**** BEGIN LOGGING AT Mon Oct 16 03:00:02 2017 Oct 16 03:25:04 So `a` depends on `b` and `b` depends on `a`, therefore it's impossible to install either, problem solved. Oct 16 03:25:38 package managers are usually a little smarter than that Oct 16 03:25:44 modern ones at least Oct 16 03:26:16 but yeah, if a kernel/modules are part of a package, they really should be part of the same package. Oct 16 03:26:40 Since they're meant to be artifacts of the exact same build process. Oct 16 03:26:52 make && make install && make modules_install Oct 16 03:31:22 (note that my point about not having the package manager "install" the modules/kernel still stands; it would be logical for something like `apt-get install kernel-installer` to give you some "install-kernel" command, which copies modules into /lib/modules and a kernel into /boot, and maybe updates some bootloader configuration) Oct 16 03:32:18 (that way, when you uninstall/upgrade the "kernel-installer" package, it doesn't try to remove the old kernel/modules .. just the ability to install them again) Oct 16 03:34:10 Anyhow, if I end up being in a position to do updates to CSSU-devel I will make some changes to solve that Oct 16 03:41:57 merlin1991: ping Oct 16 03:42:29 well... NEVER run apt-get zpgrade or apt-get dist-upgrade on maemo Oct 16 03:43:35 regardless of all that, it shouldn't ever be possible to install a package in a way where some other thing is also needed for the system to work but that other thing doesn't get installed Oct 16 03:44:02 hmm, good luck Oct 16 03:44:09 why do you say that? Oct 16 03:44:51 because that's a question of dependencies and maemo isn't built in a way that handles dependencies so you could do upgrade or dist-upgrade. Oct 16 03:46:53 I am not interested in fixing it so you can do upgrade or dist-upgrade, I am interested in making it so its impossible to do apt-get install blah and end up with a system that doesn't work (regardless of what blah is) Oct 16 03:46:56 and like you're not supposed to do a apt-get upgrade, you as wel could install some conflicting lib or module or the like with dpkg -i pkgname and that would also not get handled gracefully if the dependencies inside the packages are not made to cope with it Oct 16 03:47:51 CSSU-devel *is* broken Oct 16 03:49:23 maemo-extras-devel has broken packages. You're not supposed to enable either of both repos and if you do neverthrless you MUST NOT do apt-get upgrade. You will be able to install the nroken packages with apt-get install. FIX: do not enable those repos Oct 16 03:58:03 to make it utterly clear: the kernel modules are supposed to have a dependency to kernel (obviously) but - partially due to the way kernel gets flashed on N900 rather than installed as a package - this can't get done in maemo Oct 16 04:00:44 maemo has metapackage that consists of all the needed and paired things to make a working system. and you install that package instead of doing apt.get upgrade. The wrong kernel modules wouldn't get installed by installing the PR metapackage Oct 16 04:02:37 you could apt-get install wrong kernel modules on any system I guess, when you have wrong repos enabled and point to the wrong kernel modulles. that's not specifically maemo and not a bug Oct 16 04:03:15 Yeah fair enough I guess Oct 16 04:04:18 O hosed my PC linux by installing some libs from a newer distro release Oct 16 04:04:21 I* Oct 16 04:04:42 If you install anything manually from any of the CSSU repos instead of installing cssu (stable, testing or thumb) the recommended way and getting mp-fremantle-community-pr and its exact dependencies) its your own fault if the system is screwed Oct 16 04:05:19 which really is a bug I'd say, at least as far as shared object libs are concerned, since tthey are supposed to be able to coexist Oct 16 04:06:09 jonwil: pretty much yes :-) Oct 16 04:07:16 Its my own fault for screwing up my N900 by not reading the warnings never to use apt-get upgrade and by not checking what packages it wanted to install. Oct 16 04:08:26 But definitely we should update cssu-devel with newer packages and we should figure out what packages in cssu-devel have been tested enough to be safe to promote to cssu-testing Oct 16 04:08:39 well, we can try to fix that sick maemo-*-PR metapackage scheme as much as possible Oct 16 04:08:57 I am sure the latest version of maemo-security-certman should be put in cssu-testing for example Oct 16 04:09:01 yes, that too Oct 16 04:09:55 Speaking of which, there is probably a newer certdata.txt that I could pull into maemo-security-certman :) Oct 16 04:12:34 actually, no there is not Oct 16 04:12:52 I just pulled the latest one and it matches the one I pulled last time I updated things Oct 16 04:34:48 jonwil: pong Oct 16 05:20:43 jonwil: ignoring multiple warnings because 'it wont happen to me' isnt smart ;) Oct 16 05:20:52 1/ you ignored warnings about -devel repos Oct 16 05:21:04 2/ ignored warning of not using apt-get update Oct 16 05:21:24 but it's fine, you've learned a bit in the process Oct 16 05:21:45 (fear greatly increases remembering ;) Oct 16 08:30:25 DocScrutinizer05: My phone was "charging" all night -- when I checked it this morning the LED was still blinking yellow. When I disconnected it from the charger, the battery level was just above empty. Oct 16 08:30:39 I am not sure what happened there. Oct 16 08:30:40 uh-oh Oct 16 08:30:49 what charger/cable? Oct 16 08:31:08 your phone might be using more juice than provided Oct 16 08:31:38 KotCzarny: It could be that the WLAN was running all night, that's why I rebooted Oct 16 08:31:50 KotCzarny: The charger is the Nokia AC-2E Oct 16 08:31:52 wifi alone isnt enough Oct 16 08:32:14 did you set it not to turn of the display? maybe some application running all the time? Oct 16 08:32:32 5.3 V, 500 mA output Oct 16 08:32:56 The browser was left open, other than that I don't think anything else was running Oct 16 08:33:23 definitely 500 might be enough, but best to use 900mA or better Oct 16 08:33:31 close browser Oct 16 08:33:38 I rebooted, so... Oct 16 08:33:42 browser closed :) Oct 16 08:33:49 any widgets on desktop? Oct 16 08:34:00 anything weird you have installed? Oct 16 08:34:07 KotCzarny: You mean open applications? Oct 16 08:34:22 well, you can have widgets Oct 16 08:34:33 tiny apps that take place on the desktop Oct 16 08:34:39 not recommended usually Oct 16 08:35:50 there was also a battery/current monitoring script, you might want to run it to see Oct 16 08:35:57 KotCzarny: I did the lazyflashing procedure yesterday with Joerg and installed CSSU Oct 16 08:36:00 ~bq2700 Oct 16 08:36:05 ~bq2700.sh Oct 16 08:36:08 hrm Oct 16 08:36:16 bot is asleep maybe Oct 