Daanct12 has quit [Remote host closed the connection]
<norayr>
uvos, thank you. let me show you the screenshot about zram.
<norayr>
so i hope if i backup the home, then back it up back to the newly installed system, then reboot, i'll get my wifi data back?
<norayr>
hmm, i cannot make a screenshot. every time i was powering droid on, the browser would appear with the message about update information
<norayr>
and zram part in it.
<norayr>
so when booted leste on droid4 last several times, i got that browser window with recommendation to do with rc-update del zram atd then on the same line, rm /etc/init.d/zram. but wouldn't it just disable tce service?
<norayr>
but before that there was a suggestion to edit zram configuration file.
noidea_ has quit [Remote host closed the connection]
Daanct12 has quit [Quit: Leaving]
ceene has quit [Remote host closed the connection]
noidea_ has joined #maemo-leste
xmn has joined #maemo-leste
<sicelo>
Wizzup: uvos, he is referring to the salutem message that appeared after upgrade. I didn't read it myself, but either norayr got confused by it, or it's confusing to begin with. Someone should review it
<Wizzup>
want to be that person? :)
<buZz>
rsync: write failed on "/home/amprolla/images/droid4/20220417/maemo-leste-1.0-armhf-droid4-20220417.img.xz": No space left on device (28)
<buZz>
i wondered why droid4 image was lagging two weeks
<buZz>
:D
<Wizzup>
yeah. the vm is constantly out of space.
<Wizzup>
we need to migrate the infra to another machine IMHO
<buZz>
need a disk?
<buZz>
ah
<Wizzup>
it's not our infra
<Wizzup>
it's maemo.org infra
<Wizzup>
the I/O is also really really slow
<Wizzup>
so it's better to move to some of our own stuff
<buZz>
hmhm
<buZz>
[35m [00m [01;37m . [00m [22;37mBacking up the entire Internet... [00m
<Wizzup>
but it takes time to do it
<buZz>
ah, that step might take a lot of space
<buZz>
:)
<Wizzup>
and we need to ask the maemo folk to change dns
<buZz>
hmmf, first boot of the 20220410 image, and maemo showed me the 'unlock swiper' once, and then hung and now just rebooted? hmm
<buZz>
i'll leave it on power for a while, hope i can get this fabled 'two day standby'
<buZz>
:P
<buZz>
yeah wtf, battery is 3.2V and leste decides to shut down instantly, instead of going to that 'charging mode'
<buZz>
:(
<buZz>
hmmm yep, its -never- hitting any charging
<bencoh>
I think I have a similar issue, but I'd need to investigate here
<buZz>
omg, it hit 3.3V and continued after ~6 bootcycles
<buZz>
ffs
<buZz>
:P
<bencoh>
(right now mine might be completely discharged actually, I didn't manage to boot it last I tried, I'll have t otry aain with a wallcharger)
<buZz>
wherever that threshold it, i wanna lower it to 3.0 or maybe 2.8
<buZz>
bencoh: do you have a usb condom?
<buZz>
(only power pins connected through, not data)
<buZz>
there seems to be two results to connecting USB , one is 'get some kinda charge mode' the other is 'haha just boot!'
<buZz>
usb condom can get the second type to change into first type
<bencoh>
I think my wallcharger shorts D+/D- to the ground yeah, I'll have to check
<buZz>
i wonder if other mapphones behave similar on that
<buZz>
and pop, it just rebooted
<buZz>
-_-
* buZz
assumes it just went under X volt and decided to powerdown, then on 'wrong charger' making it instaboot again
<buZz>
booted up to hildon, then turned on 'the white LED of powerdown'
<buZz>
lol
<sicelo>
Yes ,d+/- short is also effective on droid4
<buZz>
sicelo: for which of the two modes i described?
<buZz>
the latter, right?
<sicelo>
But yeah, d4 boot can be heavily influenced by alignment of the planets
<sicelo>
buZz: which two?
<buZz>
15:29:24 < buZz> there seems to be two results to connecting USB , one is 'get some kinda charge mode' the other is 'haha just boot!'
<sicelo>
I've never seen droid4 wait to charge ... i only always see it boot on usb connected
<buZz>
sicelo: well, i just described how you can see these two different things
<buZz>
read up ;)
<sicelo>
I read. I'm saying I've never seen your first scenario. Or i don't understand it. What do you mean by 'some kind of charge mode'?
