**** BEGIN LOGGING AT Sun Mar 26 03:00:04 2017
Mar 26 04:32:09 <Zungo> thanks maemo talk bot for emailing me a happy birthday
Mar 26 04:32:10 <Zungo> :D
Mar 26 04:32:19 <Zungo> hmm
Mar 26 04:32:21 <Zungo> yesterday?
Mar 26 04:32:24 <Zungo> oh nvm
Mar 26 10:45:48 <zGrr> moin ;)
Mar 26 17:01:15 <dreamer> meuh, lost my n900 this weekend (still no clue how .. )
Mar 26 17:01:22 <dreamer> sad day :(
Mar 26 17:03:44 <KotCzarny> [']
Mar 26 17:04:57 <Sicelo> didn't have smscon on it?
Mar 26 17:10:19 <bencoh> :/
Mar 26 17:11:37 <bencoh> meanwhile I'm glad to announce I have A2DP working between maemo5 and a bluez-alsa (bluez5/alsa plugin) device
Mar 26 17:11:45 <bencoh> (with a slightly patched version of bluez-alsa)
Mar 26 17:23:38 <Sicelo009N> n900 is source or sink
Mar 26 17:25:03 <bencoh> source in that case
Mar 26 17:27:10 <KotCzarny> i think inz was interested in such setup
Mar 26 17:27:57 <bencoh> KotCzarny: being able to receive audio from n900 to linux box?
Mar 26 17:28:32 <KotCzarny> hum
Mar 26 17:28:45 <bencoh> or the other way around?
Mar 26 17:31:37 <totalizator> I had it few years ago working almost ootb in Ubuntu, just a matter of config
Mar 26 17:32:08 <totalizator> with pulse audio ofc
Mar 26 17:33:00 <KotCzarny> evilaudio
Mar 26 17:33:05 <bencoh> the good thing here is that it doesn't need PA on sink :)
Mar 26 17:33:33 <KotCzarny> ahm, he needed n8x0 as a bt receiver
Mar 26 17:33:45 <KotCzarny> so completely different. nvm
Mar 26 17:42:22 <totalizator> troubleshooting time - my n900 battery is dying surprisingly fast and syslog is full of: "3422.782367 wlan0: driver reports beacon loss from AP cfuu87c - sending probe request"
Mar 26 17:42:26 <totalizator> could this be the case?
Mar 26 17:43:23 <KotCzarny> wifi noise?
Mar 26 17:43:37 <totalizator> no idea
Mar 26 17:43:46 <totalizator> and what you mean by that?
Mar 26 17:43:50 <KotCzarny> nothing in top/powertop?
Mar 26 17:44:03 <totalizator> nothing 
Mar 26 17:44:37 <KotCzarny> wifi noise == lots of activity on interface not directed to you, but disrupting/waking up the chip
Mar 26 17:44:58 <KotCzarny> disable wifi?
Mar 26 17:45:04 <KotCzarny> change channel?
Mar 26 17:45:16 <totalizator> but originating from my network or not necessarily? 
Mar 26 17:45:31 <KotCzarny> not necessarily
Mar 26 17:45:39 <totalizator> yep, when I disable wifi it's better (I guess)
Mar 26 17:45:45 <bencoh> in your case it might very well be what's logged
Mar 26 17:45:47 <KotCzarny> run kismet and see what's happening on your channel
Mar 26 17:45:47 <totalizator> hence I'm asking what it means
Mar 26 17:46:26 <bencoh> ie wl keeps reprobing AP and waking you up at the same time
Mar 26 17:46:27 <totalizator> and I'm not using any IM on n900
Mar 26 17:46:39 <KotCzarny> im?