16 08:36:17 ~bq27200 Oct 16 08:36:18 well, bq27k-detail is http://maemo.cloud-7.de/maemo5/usr/local/sbin/bq27k-detail2, or ~bq27 Oct 16 08:36:19 yeah, cssu is recommended Oct 16 08:36:39 i suspect browser was running some weird js Oct 16 08:37:04 I have definitely had instances when a web page open in the browser drains my battery real fast Oct 16 08:37:09 you can run the bq script as root with '1' as a param to observe current usage Oct 16 08:37:14 KotCzarny: that's possible, the whole procedure (despite all the claims that it would "not take long") took about four hours Oct 16 08:37:50 The n900 browser is a massive resource hog unfortunately Oct 16 08:37:56 is your network slow? last time i was flashing/installing cssu wasnt longer than 30 min Oct 16 08:38:03 that including me reading the steps Oct 16 08:38:49 KotCzarny: I'm talking about the entire procedure Oct 16 08:39:13 yup Oct 16 08:39:20 backups (many), package installs, flashing, restores, etc. Oct 16 08:39:38 well, i dont know about backups, just flashing+cssu Oct 16 08:39:49 Can i install bq27200 from the HAM? Oct 16 08:40:06 nope, it's just a shell script Oct 16 08:40:11 download with wget Oct 16 08:40:15 jonwil: is it the browser or the wifi module that's the resource hog? ;-) Oct 16 08:40:15 then run in shell Oct 16 08:40:19 can be ran via ssh Oct 16 08:40:20 browser Oct 16 08:40:29 wifi isn't a buge resource hog in my experience Oct 16 08:40:33 huge Oct 16 08:40:38 jonwil: that's not what I've been told Oct 16 08:40:46 wifi can be resource hog in crowded areas Oct 16 08:40:50 is the battery maybe just dead? Oct 16 08:41:02 Yes WiFi can be a resource hog if the network is congested Oct 16 08:41:06 jonwil: when I first got the phone I left the WLAN on and the new, fully charged battery was dead by the end of the day Oct 16 08:41:08 wizzup, he wouldnt be able to flash then Oct 16 08:41:11 so I turned it off Oct 16 08:41:47 my experience is that WiFi doesn't drain the battery Oct 16 08:41:54 Wizzup: the battery is not that old Oct 16 08:42:14 jonwil: do you live in the country? Oct 16 08:42:15 but you may have something about the WiFi environment in your area that might be a battery hod Oct 16 08:42:20 I live in a city Oct 16 08:42:26 also, dont enable 'periodically search for networks' Oct 16 08:42:32 THAT's a battery hog Oct 16 08:42:34 KotCzarny: That was on, yes Oct 16 08:42:42 KotCzarny: and yes, I turned that off, too Oct 16 08:42:45 Yeah that is definitely the battery hog there Oct 16 08:42:52 regular old WiFi not so much Oct 16 08:42:56 Well, that is on by default :) Oct 16 08:43:11 so... unless you know this and actively turn it off... Oct 16 08:44:11 apart from that, you can install battery-eye app Oct 16 08:44:26 from my experience it's very good to observe battery history Oct 16 08:44:43 and isnt like many of the others, not eating battery itself Oct 16 08:45:00 also provides nice statistics Oct 16 08:45:20 Where is the "periodically search for networks" option? Oct 16 08:45:27 Ok, I'll make a note Oct 16 08:45:30 I see "search interval" at 10 minutes, is that it? Oct 16 08:45:36 While I have you all here, isn't this phone supposed to have an FM receiver? Oct 16 08:45:44 I can't find an app for it Oct 16 08:45:51 jonwil: internet connections Oct 16 08:45:59 then 'switch to wifi when available' Oct 16 08:46:09 and search interval Oct 16 08:46:21 well, mostly 'search interval' Oct 16 08:46:24 I have that option enabled with search interval at 10 minutes and I dont get battery hog Oct 16 08:46:36 rhombus: the radio only works if you put a headphone in, it uses the wire as antenna Oct 16 08:46:37 I had it enabled at 10 min Oct 16 08:46:53 buZz: But I can't even find an app for it :) Oct 16 08:47:01 there is app for fm radio Oct 16 08:47:04 well thats something else, cant help you there :P Oct 16 08:47:21 KotCzarny: in the community catalogs? Oct 16 08:47:24 called.. 'fm radio' Oct 16 08:47:25 ;) Oct 16 08:47:37 dont remember, probably Oct 16 08:47:41 such a strange name Oct 16 08:47:55 it was a few years since i last installed things on my n900 Oct 16 08:48:27 usually most things come from extras Oct 16 08:48:38 and many times newest versions in extras-devel Oct 16 08:48:48 while extras should be safe to leave enabled Oct 16 08:49:02 extras-devel should be only enabled to install specific apps Oct 16 08:49:21 and only if you trust devel and want latest version Oct 16 08:50:22 https://wiki.maemo.org/N900_FM_radio_receiver#User Oct 16 08:50:27 also, you should read wiki more Oct 16 08:50:31 my stupid wlan at work has a super tight firewall filter, the install fails Oct 16 08:50:34 I'll have to do this at home Oct 16 08:50:49 http://maemo.org/packages/view/fmradio/ Oct 16 08:50:55 KotCzarny: I am fond of reading, but the Wiki is a bit of a mess Oct 16 08:51:01 it is Oct 16 08:51:09 but also great source of useful information Oct 16 08:51:16 KotCzarny: There's lots of stuff there that is totally outdated and for a new user it is confusing and sometimes even dangerous Oct 16 08:51:23 indeed Oct 16 08:51:26 without Joerg's help yesterday I would have had no chance Oct 16 08:51:31 nah Oct 16 08:51:35 you would be fine Oct 16 08:51:36 If people point me to the right info, yes :) Oct 16 08:51:43 just you would learn a lot more Oct 16 08:51:49 (at the expense of more time) Oct 16 08:51:54 (and errors ;) Oct 16 08:52:04 KotCzarny: yes and yes :) Oct 16 08:52:10 usually you can check user's pages on wiki Oct 16 08:52:21 people bookmark important bits Oct 16 08:52:41 Ah, good tip, thanks Oct 16 08:54:18 i wonder why User:Dik is targetted by spam Oct 16 08:54:24 probably someone wrote a script Oct 16 08:54:59 and published in some spamkit Oct 16 08:57:56 sixwheeledbeast: ^ Oct 16 09:01:07 rhombus: also, is your wifi enabled with power saving mode? Oct 16 09:01:12 it tends to eat a bit without Oct 16 09:23:23 KotCzarny: Let me see Oct 16 09:28:14 KotCzarny: Where do I find that setting? Oct 16 09:28:28 in internet connections Oct 16 09:28:35 probably in advanced button Oct 16 09:28:59 KotCzarny: I don't see an advanced button, unfortunately Oct 16 09:29:48 connections -> (select your net) edit -> few times 'next' until you see 'advanced' Oct 16 09:30:56 power saving is at max Oct 16 09:31:04 is that a general setting or does that apply only to this connection? Oct 16 09:31:15 yeah, default is max Oct 16 09:31:23 but you can change it per-connection i think Oct 16 09:31:41 Ah, ok Oct 16 09:45:19 done. Oct 16 09:45:48 Been away for a bit, just reading ^^^ Oct 16 10:08:01 https://www.krackattacks.com Oct 16 10:08:12 well, rip n900 wifi Oct 16 10:08:19 maemo* Oct 16 10:11:26 yeah ... Oct 16 10:13:14 unless we switch to