<buZz>
this mode is ; droid 4 is powered off, you connect USB and you -only- get a green led for charging
<sicelo>
Never seen it
<buZz>
sicelo: i described how you can get there
<buZz>
after a while it gets into another charging mode where the green led only flashes ~50ms every ~1s
<buZz>
might be some trickle charge, or just a bug
<buZz>
sicelo: TLDR ; 15:29:35 < buZz> usb condom can get the second type to change into first type
<buZz>
15:28:51 < buZz> bencoh: do you have a usb condom?
<buZz>
15:29:00 < buZz> (only power pins connected through, not data)
<buZz>
nothing shorted
<sicelo>
I don't have a 'condom 'so maybe that's why I've never seen it.
<buZz>
hmhm ,well try different usb powersupplies? some seem to hit 'mode1' , others 'mode2'
<buZz>
i am not so sure 'mode2' (the booting one) even enables the charger
<sicelo>
Does it get into that mode reliably? I can go through the trouble of making a condom if it's really reliable, otherwise not worth it :-)
<buZz>
yes, both modes are 100% reliable to hit
<buZz>
and i'm kinda sure actually -booting- in 'mode1' makes it 100% unaware any usb powersource is connected
<buZz>
sicelo: i cant say its a reliable way to charge!
<sicelo>
I'll make it then.
<buZz>
i've been having endless troubles getting a 'empty' droid4 to start back up into a charging setting
<buZz>
sicelo: i grabbed the 'newest' <2euro 230V-USB psu from action.com
<buZz>
this one does 'mode1' all the time
<buZz>
and i have the ~3yo model of the <2eu 230V-USB psu from them , which does 'mode2' all the time (and if i plonk a usb condom between, it hits 'mode1' all the time'
<sicelo>
For a quick test, I can cut the data tracks on my power bank that is able to reliably start droid4 in whay you describe as mode2
<sicelo>
This power bank has shorted data pins
<buZz>
the 'no data pins' mode is mode2
<buZz>
maybe i should start a wikipage about this?
<buZz>
or a github issue, hmm
<buZz>
i really wish we could get kexecboot to just 'hang' and switch to a charging mode when <x voltage
<buZz>
now booting in 'mode1' , leste detects power on the usb port
<buZz>
(i just got my datasim, trying to test its functionality
<buZz>
eh, leste -did- detect power, at least, green led was on as like charging, BUT, it just randomly rebooted while in hildon and hildon didnt show the charging status
<buZz>
man, what a dance :P
<buZz>
i'm -guessing- the simcard makes the modem startup which draws so much current that leste murders itself :P
<buZz>
maybe i should just dig up a random 'li+ charger' and hook it to the battery terminals
<sicelo>
I'm confused now. I said on d+/- shorted charger, d4 boots up always for me. What I'd like is emergency charging where it doesn't boot. I've never seen this (is cpcap even able to regulate charging without software control?)
<buZz>
i see droid4 now in 'mode1' doing 50mA draw
<buZz>
i dont believe those 50mA are -just- the green led it turned on :P
<buZz>
but, well, very tempted to just hookup 'failsafe charging'
<buZz>
which is the latest image to have been built
<bencoh>
I still don't have whatsapp because I don't want to end up with people sending me messages, and me missing those
<_uvos_>
buZz: devel contains several bug fixes related to charging
<_uvos_>
it wokrs pretty mutch 100% there
<buZz>
_uvos_: ok, i dont have those nor can i get it booted up to even install those ;)
<_uvos_>
use android
<_uvos_>
to charge
<buZz>
is there a -devel prefab .img.xz?
<buZz>
my android install is broken since ages
<_uvos_>
sicleo: yes cpcap is fully hw charging
<buZz>
did we break it then? or just forgot to configure it in the right modes?
<_uvos_>
its just mbm sets the tresh really low for cpcap to boot with micro usb pluged in
<buZz>
as i cant seem to get any real charging beside the ~50mA through 'unbooted' setup
<buZz>
maybe thats configured for eh , non-battery setups? :D
<sicelo>
Wizzup .. isn't anbox no longer maintained anymore?
<_uvos_>
buZz: no thats how it works
<buZz>
_uvos_: so, mbm configured the cpcap? to what?
<Wizzup>
sicelo: doesn't matter, it worked
<buZz>
can we change those to work better? :)
<sicelo>
buZz: bq24 can completely handle empty battery with no issues for me. Plus .. you can yank it out easily if you must :-)
<buZz>
eh
<buZz>
not sure what you're saying
<_uvos_>
cpcap allways trickle charges, it trickle charges because it cant negotatiate for more power via usb by itself (you have to negotiate high power class even for 500mA)
<buZz>
sicelo: you're saying 'your issues dont exist' ? :D
<_uvos_>
mbm sets the thresh voltage
<buZz>
_uvos_: to what?