Mar 26 17:47:06 <totalizator> instant messaging
Mar 26 17:47:15 <KotCzarny> doesnt matter
Mar 26 17:47:40 <KotCzarny> its not about your traffic
Mar 26 17:48:08 <KotCzarny> its about your device waking up because losing sight of your ap due of interference on your wifi channel
Mar 26 17:48:10 <totalizator> there is almost no other traffic on *my* network most of the day
Mar 26 17:48:25 <KotCzarny> not your network, again
Mar 26 17:48:50 <totalizator> sure, there is lots of other networks around 
Mar 26 17:49:10 <totalizator> and I believe the channel is set to non-overlapping one 
Mar 26 17:49:37 <totalizator> but it doesn't matter as there is so many other networks around utilizing almost every channels 
Mar 26 17:50:09 <KotCzarny> might be someone using illegal antennas with too much gain
Mar 26 17:50:51 <totalizator> and all my other wifi devices are behaving quite ordinary
Mar 26 17:51:25 <totalizator> then I have to test it some other place
Mar 26 17:52:13 <totalizator> it's what I loved about my n900 - that it could stay idle for a few days without charging with wifi on
Mar 26 17:52:21 <Sicelo009N> iirc, there's also a bug report somewhere for that, bugs.maemo.org
Mar 26 17:52:58 <totalizator> yes, I saw that 
Mar 26 17:53:12 <totalizator> but there is no conclusion 
Mar 26 17:53:22 <Sicelo009N> if you think the problem is on N900 itself, have you tried rebooting? or unload/reload the driver?
Mar 26 17:53:48 <totalizator> https://bugs.maemo.org/show_bug.cgi?id=6615
Mar 26 17:53:50 <povbot> 04Bug 6615: Battery Dies Under 6 Hours with Very Moderate Use (Static IP?)
Mar 26 17:54:18 <totalizator> yes, I have rebooted it many times
Mar 26 17:55:29 <totalizator> well, I guess I have to backup my sms db and reflash
Mar 26 17:55:44 <totalizator> once/10 years won't hurt, huh?
Mar 26 17:55:44 <KotCzarny> or just reboot?
Mar 26 17:56:06 <totalizator> I did - it's the same spam in syslog
Mar 26 17:56:11 <KotCzarny> hmm
Mar 26 17:56:18 <KotCzarny> change wifi channel?
Mar 26 17:57:17 <totalizator> I would rather not as my AP is tweaked hard to get maximum bandwidth 
Mar 26 17:57:37 <totalizator> but what I might actually try is to find my old Linksys router
Mar 26 17:58:16 <totalizator> which I remember the 900 has been paired with most of the past years
Mar 26 17:58:40 <Sicelo009N> if your N900 behaves fine in other networks, then of course you'll need to revisit your AP settings :)
Mar 26 17:59:07 <Sicelo009N> just recently had to adjust DTIM on mine for example, and i see better results since 
Mar 26 18:02:23 <Sicelo009N> and can't you save the existing config on your ap before making adjustments such as suggested by KotCzarny? ;)
Mar 26 18:05:15 <totalizator> I'll try but wifi in the city is like jungle survival ;) endless fighting for a non occupied channels 
Mar 26 18:06:49 <KotCzarny> it wouldnt be if all obeyed the law/regulations
Mar 26 18:07:03 <KotCzarny> and not set wifi tx to the max by default
Mar 26 18:07:41 <TheKit> does someone have newer glibc deb for Maemo?
Mar 26 18:07:44 <KotCzarny> but hey, 'i have to have max bWwWWwww! no matter if i use it once a week, and otherwise it would spam overpowered beacons'
Mar 26 18:07:56 <KotCzarny> thekit: need newer kernel first
Mar 26 18:08:33 <TheKit> I don't have N900, so that's no the biggest issue :)
Mar 26 18:08:36 <TheKit> *not
Mar 26 18:09:11 <TheKit> I just found https://talk.maemo.org/showpost.php?p=1354484&postcount=29 this topic, but the links are dead
Mar 26 18:09:20 <totalizator> well, the real issue is that at 2,4GHz you are unable to use max speed anyway if you are following the regulatory reqs
Mar 26 18:10:14 <totalizator> 5GHz is a good solution as almost nobody is using it (around me) 
Mar 26 18:10:26 <KotCzarny> yeah. yet.