wpa_supplicant (and then a new version) :) Oct 16 10:15:48 or patch the current implementation (is it closed?) Oct 16 10:15:59 The weaknesses are in the Wi-Fi standard itself, and not in individual products or implementations. Therefore, any correct implementation of WPA2 is likely affected Oct 16 10:16:01 ;) Oct 16 10:16:39 KotCzarny: apparently there is some kind of workaround to the weakness Oct 16 10:16:46 bencoh: closed Oct 16 10:16:47 (backward-compatible) Oct 16 10:30:40 Damn. I think my battery is damaged :( Oct 16 10:31:05 ~polarcell Oct 16 10:31:05 somebody said polarcell was http://www.ebay.co.uk/itm/390402445382 BL-5J Replacement Oct 16 10:31:23 I have a variocell, it's been great up until now Oct 16 10:31:32 but it's only 5 months old! Oct 16 10:31:41 yup, that's how those cheapos work Oct 16 10:31:47 great at the first sight Oct 16 10:31:52 then deteriorate quickly Oct 16 10:31:58 Cheapos? It was the only thing available! Oct 16 10:32:11 i still own and use ORIGINAL nokia battery that came with my n900 in 2009 Oct 16 10:32:22 My original battery was totally useless Oct 16 10:32:54 It took two days to fully charge and then it held its charge for about six hours Oct 16 10:33:12 I have had good success with several polarcells from wannasee Oct 16 10:33:20 but aslo get good charger Oct 16 10:33:27 500mA isnt going to cut it Oct 16 10:33:45 beware of cheapos too. Oct 16 10:33:49 reminds me I should probably order a spare battery and compatible charger Oct 16 10:33:50 better get original nokia charger Oct 16 10:34:00 expected life is only 2-3 years Oct 16 10:34:20 I'm using the charger that came with the phone, an original Nokia charger Oct 16 10:34:27 actually I dont really need a compatible charger, only a compatible cable (ie D+/D- shorted to ground) Oct 16 10:34:42 The battery was working beautifully until I did this reflash Oct 16 10:37:29 Stellar, these Varocells aren't even available anymore Oct 16 10:38:10 Is there any point in buying an old Nokia BL-5J, or one in original packaging? They've gotta be ancient by now Oct 16 10:38:51 I was looking at them magnetic USB cables but I imagine they are not D shorted. Oct 16 10:40:01 When I search for BL-5J on ebay, nothing but junk comes up Oct 16 10:40:31 Are there other model numbers that are equivalent? Oct 16 10:42:59 rhombus: why not just get a Polarcell? Oct 16 10:43:26 I imagine if anyone claims to be selling a BL-5J nowadays, it's probably fake. Oct 16 10:43:27 Maxdamantus: There isn't one available -- that link comes up with a "CURRENTLY SOLD OUT" page Oct 16 10:43:56 and the links below are to UK sellers, that's not much use to me Oct 16 10:49:01 Ok, where do I get a decent charger? Can I just use a generic (micro-)USB charger? Oct 16 10:49:15 I don't want cheap junk Oct 16 10:51:52 They seem to be advertising them on Amazon still, maybe they'll ship to other countries through there. Oct 16 10:53:42 Do you mean a charger to just plug into the phone? Oct 16 10:55:03 http://www.wannsee-electronic.de/handyakkus/nokia/polarcell_hochleistungsakku_fuer_nokia_n900_ersetzt_originalakku__bl-5j_i2_2090_0.htm Oct 16 10:55:11 aiui, you can get around the D shorting issue through software. Oct 16 10:55:29 Maxdamantus: there must be suppliers on the Continent Oct 16 10:55:43 No idea if there is an english site but maybe a good idea to replace the factiod with the company site over ebay. Oct 16 10:56:03 there were still new bl-5j produced by nokia for some nokia asha series Oct 16 10:56:10 that was year ago or so Oct 16 10:56:21 sixwheeledbeast: No worries, I can read the German Oct 16 10:56:35 i can't Oct 16 10:56:43 KotCzarny: Produced by Nokia, or by "Nokia"? ;-) Oct 16 10:57:43 I think you can just do something like: echo 1800 >/sys/class/power_supply/bq24150a-0/charge_current Oct 16 10:57:48 since nokia is no more since ~2012, by whatever m$ ordered it to, i presume factories didnt change Oct 16 10:58:19 after plugging in a non-shorted supply .. not sure though, I've just shorted the supplies I happen to use. Oct 16 10:58:24 also, my n900 came with 900mA charger Oct 16 11:01:04 KotCzarny: Can you direct me to a suitable 900 mA charger? Oct 16 11:01:06 Hmm .. with those magnetic cables, do you actually need to short them? Shouldn't they just function as normal USB cables, so you'd just have it shorted at the type A port, or have it plugged into a computer that it can negotiate a current with? Oct 16 11:01:54 https://wiki.maemo.org/N900_Accessories#Chargers Oct 16 11:02:33 oh, it's 1200mA, not 900 Oct 16 11:02:50 most likely my n800 charger is 900 one Oct 16 11:03:28 AC-10U Oct 16 11:04:20 but if you have htc/samsung etc, you can use that one Oct 16 11:04:34 just use the original, not noname one Oct 16 11:05:20 AC-10U will have a NEMA 1-15 plug, no? Oct 16 11:05:26 It says here for US N900 Oct 16 11:06:26 I need a type F plug Oct 16 11:07:02 ac-10e seems eu version Oct 16 11:07:09 ah, thank you Oct 16 11:08:54 these reviews are not encouraging Oct 16 11:09:16 maybe fakes? Oct 16 11:09:32 Entirely possible :( Oct 16 11:10:02 i havent seen anyone on #maemo saying his/hers charger breaking bad Oct 16 11:10:03 Is this a standard charger format? Does it have to be a Nokia charger? Oct 16 11:10:19 as i've said, even cheapo fake will charge Oct 16 11:10:30 but it's best to use branded one Oct 16 11:10:39 and yes, doesnt have to be nokia branded Oct 16 11:10:52 just that nokia did solid products Oct 16 11:11:01 The reviews on the German Amazon site are savage -- ("broke after three weeks", "complete junk", etc) Oct 16 11:11:21 they are probably poorly refurbished previously defective units Oct 16 11:12:14 oh, and some more "fake" and "really original? I doubt it" "not recommended" Oct 16 11:12:16 so much for that Oct 16 11:12:31 So a standards compliant micro-USB charger will do the job Oct 16 11:12:37 I'll get one of those then Oct 16 11:13:39 Though I think there's some chance a "standard-compliant" charger wouldn't be standards-compliant enough. Oct 16 11:14:02 if it's original and branded, it should be ok Oct 16 11:14:16 try borrowing some charger Oct 16 11:14:16 in that it might not have its data pins shorted, which means the device needs to negotiate a current with a computer. Oct 16 11:14:20 Maxdamantus: one issue is that the Nokia charger has an output voltage of 5.3 V listed, wheres the standard chargers have 5 V output Oct 16 11:14:34 The voltage shouldn't be an issue. Oct 16 11:15:11 i often charge n900 straight from my thinkpad too Oct 16 11:15:15 no issues Oct 16 11:15:27 (but thinkpads can easily supply more than 500mA) Oct 16 11:16:32 Mm, I've done that too, including using the charging port thing, which I think acts