the CPUs and the RAM again
<Pali>
do you know any other web browser which will not eat too many RAM? :d
<DocScrutinizer05>
not like 5GB in some 15..20s
<kerio>
meanwhile, kernel_task is using 1.27GB of ram here
<kerio>
after 14 days of uptime
<kerio>
launchd is using 18mb
<DocScrutinizer05>
I've seen this RAM hogging in konqueror a few times during last 9 months. Meanwhile I know that first thing to do is switch to console (alt-ctrl-F1) or I won't get there anymore since any other key event clogs the queue
* DocScrutinizer05
probably should enable systemkeys
<DocScrutinizer05>
though I wonder if those would really help me out in such situation
<kerio>
go nuclear
<DocScrutinizer05>
I at least could force the system to a graceful reboot
<kerio>
sysrq+e
<DocScrutinizer05>
a few weeks I was desperate and pissed enough to seriously consider going SLED, but then I learned they don't support kmail (kde?) anymore and also seem to already have adopted systemd
<Pali>
or sysrq+f
illwieckz has quit [Quit: Ex-Chat]
<kerio>
DocScrutinizer05: install gentoo :3
mvaenskae has joined #neo900
<DocScrutinizer05>
soon
<kerio>
or maybe funtoo
<DocScrutinizer05>
devuan
<mvaenskae>
too soon?
<mvaenskae>
cheers :)
<mvaenskae>
devuan, funtoo.... what's the topic? :)
<DocScrutinizer05>
KDE and systemd fsckup
<mvaenskae>
oh, the usual
<DocScrutinizer05>
lol, yeah
illwieckz has joined #neo900
<kerio>
the usual topic, systemd rants and poettering physical elimination plans