Mar 26 18:11:05 <totalizator> even in the crowded places imho
Mar 26 18:12:18 <KotCzarny> we are baking ourselves with e-m
Mar 26 18:15:26 <totalizator> I don't want to start with that :D
Mar 26 18:15:33 <KotCzarny> :)
Mar 26 18:16:06 <KotCzarny> i have 1 or 2 neighbours with stronger signal than my tp-link router 2m away ;)
Mar 26 18:16:38 <KotCzarny> so you know, if your wifi doesnt work as expected, it's.. people
Mar 26 18:16:39 <totalizator> nevertheless I stopped carrying my phone in a pocket some time ago 
Mar 26 18:16:39 <KotCzarny> ;)
Mar 26 18:16:57 <Sicelo> what do you do instead?
Mar 26 18:17:24 <KotCzarny> me? ethernet cable
Mar 26 18:17:28 <totalizator> in my backpack or ...outer pockets
Mar 26 18:17:28 <KotCzarny> or usb
Mar 26 18:18:03 <totalizator> yep, I could live without wifi - totally
Mar 26 18:19:36 <totalizator> but I don't believe that any wireless device could be harmful when not with a direct contact to the body
Mar 26 18:20:07 <totalizator> like ultra low signal phone sticked to your butt
Mar 26 18:20:07 <KotCzarny> um, power is inversely proportional to distance^2
Mar 26 18:20:25 <KotCzarny> so the further away it is, the better, but still, if its strong to begin with.. ;)
Mar 26 18:21:07 <KotCzarny> offtopic: internet radio browser addition in oscp is the best feature since make-rescursive-list
Mar 26 18:21:21 <totalizator> hyhy :3
Mar 26 18:21:42 <totalizator> I agree!
Mar 26 18:24:06 <totalizator> but mostly by the inclusion of the "top voted" category 
Mar 26 18:24:23 <KotCzarny> top click isnt bad too
Mar 26 18:24:31 <KotCzarny> for something new
Mar 26 18:27:19 <totalizator> well adding a few hard coded wouldn't hurt as well, huh?
Mar 26 18:27:41 <KotCzarny> um, thats what ~/.fav-1.m3u etc are for
Mar 26 18:27:54 <KotCzarny> and favX= in config too
Mar 26 18:42:04 <DocScrutinizer05> totalizator: ((and all my other wifi devices are behaving quite ordinary)) well, do your other devices use power saving mode aka PSM on wifi? or are they simply shutting down wifi completely as soon as you 'lock' them, or even are PCs or similar that have no issues with keeping wifi using a few watts of power?