as a dumb supply (so the N900 will be able to draw 1.6 A or something from it). Oct 16 11:16:51 (I think that's only while the thinkpad is asleep/off) Oct 16 11:17:04 man, it just gets worse: "The third time I plugged it in, there was a dull thud... and the charger was dead." Oct 16 11:17:11 maxd: if main battery isnt charging, there is a lots of power available Oct 16 11:18:03 KotCzarny: yeah, but afaik you're not normally going to draw more than 500 mA from a computer, since the device won't try to negotiate more than 500 mA. Oct 16 11:18:29 maxd: i was running 1A and more hdds via usb with thinkpads Oct 16 11:18:36 those are really solid machines Oct 16 11:18:39 (were.) Oct 16 11:18:46 KotCzarny: I'm just talking about the current draw from an N900. Oct 16 11:19:06 KotCzarny: yes. I will hang on to my T500 for as long as possible, it was the last decent model Oct 16 11:19:13 KotCzarny: sure, the Thinkpad has more than enough power available, but I don't think the N900 will draw more than 500 mA from a computer, unless you force it to. Oct 16 11:19:28 (that is, more than 500 mA from a smart power supply) Oct 16 11:19:34 maxd: still better power quality than any noname charger Oct 16 11:20:44 Well, if it supplies enough current at the right voltage, it should be the same as any other charger. Oct 16 11:22:03 a huge industry, producing tonnes of electronics that turns to junk in just a few months Oct 16 11:22:05 very discouraging Oct 16 11:22:29 rhombus: i use my n900 since 2009 Oct 16 11:22:40 got 3 spares just in case Oct 16 11:22:49 and yeah, I can confirm that it behaves the way I described. Oct 16 11:22:50 KotCzarny: I wasn't referring to the N900 or the Thinkpad T_series prior to 2009 :) Oct 16 11:22:58 plug it into computer, charge_limit = 500 Oct 16 11:23:09 plug it into power supply with data pins shorted, charge_limit = 1800 Oct 16 11:23:14 The rest of the industry... all I can say is, holy $%& Oct 16 11:23:37 and iirc, the Thinkpad only acts as a dumb supply when it's off/asleep. Oct 16 11:23:55 otherwise it's just a computer, and a USB device shouldn't ask for more than 500 mA. Oct 16 11:23:59 at least until USB 3 Oct 16 11:24:16 (which N900 obviously doesn't implement) Oct 16 11:26:09 btw. in idle n900 should take ~10mA Oct 16 11:26:16 You can always just force the N900 to draw up to 1800 mA by writing "1800" into the "charge_limit" sysfs node, which might be okay. Oct 16 11:26:32 that's why even 500mA would work. unless something eats a lots of power Oct 16 11:27:29 Well, the advantage would be faster charging. Oct 16 11:27:42 yes, but he had a problem charging overnight Oct 16 11:27:57 If it used 0 mA, it would still take at least 3.2 hours to charge a 1600 mAh battery at 100% efficiency. Oct 16 11:28:13 (when drawing 500 mA) Oct 16 11:29:29 KotCzarny: Maxdamantus As I said, the charger worked wonderfully right up until yesterday, when I did the flashing Oct 16 11:29:35 The battery was full when the process started Oct 16 11:29:59 I did not have the phone plugged in during the updates, as I assumed that a full battery would mean I had more than enough time Oct 16 11:30:10 by the time the process was complete the battery was almost totally empty Oct 16 11:30:19 And now, it is not charging Oct 16 11:30:22 anyway, in full use with screen on battery suffices for ~4-10 hours Oct 16 11:30:31 I don't think the charger is to blame Oct 16 11:30:40 I think the battery has checked out Oct 16 11:31:10 I have no way of knowing this for certain, but the phone recognizes that a charger has been connected when I plug it in and indicates that it is charging Oct 16 11:31:25 you can try leaving it OFF for 30 minutes Oct 16 11:31:29 I would use that script you recommended but I am at work and I can't use the network Oct 16 11:31:30 and see if it charged Oct 16 11:31:33 a bit Oct 16 11:31:36 rhombus: it needs to be able to boot the OS to charge effectively. Oct 16 11:31:54 KotCzarny: you mean charge it in off state? Oct 16 11:31:55 maxd: he left it installing cssu Oct 16 11:32:01 rhombus: it will only draw a small amount of current (I think 100 mA) through just hardware. Oct 16 11:32:07 KotCzarny: when I went to bed the installlation was complete Oct 16 11:32:36 Maxdamantus: as I've said, I haven't any issues charging the phone -- on or off -- until this reflash Oct 16 11:33:01 This is not an existing problem, this started last night Oct 16 11:33:11 I will try turning the phone off and charging it that way Oct 16 11:33:48 and order two replacement cells Oct 16 11:34:14 rhombus: hang on, how are you determining that charging isn't working? Oct 16 11:34:35 Maxdamantus: It charged over night and the battery was still almost empty Oct 16 11:34:47 rhombus: maybe installing CSSU included using replacement-bme, which can be misleading if the battery is uncalibrated. Oct 16 11:34:49 Maxdamantus: then I charged for 2 hours this morning, after a reboot, and still nothing Oct 16 11:34:54 ya. that too Oct 16 11:35:13 Maxdamantus: oh, man, really? (I wish someone had mentioned this when I did the update) Oct 16 11:35:52 ~calibration Oct 16 11:36:08 ~battery-calibration Oct 16 11:36:14 ~listkeys battery Oct 16 11:36:15 Factoid search of 'battery' by key (11 of 17): true, but the lack of extra battery #DEL# ;; #maemo battery ;; flips ibot's battery disconnect switch ;; batterytest ;; wmbattery ;; battery disconnect switch ;; as battery acid #DEL# ;; battery damage ;; battery-usage ;; batteryfaq ;; heh, no bottom view. can't tell if the battery #DEL#. Oct 16 11:36:34 ~listkeys calibration Oct 16 11:36:35 Factoid search of 'calibration' by key (1): bq27k-calibration. Oct 16 11:36:42 ~bq27k-calibration Oct 16 11:36:42 well, bq27k-calibration is http://talk.maemo.org/showthread.php?p=1247595#post1247595 Oct 16 11:37:01 rhombus: lsmod | grep bq27 Oct 16 11:37:22 rhombus: if that shows the module, it'll be using the bme replacement. Oct 16 11:37:45 Maxdamantus: ok, hang on, I have to boot the phone, it was off Oct 16 11:38:02 also, you are right, it was not charging when off. Why is that? That worked last night Oct 16 11:38:02 Maxdamantus: but one thing is weird, because he have said, he left it charging overnight and in the morning it was still orange Oct 16 11:38:05 which uses information stored in the batter calibration chip to show charge, instead of just doing something weird with the current voltage. Oct 16 11:38:32 KotCzarny: hm, I guess that might be weird. Oct 16 11:38:34 rhombus: no, phone never goes into off mode, it fakes being off Oct 16 11:38:47 he meant when 'laptop is off' Oct 16 11:38:55 oh, sorry Oct 16 11:38:58 (( also, dont enable 'periodically search for networks' THAT's a battery