<_uvos_>
to boot
<buZz>
where does it set it?
<_uvos_>
cpcap registers
<buZz>
line of code?
<sicelo>
I'm saying the pain i experience with d4 is absolutely new to me
<_uvos_>
mbm is closed sourcep
<_uvos_>
(its the bootloader)
<buZz>
sicelo: alright, not sure what use that mentioning is then :D
<_uvos_>
anyhow
<buZz>
_uvos_: are you saying even android users back in the day would suffer this?
<buZz>
the settings are what they used aswell?
<_uvos_>
yes no
<_uvos_>
so the problem is
<_uvos_>
that kexecboot is slow
<_uvos_>
and mbm thresh is very low
<buZz>
and doesnt enable charger
<_uvos_>
that was fine
<_uvos_>
because android booted fast
<_uvos_>
but our boot takes so long
<buZz>
hmhm
<_uvos_>
that a battery just over mbm thresh dies
<_uvos_>
anyhow this is mostly solved
<_uvos_>
in devel
<buZz>
just not on a image anyone can install ;)
<_uvos_>
by improving leste behavior
<sicelo>
buZz: i was responding to [16:16] buZz: i dont remember n900 doing much better on this, but havent used millenialPDAs a lot ;)
<sicelo>
Anyway
<buZz>
sicelo: ah, bq24 is relevant to that?
* buZz
no clue
<buZz>
:)
<sicelo>
Of course. It's the charging chip. Sorry i thought you were aware
<_uvos_>
you cant really do any better than cpcap, if the sw is implemented as the booloader expects
<_uvos_>
without violating usb spec
<_uvos_>
(which is common tbf.)
<sicelo>
Anyway, with d4, i noticed that sometimes it won't charge when it boots, and doesn't 'see' charger is connected (no green led).
<sicelo>
So now when battery is very low, and it starts booting, i found removing and reconnecting cable at some 'magical' point in time during boot helps
<_uvos_>
this happens when the low vbus voltage failsafe trips
<_uvos_>
this is set at 4.2V
<_uvos_>
also there was a bug (maybe still in stable) that cause this too
<sicelo>
I've never used stable
<buZz>
yeah , plus the '3.2V' threshold
<buZz>
which is way too high
<_uvos_>
no its not
<buZz>
yeah
<buZz>
it is :D
<_uvos_>
no
<buZz>
whats the voltage of a empty li-ion ?
<_uvos_>
3v
<buZz>
2.8V indeed
<_uvos_>
no
<buZz>
not half a volt higher
<Wizzup>
my droid also often seems to turn off way too soon, esp since I went to asia, and after that I even swapped the batt
<_uvos_>
also thats open circut
<buZz>
_uvos_: batteryuniversity.com
<_uvos_>
buZz they are wrong
<_uvos_>
anyhow this is not the reason for the tresh anyhow
<_uvos_>
the reason its so high
<_uvos_>
is again because we take so long to boot
<_uvos_>
we must guarentee that there is enough charge to boot again
<_uvos_>
since mbm's tresh is too low
<_uvos_>
btw the capacity below 3.2v is almost nothing
<_uvos_>
your not loosing any real capcacity
<_uvos_>
Wizzup: how do you define way to soon in this case
<Wizzup>
not sure, I'm wondering if there's some hw problem tbh
<_uvos_>
ok
<sicelo>
iirc spinal reported faster boot times with leste+systemd ;-)
<_uvos_>
also btw disregarding how bad 2.8v is for the batter
<_uvos_>
d4 hw outright dies at that voltage
<_uvos_>
if you try on lab psu
<_uvos_>
no doubt
<_uvos_>
systemd is way more efficant
<_uvos_>
since its dosent have to execute scripts
<buZz>
yeah just run 28947923878923 binaries
<buZz>
what a improvement
<_uvos_>
well perf wise it is
<Wizzup>
kinda ot
<buZz>
just distupgrade to debian then :P
<buZz>
weee finally found a random li+ charger
<Wizzup>
ok, let me see if I can fix the repodiff tool to be more useful
<Wizzup>
might rewrite it in python
<buZz>
yay charging works, at ~1A, also my droid4 now has -two- usb ports :P
<Wizzup>
hm my d4 just powered off without clear reason
<Wizzup>
maybe it really is some hw problem
<Wizzup>
battery is pretty much full
<buZz>
hmmm , 'apt install ofone' isnt a thing, i guess
<buZz>
eh, not sure, even apt search ofone doesnt fine anything? just some hashing algorithm
<buZz>
find*
<buZz>
oooo, 'ofono'
<buZz>
derp
<Wizzup>
why ofone?
