<freemangordon>
arno11: does 'killall mission-control-5' fix the issue?
<arno11>
(in haze i see purple/certificate/x509/ca-WARNING **: 14:27:49.403: Lazy init failed because an X.509 Scheme is not yet registered. Maybe it will be better later.)
<freemangordon>
that's fine
<arno11>
@mission control, let me check
<freemangordon>
just tell me if restarting mc fixes the issue
<arno11>
no change unfortunately
Livio has joined #maemo-leste
<freemangordon>
wait, WTYM no change?
<freemangordon>
does mc-tool now lists the account?
<freemangordon>
*list
<arno11>
yeah now the account is visible
<freemangordon>
ok, that explains it
<arno11>
i suppose it should work now if i killall hsm
<freemangordon>
yes
<freemangordon>
wait
<freemangordon>
check if it is vissible in accounts ui
<freemangordon>
it shold be
<freemangordon>
so, my theory is:
<arno11>
yeah it works now
<arno11>
well done
<freemangordon>
on boot, haze crashes so mc cannot add it
<freemangordon>
now, we have to find out why haze crashes
<arno11>
ok
<freemangordon>
lemme test something
<arno11>
ok
<Wizzup>
freemangordon: if you're in vm disable apparmor
<Wizzup>
if you're using the telegram thing
<Wizzup>
12:04 < freemangordon> I wonder if tp-ring onlines the modem
<Wizzup>
it does not
<Wizzup>
12:03 < freemangordon> ok, but what would happen if tp-ring crashes?
<Wizzup>
restart the phone :)
<freemangordon>
sure, but sms will be missed
<freemangordon>
Wizzup: I am not using telegram thing
<freemangordon>
also, it seems to crash tp-haze only in early boot stages
<freemangordon>
I am trying to find a way to gather tp-haze logs whyle starting-up
<freemangordon>
right, but still those should be deleted
<freemangordon>
" Although data stored in /var/tmp is typically deleted in a site-specific manner, it is recommended that deletions occur at a less frequent interval than /tmp."
apac has quit [Ping timeout: 252 seconds]
<freemangordon>
but in leste vm I see files back from 2022
<Wizzup>
it's about 100M on my laptop and also does back to 2019
<Wizzup>
let me check my d4
<Wizzup>
18MB on my d4 (mostly some mafw-gst stuff)
<Wizzup>
I don't think is very impactful right now
<freemangordon>
ok
<Wizzup>
one thing we do also need to fix is setting up relatime on all devices
<Wizzup>
some devices don't have a / entry in their fstab even
<Wizzup>
so this can get a bit tricky
<freemangordon>
yes, we must do that
<Wizzup>
yes, but if there is no /etc/fstab entry, wheree :)
<dsc_>
my problem with /tmp/ is writing to disk, trashing the SD card
<dsc_>
but maybe its not an issue
<Wizzup>
what are you writing that causes this much load?
<dsc_>
just in general
<gnarface>
for something like a phone it's probably acceptable to mount /tmp as tmpfs
<gnarface>
i think that's even the default on raspbian, isn't it?
<gnarface>
typically you wouldn't have much in there, so it doesn't need much space
<gnarface>
that should minimize related disk load
<gnarface>
i think on a system with a mysql instance running, i once hit a 32MB limit i'd set, but otherwise under normal use i don't think i've seen it go above 16MB (iirc raspbian defaults to 64MB?)
<gnarface>
obviously ymmv
ungeskriptet has quit [Read error: Connection reset by peer]
ungeskriptet has joined #maemo-leste
ungeskriptet has quit [Remote host closed the connection]
ungeskriptet has joined #maemo-leste
<freemangordon>
ok, rtcom-presence-ui should now properly react to mc being restarted or to new cm being installed
<freemangordon>
will do the same for addressbook
<freemangordon>
none of these will fix haze issue :)
<freemangordon>
Wizzup: so, what will conversations do if mc got killed?
<freemangordon>
or, you think I shall try to re-init everything?
<freemangordon>
arno11: thanks
<arno11>
np
<freemangordon>
are yuo sure purple debug is enabled?