hog)) ONLY when WLAN is not logged in to an AP Oct 16 11:39:14 KotCzarny: But then it should still charge when off, no? The yellow LED was not on even though the phone was plugged in Oct 16 11:39:19 and the charger stopped whistling Oct 16 11:39:28 when I turned it on again, the whistling returned Oct 16 11:39:52 battery-eye DRAINS battery Oct 16 11:39:54 btw. it was ON or pulsing? Oct 16 11:40:30 DocScrutinizer05: nope, afair it only takes snapshots when system events occur, and the graph isnt continuous Oct 16 11:40:34 KotCzarny: When? Just now? Just now it was totally off Oct 16 11:40:39 no, in the morning Oct 16 11:40:46 Maxdamantus: lsmod | grep bq27 returns nothing Oct 16 11:41:09 KotCzarny: in the morning it was yellow... I *think* it was pulsing Oct 16 11:41:28 ok. because sometimes daemon responsible hangs Oct 16 11:41:35 DocScrutinizer05: It may drain battery, but the key point here is that this worked fine until the reflash Oct 16 11:41:42 DocScrutinizer05: with battery-eye on Oct 16 11:42:01 KotCzarny: Ok, well, I've rebooted this morning Oct 16 11:42:19 bq27 module is not running, so can I assume it is using the old battery manager? Oct 16 11:42:23 sorry, wall of text with lots of ... Oct 16 11:42:46 I can't see what's happening and what all the incorrect assumptions were Oct 16 11:43:46 basically he claims his battery didnt charge overninght after cssu install Oct 16 11:43:56 *overnight Oct 16 11:44:16 you might recall I tried to make sure your device charged. That was for a reason. Maybe we should have given it another boot. Then a 500mA charger is not compatible with N900 anyway Oct 16 11:44:37 but it should charge even with 500mA Oct 16 11:44:41 I assumed you had a compatible charger Oct 16 11:44:57 also, he says it was working before Oct 16 11:45:00 KotCzarny: no. 500mA has no D+- short Oct 16 11:45:16 but it still would charge, only slower Oct 16 11:45:30 not 'almost empty' after whole night of charging Oct 16 11:46:20 no Oct 16 11:46:30 it would charge with max 100mA Oct 16 11:46:55 still WOULD charge Oct 16 11:47:06 and yes, that can and will result in battery almost empty after a night Oct 16 11:47:16 when display backlight on Oct 16 11:47:25 did he left lcd on? Oct 16 11:47:34 do you know he didn't? Oct 16 11:47:35 DocScrutinizer05: So whoever sold me the phone wasn't totally honest -- the charger is not original? Oct 16 11:48:02 it's maybe a genuine Nokia charger but not for that device Oct 16 11:48:15 maybe the cable is just 2 wires? Oct 16 11:48:25 is the cable unusually thin? Oct 16 11:48:36 DocScrutinizer05: but... how many times do I have to repeat this? This worked fine until yesterday. Was I imagining things? Oct 16 11:49:01 It would take about eight hours to charge from totally empty to totally full Oct 16 11:49:18 I routinely did partial charges (from say 50 - 60 % to 100 %) in the evenings or over night Oct 16 11:49:24 btw. can you access www at all? Oct 16 11:49:27 and a charge held typically two to three days Oct 16 11:49:32 yeah, charging has a number of modes Oct 16 11:49:34 KotCzarny: let me check Oct 16 11:49:55 [2017-10-16 Mon 13:44:16] you might recall I tried to make sure your device charged. That was for a reason. Maybe we should have given it another boot. Oct 16 11:50:48 wild guessing, anecdotal "evidence", and inappropriate measures won't "fix" this now Oct 16 11:51:06 give the device a reboot! Oct 16 11:51:14 again? I rebooted this morning Oct 16 11:51:19 then charged for two hours Oct 16 11:51:25 no effect, battery shows as empty Oct 16 11:51:29 https://wiki.maemo.org/N900_Hardware_Charge_Meter/bq-script Oct 16 11:51:39 dont know if its the latest/proper version Oct 16 11:51:44 aah ok. This morning then fine Oct 16 11:51:54 unplig and replug USB Oct 16 11:52:06 KotCzarny: I'm on the company guest network, it has a portal Oct 16 11:52:15 Let me see if I can use the employee wifi Oct 16 11:53:11 do you have any data plan on gsm? Oct 16 11:53:12 :) Oct 16 11:53:34 ohmy Oct 16 11:55:32 D+- short means fastcharging. 500mA is no fatscharging, so a 500mA charger will most likely not have D+- short. And device is meant to try USB ENUM when it sees VBUS without D+- short Oct 16 11:56:02 hook it up to a PC so you know it's charging with 500mA Oct 16 11:56:13 damn, I can't authenticate Oct 16 11:56:32 maybe do that stuff tonight then? Oct 16 11:56:43 DocScrutinizer05: yeah, i will have to do it then Oct 16 11:56:58 instead of draining your battery further by futile installaion efforts Oct 16 11:57:07 (I don't understand why this authentication doesn't work, but oh well) Oct 16 11:57:34 I would connect the phone to my thinkpad but lack a suitable microUSB cable Oct 16 11:57:40 more stuff to add to the shopping list Oct 16 11:58:21 close all apps and lock/dim the screen, then atach it to that charger. This way it should still charge a little bit Oct 16 11:59:05 and yeah, disable WLAN Oct 16 11:59:18 also, funny, you can access irc but not www Oct 16 11:59:25 scanning for APs really is a terrible battery hog Oct 16 12:00:27 the more APs in vicinity the worse Oct 16 12:01:22 there's a reason the system does this only periodically Oct 16 12:01:58 disable periodic scanning, that's OK then Oct 16 12:04:52 You should still be able to get 500 mA from a 500 mA charger, by doing `echo 500 >/sys/class/power_supply/bq24150a-0/charge_limit` after plugging it in. Oct 16 12:06:31 (you can also read that file to see how much it's willing to draw; will probably say 100 after plugging it into that charger) Oct 16 12:11:30 ~power Oct 16 12:11:30 well, power is http://wiki.maemo.org/N900_Hardware_Power_Consumption Oct 16 12:12:07 (on phone) Oct 16 12:12:48 >>WLAN on - no powersave *) - 200 mA<< *) or no associated AP which controls powersave Oct 16 12:13:41 NB: >>WLAN transmit 802.11b - 150 mA<< - only Oct 16 12:13:55 b Oct 16 12:14:06 so WLAN RX without AP sucks more power than constant TX Oct 16 12:32:02 (( It would take about eight hours to charge from totally empty to totally full)) doesn't match charging with 500mA which shouldn't take longer than 3h to charge a 1300mAh battery. Rather matches charging a maybe 20% battery (shows as empty) to 80% (shows as full) at 100mA, particularly when the battery capacity has aged and is maybe only 1000mAh now Oct 16 12:32:47 and 100mA charging is what you'd expect to see from a charger without D+- short Oct 16 12:33:05 DocScrutinizer05: ok Oct 16 12:33:05 see what maxd said Oct 16 12:33:13 so I need a new charger Oct 16 12:33:18 connect that charger, then check the sys file Oct 16 12:33:34 also note that charging details changed from PR1.2 to PR1.3 iirc Oct 16 12:33:34 if it's 100, set it to 500 and observe Oct 16 