<Wizzup>
oh
<Wizzup>
it dhould already be installed
<buZz>
ah, it is
<Wizzup>
(outside, hard to type)
<buZz>
but , i just wanna try out my simcard in leste
<Wizzup>
my d4 turned off again while trying to copy pstore, I guess it is battery re.ated
<Wizzup>
will try with serial later
<buZz>
Wizzup: maybe confirm the screws are well tight?
<buZz>
> For basic usage, there is the "ofono-scripts" package.
<buZz>
meh, nope
<buZz>
so eh, how can i get just cellular data working on this?
Bratch has joined #maemo-leste
l_bratch has quit [Remote host closed the connection]
<buZz>
droid4 seems to negotiate 500mA charging, sad :)
<buZz>
my external charger is a lot faster ^_^
<bencoh>
I'm not certain there is any real negociation if it settles for 500 :)
<bencoh>
negotiation*
Pali has joined #maemo-leste
vagag has joined #maemo-leste
Twig has quit [Ping timeout: 256 seconds]
Twig has joined #maemo-leste
_uvos__ has quit [Ping timeout: 256 seconds]
<freemangordon>
bencoh: there is no, usb driver on d4 needs love
<buZz>
i'm not really aware how i can test or use my modem in droid4 :P i installed all the 'connui' stuff with 'cellular' in the name, but its telling me nothing is active
<Wizzup>
do you have libicd-network-ofono ?
<Wizzup>
that ought to power the modem
<Wizzup>
and potentially online it
<buZz>
i do not
<buZz>
*installs*
<buZz>
*reboots*
<sicelo>
buZz: development on a device? i wonder the -devel note doesn't belong somewhere else instead :-/
<buZz>
any suggestions?
<buZz>
you dont feel -devel is for development? :P
<sicelo>
not necessarily 'development' on device :-)
<sicelo>
will look at wiki in a bit and make a suggestion
<buZz>
rebooting after installing the libicd-network-ofono got me a 'youfone' display at top and a 2G icon
<Wizzup>
buZz: sounds like you're connected to the network
<buZz>
anything i can do to get gprs data working? :P
<buZz>
15:21 < jonwil> Its also using /etc/gprs/icd-gprs-down.sh and /etc/gprs/icd-gprs-up.sh
<buZz>
i wonder ...
<buZz>
ok, doesnt seem that relevate :P
<sicelo>
buZz: just like on fremantle. just click the status, and choose your mobile operator :-)
<Wizzup>
buZz: no, it's much easier than that
<Wizzup>
like sicelo says
<Wizzup>
but the firs time will wait (it will connect, just reports ux fail)
<Wizzup>
if you try again in 10-15s it should succeed
<buZz>
click the status? eh
<buZz>
which status where? :P i cant click the 'youfone' up top, and on tapping the '2G' icon i dont see anything related, 'internet connections' also doesnt show gprs/gsm as a option
<buZz>
maybe its because of the connui*cellular* stuff i installed before?
vagag has left #maemo-leste [Error from remote client]
<Wizzup>
buZz: status area
<Wizzup>
buZz: ok @ internet connections
<buZz>
i just get four 'icons' on statusarea tapping , 'battery' 'clock&alarms' 'profile' and 'internet connection'
<buZz>
which tells me 'not connected' although i can select wifi networks in there and it'll work
<Wizzup>
buZz: yeah so there ought to be a gprs related network where the wifi networks list is
<Wizzup>
btw, you know there's a meta pkg for all of this ight
<buZz>
i do not! whats it called?
<Wizzup>
hildon-connectivity-modile
<Wizzup>
it will get you connui-iapsettings-gprs amongst other things
<buZz>
'depends on sphone but its not going to be installed'
<buZz>
'depends on conversations but its not going to be installed'
<Wizzup>
it depends on conversations? hrm
<Wizzup>
buZz: do you have -devel ?
<buZz>
yeah
<Wizzup>
then it really should just install
<buZz>
from the 20220410 image
<Wizzup>
did you upgrade / dist-upgrade to be sure?
<Wizzup>
apt update && apt dist-upgrade
<buZz>
i installed a bunch of stuff with 'cellular' in the name that i thought was revelant