<arno11>
yep
<arno11>
and i removed old tmp files
<freemangordon>
hmm, I see no diff with the old one
<arno11>
yes indeed...
<freemangordon>
that's weird
<arno11>
mhm
<Wizzup>
freemangordon: keep running and keep checking dbus for tp to come back
<freemangordon>
right, but the issue is that binary might have stalled data, like protocol objects and whatnot
<freemangordon>
tp coming back is not an issue
<freemangordon>
but we shall destroy all the account, protocol, etc objects, re-create TP e-books, attach to the new account manager and start from the beginning
<freemangordon>
not to say that account manager might be provided by th eapplication to the library
<freemangordon>
then what?
<freemangordon>
to mt it is application that shall watch for the account manager service dis-appearing from the bus and re-init the library
<freemangordon>
*to me
<freemangordon>
library has no means to know what is cached in the application
<Wizzup>
ok
<Wizzup>
how often does mc disappear?
<freemangordon>
should never happen ;)
<freemangordon>
adding new connection mamanger is another story
<freemangordon>
*manager
<freemangordon>
I will take care of that in the library
<Wizzup>
`ok
<Wizzup>
I thought maybe the problem we were seeing was mc disappearing during boot or something, but I wasn't following along
<freemangordon>
no, mc is fine
<freemangordon>
it is that tp-haze does not provide accounts information for some reason
<freemangordon>
arno11: please, double-check that purple debug is enabled and also, check if additional debug logs can be enabled for the particular pidgin plugin (telegram-tdlib)
<arno11>
i already doublecheck for purple, will look @tdlib
<Wizzup>
I don't think tdlib has much to do with purple, or do you mean purple-tdlib
<Wizzup>
so is this a telegram specific problem?
<arno11>
i don't think it is telegram specific as i encountered quite the same issues with facebook
<arno11>
it seems maemo specific
<arno11>
i don't remember any troubles with bitlebee
<arno11>
*bitlbee
<freemangordon>
I don't think it is maemo, it is either TP or libpurple issue
<freemangordon>
arno11: does it happen with xmpp haze account?
<arno11>
for xmpp, no idea
<arno11>
we should try
<arno11>
(i don't have any account)
<freemangordon>
Wizzup: could you provide a test account ^^^
<arno11>
btw no trouble with pidgin
<freemangordon>
arno11: yes, but you start pidgin way after averything is ready
<arno11>
yeah
coffeecreature has quit [Remote host closed the connection]
<freemangordon>
hmm, lemme slow down my vm, to see if it will happen
<arno11>
ah good idea :)
coffeecreature has joined #maemo-leste
<freemangordon>
ok, 1 cpu/10% execution cap
<arno11>
if slowness is the problem, it could be also ram/swap
<freemangordon>
yeah, but I think now vm runs even slower than n900 :)
<arno11>
lol ok
<freemangordon>
it still boots :)
<arno11>
:)
<freemangordon>
and I have fb acccount there, so lets see
<arno11>
ok cool
<freemangordon>
no, it seems fine :(. lemme slow it down even more
<freemangordon>
hmm, ok, actually I see no presence icon
<arno11>
ah
<freemangordon>
however, the account is visible in mc-tool
<freemangordon>
and in accounts-ui
<arno11>
yes it was the same for me iirc
<arno11>
*with facebook
<freemangordon>
restarting h-s-m makes it appear :)
<arno11>
yep
<freemangordon>
no, that's with xmpp account
<freemangordon>
lemme see if I can speed-up the machine a bit
<freemangordon>
and still make it happen
<arno11>
ok
<arno11>
what you see with xmpp is exactly what i saw with fb
<freemangordon>
ok, I have a test-case here
<freemangordon>
right
<freemangordon>
however, the account is visible in mc-tool and in accounts-ui
<freemangordon>
so there is another issue with telegram
<freemangordon>
but lets fix that one first
<arno11>
yes but that's because login is differnet
<freemangordon>
arno11: thanks, will get back to you when I have something
<arno11>
if you solve this issue i'm pretty sure it will be ok with tg as well
<freemangordon>
could be, lets see
<arno11>
i.e i saw the same with facebook and solved it by starting hsm later. same result with tg