12:33:41 KotCzarny: that works only if bq27 is installed Oct 16 12:33:54 you can try switching to cssu-thumb Oct 16 12:34:12 or cssu-testing Oct 16 12:34:13 errwut? Oct 16 12:34:33 ahm Oct 16 12:34:37 rhombus: please don't touch *thumb! (for now) Oct 16 12:34:49 thought cssu comes with newer module already Oct 16 12:35:46 no, CSSU is about fixing stuff Oct 16 12:36:02 hmm, where does that module coming from? Oct 16 12:36:09 PK Oct 16 12:36:15 i have pk53 Oct 16 12:36:19 and no such sys file Oct 16 12:36:31 it needs to load that module by hand? Oct 16 12:36:38 then the module is most probably blacklisted Oct 16 12:36:58 what was the module named again? Oct 16 12:37:16 no idea Oct 16 12:37:33 only mousedev is in blacklist Oct 16 12:38:33 ./bq27x00_battery.ko Oct 16 12:40:47 ok, off the phone Oct 16 12:41:03 (( see what maxd said)) sorry, I don't see that Oct 16 12:41:30 funny, seems that module is not used, and bq script works fine without it Oct 16 12:41:40 KotCzarny: I'm on IRC on my laptop on the LAN Oct 16 12:41:54 yes, BQ script works fine *only* without that module Oct 16 12:41:57 rhombus: download bq script on laptop then scp ? Oct 16 12:42:07 +1 for scp Oct 16 12:42:52 please leave rhombus alone with messing up his phone while no connection to internet Oct 16 12:43:09 whoa, just noticed something Oct 16 12:43:22 When I press on the status bar, I see Oct 16 12:43:23 bq script is safe Oct 16 12:43:27 a notifier "not enough charging power"? Oct 16 12:43:48 "Battery: 13% Battery charging 168/1234 mAh 3 hours" Oct 16 12:44:04 KotCzarny: don't you tell the inventor of bq script about its properties ;-) Oct 16 12:44:19 oh, tell me more, what are the risks? :P Oct 16 12:44:39 (reading all your previous messages) Oct 16 12:45:56 KotCzarny: I cannot get on the company intranet, I have to do this at home this evening Oct 16 12:46:23 Ok, I have closed all apps and dimmed the screen. Internet connections says "no connection" Oct 16 12:47:00 make sure you set to "scan for new APs: never" or somesuch, in internet/WLAN settings Oct 16 12:47:00 I will have to see about a replacement charger, the Nokia ones seem to be fakes Oct 16 12:47:16 s/never/manually/ Oct 16 12:47:17 DocScrutinizer05 meant: make sure you set to "scan for new APs: manually" or somesuch, in internet/WLAN settings Oct 16 12:48:17 DocScrutinizer05: I have : "Connect automatically: Always ask" Oct 16 12:48:31 is that the right place? Oct 16 12:48:42 next to it, moment please Oct 16 12:48:47 * DocScrutinizer05 fetches phone Oct 16 12:50:05 Einstellungen - Suchinterval: Nie Oct 16 12:50:15 Einstellungen - Internet - Suchinterval: Nie Oct 16 12:50:33 From which Settings applet? Oct 16 12:50:37 yes Oct 16 12:50:55 settings - internet - >> Oct 16 12:50:58 from which one? Oct 16 12:51:01 Internet connections? Oct 16 12:51:04 yes Oct 16 12:51:13 Because there, I see only Oct 16 12:51:30 "Automatisch verbinden: Immer fragen" Oct 16 12:51:32 then on the right Oct 16 12:51:36 "Verbindungen" Oct 16 12:51:39 and "Speichern" Oct 16 12:51:49 "Suchintervall" is greyed out Oct 16 12:52:20 (And it says 30 min) Oct 16 12:52:49 Ok, now I set "WLAN" and changed the interval to never Oct 16 12:52:57 and then set it back to Always ask Oct 16 12:53:23 I'm going to leave this be until I get home tonight Oct 16 12:53:36 DocScrutinizer05: Until then, can you suggest a better charger? Oct 16 12:53:55 three buttins: "Autimatisch verbinden", "Suchinterval" and a checkbox "[ ] Bei verfuegbarkeit in WLAN wechseln" - values: "Beliebige Verbindung", "Nie", "[x]" Oct 16 12:54:26 DocScrutinizer05: What's wrong with "Immer fragen"? Oct 16 12:54:34 up to you Oct 16 12:54:50 Otherwise it will automatically try to connect, no? Oct 16 12:54:54 yes Oct 16 12:55:07 but not via WLAN when interval is set to "Nie" Oct 16 12:55:15 I don't have a data plan, so "always ask" is fine, it worked fine last night Oct 16 12:55:21 yes Oct 16 12:55:36 It tells me there is no connection, shows me a list, and lets me choose -- that's perfectly acceptable Oct 16 12:56:23 DocScrutinizer05: Can you suggest a charger? Oct 16 12:56:33 DocScrutinizer05: There are loads of Nokia fakes out there Oct 16 12:56:56 they are usually good fakes or even originals Oct 16 12:57:18 DocScrutinizer05: The product reviews I've been reading suggest otherwise Oct 16 12:58:21 buy some nokia phone from telco? ;) Oct 16 12:58:35 but no, I can't recommend a charger since I usually get them with phones Oct 16 12:58:45 KotCzarny: you're suggesting that any Nokia charger will do Oct 16 12:58:59 KotCzarny: I need not limit myself to the AC-10E, then Oct 16 12:59:15 KotCzarny: Do I have to start soldering shit? Because I don't have time for that Oct 16 12:59:25 rhombus: for slow charging any will do Oct 16 12:59:34 basically any charger will do as long as it has the (usual) 1000mA+ and it has (needs check) a short between D+ and D- data line of USB Oct 16 12:59:44 for fast charging, it nees to follow a standard Oct 16 13:00:10 DocScrutinizer05: is that some sort of standard? Oct 16 13:00:20 and those without D+- short are useless even for slow charging Oct 16 13:00:39 DocScrutinizer05: What does it need to support fast charging? Oct 16 13:00:40 rhombus: yes, it's one of 4 competing standards Oct 16 13:00:50 DocScrutinizer05: What's the standard called? Oct 16 13:01:05 they don't have names, it's a hige PITA Oct 16 13:01:12 huge* Oct 16 13:01:17 funny as it is, i havent ever seen a charger that wouldnt work with my n900 Oct 16 13:01:43 most chargers now follow the Apple standard which has a resistor network on D+/- Oct 16 13:01:44 ie. my gf's samsing travel charger works as expected and is doing fast charge Oct 16 13:01:52 *samsung Oct 16 13:02:02 KotCzarny: so I can get a Samsung travel charger, done Oct 16 13:02:12 um, results may vary Oct 16 13:02:13 ;) Oct 16 13:02:20 better just borrow one Oct 16 13:02:23 before buying Oct 16 13:02:28 KotCzarny: argh, then why do you make a claim like that? Oct 16 13:02:36 because it works in my case? Oct 16 13:02:57 KotCzarny: why wouldn't it work in mine? Oct 16 13:03:08 how many chargers did you try? Oct 16 13:03:24 KotCzarny: I haven't tried any apart from the one that was packaged with the phone Oct 16 13:03:39 Because I don't want to wreck the phone, for starters Oct 16 13:03:44 nah Oct 16 13:03:47 you wont Oct 16 13:03:49 rhombus: you best get a charger with USB A receptacle for regular USB cable. Then either check (and if needed fix) D+/- short, or mod a USB cable by opening the outer sleeve carefully and cutting+conecting the green and yellow cable Oct 16 13:04:12 DocScrutinizer05: can D-short cables be purchased? Oct 16 13:04:24 yes, but tricky to find Oct 16 