Mar 26 18:42:56 <DocScrutinizer05> first of all make sure your PSM is set to a mode that works with your APN
Mar 26 18:44:47 <DocScrutinizer05> see >><Sicelo009N> just recently had to adjust DTIM on mine for example<<
Mar 26 18:47:15 <totalizator> DocScrutinizer05: oh, I get it now
Mar 26 18:47:23 <totalizator> thanks
Mar 26 18:50:42 <DocScrutinizer05> totalizator: WiFi sleeps 99% of time, and wakes e.f. once per second (DTIM time) to watch AP's beacon and a possible "wake up! I got data for you!" bit in that. When AP doesn't keep exact timing or somebody spams the beacons into illegibility, N900 needs to stay RX active 100% of time, which eats **huge** amounts of battery (simplified picture but gives you the basics)
Mar 26 18:51:01 <totalizator> the only question is what and where would be "a mode  that works with my AP"
Mar 26 18:52:16 <DocScrutinizer05> advanced in WiFi settings: "power savings mode"  iirc
Mar 26 18:52:33 <DocScrutinizer05> none|min|max
Mar 26 18:53:31 <totalizator> I will try 
Mar 26 18:53:35 <DocScrutinizer05> and I suggest to keep TX at 100mW, it doesn't make a huge difference. The purpose of this setting is just for poor frenchmen who mustn't  >10mW outdoors
Mar 26 18:53:53 <totalizator> haha, ok
Mar 26 18:56:04 <DocScrutinizer05> I mean, 90mW (100 - 10) is sucking empty a full battery in sth like 50h, if your WLAN would send all the time, which it doesn't for sure
Mar 26 18:57:14 <DocScrutinizer05> it's WLAN RX that's known to be extremely power greedy
Mar 26 18:58:16 <DocScrutinizer05> and it's poorly working PSM in APs that causes exactly that
Mar 26 18:59:23 <DocScrutinizer05> plus you even got an indication: syslog is full of: "3422.782367 wlan0: driver reports beacon loss from AP cfuu87c - sending probe request"
Mar 26 19:01:12 <DocScrutinizer05> btw zzztop of course shows something: number of IRQ from WLAN
Mar 26 19:05:09 <DocScrutinizer05> and cmdline: wl12xx
Mar 26 19:05:18 <DocScrutinizer05> err
Mar 26 19:05:22 <DocScrutinizer05> and cmdline: <wl12xx>
Mar 26 19:08:18 <DocScrutinizer05> http://paste.opensuse.org/50263550
Mar 26 19:08:37 <totalizator> I will definitely try to solve and debug the issue as I see a great future ahead of my n900 ;)
Mar 26 19:10:22 <DocScrutinizer05> niote the "Average Current: 20 mA" from bq27200.sj which is the actual power consumption of my N900 while it's logged in to IRC and I'm logged in to it via ssh, all over WLAN
Mar 26 19:12:05 <DocScrutinizer05> TTE: 2083 minutes
Mar 26 19:12:23 <DocScrutinizer05> ~2083/60
Mar 26 19:12:23 <infobot> 34.716666666667
Mar 26 19:12:45 <DocScrutinizer05> hours, with a 67% remaining charge
Mar 26 19:12:56 <DocScrutinizer05> 57%, sorry
Mar 26 19:13:50 <DocScrutinizer05> without IRC, it should do a week of standby while ssh over WLAN session established but idle
Mar 26 19:16:38 <enyc> =)
Mar 26 19:24:54 <totalizator> yup, that's what I want it to be
Mar 26 20:14:18 <Sicelo> DocScrutinizer05: what is the meaning of the vol'try/non-vol fields in zzztop output, context switch section
Mar 26 20:14:41 <KotCzarny> voluntary context switches?
Mar 26 20:15:15 <KotCzarny> ie. how many times app given up the cpu control?
Mar 26 20:15:26 <DocScrutinizer05> preemtive/forced (by scheduler, due to e.g. timeslice expiry) versus voluntary task switching
Mar 26 20:15:26 <KotCzarny> vs. forced taking control away
Mar 26 20:15:44 <Sicelo> thanks. cool
Mar 26 20:15:44 <DocScrutinizer05> O'd guess
Mar 26 20:15:49 <DocScrutinizer05> I*
Mar 26 20:16:30 <KotCzarny> if its high, there is probably lots of happening in the system
Mar 26 20:16:50 <KotCzarny> which could result in high 'load' even with low cpu usage
Mar 26 20:18:16 <DocScrutinizer05> basically the voluntary context switches are same number as the events waking up a process
Mar 26 20:20:12 <DocScrutinizer05> forced aka non-voluntary context switches just mean the task is running longer than the scheduler's time slices, so they don't really count for how often the task got woken up, it stays "status: R-unning" when non-vol context switch happens to it
Mar 26 20:21:51 <DocScrutinizer05> a voluntary context switch otoh should mean the process/task left R status for going waiting for something, be it timer, completion of a poll() or wait() or whatever
**** ENDING LOGGING AT Mon Mar 27 03:00:02 2017