13:04:28 I'll need a multimeter to check the cable, yes? Oct 16 13:04:35 rhombus: do you have access to any phone charger in house/at work? Oct 16 13:04:37 no Oct 16 13:04:38 I've been meaning to purchase one anyway Oct 16 13:04:41 just borrow for a charge or two Oct 16 13:04:44 you can check with your N900 Oct 16 13:04:44 KotCzarny: my wife has one or two Oct 16 13:05:08 I'm guessing infobot will tell me how Oct 16 13:05:16 if it has microusb, just try it Oct 16 13:05:16 I was busy to find the cmdline for you but I had to abort that to kicj in here once more to stop shit happen Oct 16 13:05:46 yes, I was getting a lot of dangerous advice Oct 16 13:05:51 ;-) Oct 16 13:06:01 you have backup, no? ;) Oct 16 13:06:07 if so, fire away! Oct 16 13:06:41 What I have little of is time, and I've already spent way too much of it on this today already. If you guys are around this evening, I might see you then. I have to get back to work. Oct 16 13:06:59 rhombus: cat /sys/devices/platform/musb_hdrc/charger Oct 16 13:07:01 But thank you all for your tips. Now at least I know why the charging performance on this phone has been modest Oct 16 13:07:08 1=D+- short Oct 16 13:07:48 DocScrutinizer05: it says "1" Oct 16 13:07:59 let me verify Oct 16 13:08:36 yes, verified Oct 16 13:09:04 ok, so the charger can do fast charging? Oct 16 13:09:26 DocScrutinizer05: ^ Oct 16 13:09:34 it can't (fastcharging means > 1A), it just *claims* it could Oct 16 13:09:46 ok Oct 16 13:09:50 I.E it has D+- short Oct 16 13:09:59 but I can verify that any other charger i connect at least has D+-short Oct 16 13:10:05 it mustn't unless it can provide 1A Oct 16 13:10:11 yes Oct 16 13:10:33 so fast charging = must have D+-short Oct 16 13:10:41 no D+-short, definitely no fast charging Oct 16 13:10:42 and 1A plus Oct 16 13:10:47 right Oct 16 13:10:48 yes Oct 16 13:10:48 got it Oct 16 13:10:55 i thought no short == no charging Oct 16 13:10:56 :-) Oct 16 13:10:57 at all Oct 16 13:11:06 buZz: only 100mA Oct 16 13:11:19 it thinks it's a USB host Oct 16 13:11:20 but not without D- and D+ connected to -something- Oct 16 13:11:28 if its not connected to anything, it wont charge Oct 16 13:11:33 ok, friends, I have to work Oct 16 13:11:35 thanks again Oct 16 13:11:38 o/ Oct 16 13:11:40 see you this evening Oct 16 13:11:40 (aka just V+ and GND) Oct 16 13:11:42 o/ Oct 16 13:11:45 \o Oct 16 13:12:57 weird as it is, all ac-2e chargers i find are using barrel plug, not microusb Oct 16 13:13:07 could it be someone was fiddling with his? Oct 16 13:13:42 buZz: iirc the method is: when VBUS then set 100mA, check for D+- short and set 1A if seen, try USB host detection and start ENUM when seen Oct 16 13:13:56 weird Oct 16 13:14:00 ENUM would negotiate up to 500mA then Oct 16 13:14:09 i see zero charging happen on my selfmade powerbanks Oct 16 13:14:20 which have D- and D+ just not connected Oct 16 13:14:43 powerbanks are weird, they stop VBUS when no load Oct 16 13:14:54 * buZz points at selfmade Oct 16 13:15:01 hmm Oct 16 13:15:08 this one never stops to power the outputs Oct 16 13:15:50 sorry can't speculate Oct 16 13:15:59 thats ok :P Oct 16 13:16:07 you could trace what's going on Oct 16 13:17:35 to make it more challenging, this whole shit changed somewhere around PR1.2->PR1.3 and then again with Pali's open "BME" implementation Oct 16 13:17:54 so odds are when you got powerkernel, no charging at all is what you see Oct 16 13:18:23 depending on whether or not that bq27xx.ko and friends are loaded Oct 16 13:18:53 oh i might actually have powerkernel yeah Oct 16 13:19:11 doc, can you find microusb based ac-2e charger? Oct 16 13:19:32 wouldn't even know what that is Oct 16 13:19:56 just google nokia ac-2e Oct 16 13:20:04 he has that charger and uses it Oct 16 13:20:24 he said the charger says 500mA Oct 16 13:20:31 don't need toknow more Oct 16 13:20:46 yes, but all chargers i find use barrel plug Oct 16 13:20:54 what? Oct 16 13:21:02 never seen any Oct 16 13:21:48 pity old nokia pages are already nuked Oct 16 13:22:00 aah wait, you say it's a fake ac-2e with micro-USB instead of barrel plug at cable end? Oct 16 13:22:03 possible Oct 16 13:22:15 ac-2e might be original, just cable changed Oct 16 13:22:18 ie. some refurb Oct 16 13:22:22 :nod: Oct 16 13:23:06 500mA USB chargers are very unusual Oct 16 13:23:23 always have 70mA minimum, now 1000 Oct 16 13:23:28 700mA* Oct 16 13:23:34 and maybe he was able to charge before, because he wasnt on cssu Oct 16 13:23:44 yes Oct 16 13:23:55 which means he is going to have flatbat soon Oct 16 13:23:56 ;) Oct 16 13:24:45 yeah, with 13% SoC Oct 16 13:25:10 that's why I advised to not do lots of futile messing to download and install stuff Oct 16 13:25:12 there is a chance his wife samsung charger would work Oct 16 13:25:44 or he just uses PC to charge Oct 16 13:26:25 shouldnt that warning be added to cssu page? Oct 16 13:26:34 which will work just fine unless you installed some kernel modules that make the whole thing bootloop ;-) Oct 16 13:27:01 shouldn't that get investigated and fixed in CSSU rather? Oct 16 13:27:11 that's also fine idea Oct 16 13:27:32 but i guess bme is pali's area of competence Oct 16 13:27:40 add bogus warnings based on a hypothesis... tzzz Oct 16 13:28:14 no, actually BME is my genuine competence. Pali's domain is his BME replacement implementation Oct 16 13:28:16 not so bogus, we have seen one sample today ;) Oct 16 13:28:28 do we? Oct 16 13:28:30 Just about ready to upload a bunch of stuff to cssu-devel (a lot of which I haven't tested and dont know if it works) Oct 16 13:28:40 how do you explain his almost empty device? Oct 16 13:28:51 you don't even have any data if the thing charged at 100mA or not at all Oct 16 13:29:06 which is why I want to check with fmg before I upload some of it just to see if he thinks its OK to upload or not Oct 16 13:30:26 kids today, saying the phrase then leaving Oct 16 13:30:29 ;) Oct 16 13:30:44 WLAN on - no powersave - 200 mA; Backlight on max +150 mA Oct 16 13:31:55 you wouldn't know the difference whether it's charging @100mA or not at all. So please don't suggest to add bogus warnings Oct 16 13:32:46 but note, he claims it was working fine before installing cssu Oct 16 13:32:53 and i suspect he was actively using it Oct 16 13:33:04 rather consider how to do proper investigation on your own device Oct 16 13:33:14 i have no fake chargers Oct 16 13:33:23 no bad cables either Oct 16 13:33:31 then don't do fuzzy conclusions Oct 16 13:35:51 when I hear "I suspect... (it worked...)" I always answer "I suspect there are other explanations too" Oct 16 13:36:26 got any? Oct 16 13:36:31 see above Oct 16 13:37:17 I'm not wasting my time on suspecting a likely usage pattern a few years ago, or whatever Oct 16 13:37:46 usage pattern 'device left overnight with wifi on and charger connected' Oct 16 13:38:00 * DocScrutinizer05 headdesks Oct 16 13:38:30 read what I said: associated to AP (very low WLAN power consumption) and with dim backlight Oct 16 13:39:03 now please stop that Oct 16 13:40:45 you don't even know if power saving setting has chaged during reflash and upgrade Oct 16 13:41:38 and you don't know if backlight on or off Oct 16 13:41:45 he checked, it's at 'max' Oct 16 13:41:50 (wifi power save) Oct 16 13:41:55 basically we know nothing at all Oct 16 13:42:47 meh, wasting my time Oct 16 13:48:43 before: worked with unknown usage pattern (most likely WLAN turned off and backlight turned off) allegedly 3 days standby. NOW: reflashed and upgraded to CSSU, on a supposedly broken charger overnight, then in use for at least 6h and now at 13% charge -- "DAMN there is something wrong!!! We should add a warning that CSSU has a bug in charging" ... No Way! Oct 16 13:56:05 please google "powersupply foldback" for a story about how a 500mA charger could reduce voltage to a point where only a few mA would flow, when overload detected. Overload is possibly under very specific conditions when you attach a 500mA charger to N900 Oct 16 13:57:02 overload *not* likely with >>[2017-10-16 Mon 13:49:18] I routinely did partial charges (from say 50 - 60 % to 100 %) in the evenings or over night<< Oct 16 13:58:04 as a re: >>[2017-10-16 Mon 15:36:26] got any?<< Oct 16 15:34:21 http://paste.opensuse.org/15528693 Oct 16 15:35:11 seems device is a tad busy immediately after unplugging charger Oct 16 15:36:20 ...or the calculated corrected remaining energy in cell gets adjusted by bq27200 chip due to battery chemistry and temperature settling Oct 16 15:37:11 or it's just a random factor from IRC traffic to xchat ;-) Oct 16 15:49:03 aah, I know (maybe). It's probably nased on CACT x *current* VOLT and the latter changes with load thanks to battery ESR Oct 16 15:49:17 based* Oct 16 15:55:37 https://crocs.fi.muni.cz/public/papers/rsa_ccs17 Oct 16 15:55:51 is there a chance it helps break some maemo's core keys? Oct 16 15:56:01 ie. repo keys? Oct 16 16:02:07 what for? Oct 16 16:04:03 mv RSOC CSOC mA NAC CACD CACT TTF TTE TEMP Oct 16 16:04:05 17:50 4115 96 96 -41 1098 1098 1098 65535 1606 25 NOACT:0 IMIN:0 CI:0 CALIP:0 VDQ:1 EDV1:0 EDVF:0 Oct 16 16:04:06 17:55 4131 96 96 -13 1095 1095 1095 65535 4882 25 NOACT:0 IMIN:0 CI:0 CALIP:0 VDQ:1 EDV1:0 EDVF:0 #1098-1095=3mAh, per 5min= 36mA avg Oct 16 16:04:08 18:00 4131 96 96 -14 1094 1094 1094 65535 4496 24 NOACT:0 IMIN:0 CI:0 CALIP:0 VDQ:1 EDV1:0 EDVF:0 #1095-1094=1mAh, per 5min= 12mA avg Oct 16 16:09:48 KotCzarny: I doubt Nokia generated their signing keys for repos by using >>a software library adopted in cryptographic smartcards, security tokens and other secure hardware chips manufactured by Infineon Technologies AG<< Oct 16 16:10:57 anyway it would be worth lottle to break their signature Oct 16 16:11:10 little, even Oct 16 16:11:34 unless, for Harmattan maybe Oct 16 16:20:24 http://paste.opensuse.org/25948187 shows how the more on-spot measurements for mA and derived values like TTE reproduce noise from random system activity, while the delta-NAC is relatively steady (though corse granularity) at ~3mAh/5min aka 36mA Oct 16 16:21:54 this is idle device with WLAN on PSM-max with AP supporting that, locked, with xchat running in background Oct 16 16:27:16 1098 - 1080 NAC = 18mAh per 30min = 36mAh/h :-D Oct 16 16:28:14 http://paste.opensuse.org/737649 Oct 16 16:32:49 avrage over all spot readputs of current (8 points): 37mA Oct 16 17:22:01 http://paste.opensuse.org/9666482 54 * 50 / 90 = 36 Oct 16 17:22:05 :-)) Oct 16 17:22:27 s/50/60/ Oct 16 18:42:17 https://www.kb.cert.org/vuls/id/228519 hmm lol, hope we can get a wpa-supplicant update soon Oct 16 18:58:26 dang, maemo doesn't use wpa-supplicant, eh? Oct 16 19:45:49 let's see if TI conmes up with a new firmware for the WL12xx then. The benefits of non-FSF-RYF compliant firmware that does _nit_ live in a ROM in WLAN chip Oct 16 19:46:04 _not_, even Oct 16 19:47:42 [2017-10-16 Mon 21:23:44] is the communication "fixed" when only one of both has the patches? [2017-10-16 Mon 21:24:05] DocScrutinizer05: yes it is a man in the middle attack and you have to convince both sides to re-use a previous key. if one of the sides refuses, you are already safe Oct 16 19:59:39 DocScrutinizer05: that's the vuln i showed yesterday ... i didn't quite understand why you said it's a good thing Oct 16 20:01:07 n900 doesn't use wpa_supplicant .. but i won't be surprised if eapd (what we use) is just as vuln. someone's got to test Oct 16 20:02:36 from what i understood, it is not really a firmware issue, so we don't necessarily need wl12xx. Oct 16 20:10:03 sicelo: yes Oct 16 20:10:51 it's unclear where the WPA protocol gets handled on maemo Oct 16 20:10:58 to me, at least Oct 16 20:17:34 eapd Oct 16 20:18:25 you can use wpa_supplicant though, if you first stop wlancond, and can live with having some applications get confused about connectivity state Oct 16 22:22:59 damn, I closed the window and now I cant remember who fmg told me to talk to about cssu-testing Oct 16 23:30:35 merlin1991: Oct 16 23:39:35 sicelo: that sounds extremely promising Oct 16 23:40:04 jonwil: seen this?^^^ [2017-10-16 Mon 22:18:25] you can use wpa_supplicant though, if you first stop wlancond, and can live with having some applications get confused about connectivity state Oct 16 23:44:49 I suspect the answer is to figure out what external interfaces are exposed by the wlan system (wlancond, eapd, icd plugins, related connectivity UI bits) then figure out who else needs those interfaces and go from there. It probably isn't that hard, even the gconf keys could be figured out fairly easily with a bit of work in IDA Oct 16 23:45:28 Then you could easily rip the whole subsystem out (including the ICD plugins and related UI bits) and replace it with something else based around wpa_supplicant Oct 16 23:45:59 I am busy getting things into cssu-devel and also cleaning up my pigsty of an apartment so I wont be doing that right now Oct 16 23:48:12 If you replace things like connui-conndlgs-wlan and connui-iapsettings-wlan as well as the lower level bits it becomes even easier (a lot of the gconf keys for example are only used by that stuff and don't get used outside the wlan subsystems) Oct 16 23:48:53 Bonus is that once this work is done, all you need is a new GPRS plugin and you could replace ICD completly with something new without much effort Oct 17 00:38:55 ack **** ENDING LOGGING AT Tue Oct 17 03:00:02 2017