**** BEGIN LOGGING AT Tue Oct 17 03:00:03 2017 Oct 17 03:24:08 great, the site I am upliading CSSU-devel bits to has just gone down :( Oct 17 03:26:04 now its back :) Oct 17 04:19:01 fmg is already doing some RE work around connui Oct 17 04:59:56 Yeah I am aware Oct 17 05:00:11 but that doesn't necessarily include the wlan bits Oct 17 05:28:04 ok, that's done everything I can for cssu-devel updating Oct 17 05:29:04 Some packages wouldn't compile properly and a couple wouldn't upload (need to find out from the right people how to get them to compile or upload properly) but all the ones that worked are in cssu-devel now Oct 17 05:30:23 http://talk.maemo.org/showthread.php?p=1537037#post1537037 Oct 17 07:17:44 jonwil: what's new in microb-engine Oct 17 07:18:20 https://github.com/community-ssu/microb-engine/commit/ac5fdce5bc231334ca616461377daa7b8e2cfa63 Oct 17 07:18:24 That's the relavent commit Oct 17 07:18:33 thanks. Oct 17 07:19:05 Its a workaround for an issue with Google that Google knows about but hasn't fixed Oct 17 07:22:23 But yeah I have put everything into cssu-devel that I can (a couple things that I want to put there but cant compile or cant properly upload) Oct 17 08:39:51 DocScrutinizer05: I have an update. Sorry I wasn't around last night, I had a long workday Oct 17 08:40:16 The battery is fine. I tested the phone with a standard USB charging device. It was fully charged in three hours :) Oct 17 08:40:21 But there are some issues Oct 17 08:40:58 One is that, when I disconnected the phone, I got the prompt asking me which USB mode I want to use, mass storage mode or pc-suite mode Oct 17 08:41:07 note: this was *after* disconnecting Oct 17 08:41:40 After this, the phone thought it was connected to USB in mass storage mode and there seemed to be no way to turn this off Oct 17 08:41:49 This meant I had no access to stored files Oct 17 08:42:09 The other issue I've noticed has to do with the FM radio function Oct 17 08:42:49 I installed the "C FM radio" app. I couldn't find the "FM radio" app, it must be in a repo I haven't added Oct 17 08:43:04 when I turn it on, it works -- but it also turns on Bluetooth Oct 17 08:43:51 If I turn Bluetooth off, the radio stops working. The app runs, but I don't hear anything. Turning Bluetooth on again has no effect. Nor does restarting the app. The only solution is to reboot the phone. Oct 17 08:44:18 Is this normal behaviour? Are the Bluetooth and FM receivers linked somehow? Oct 17 08:44:50 So -- in both cases, the only solution was to reboot the phone. Doesn't sound right to me. Oct 17 08:44:53 End of update :) Oct 17 08:45:34 bluetooth/fm sit on the same chip Oct 17 08:45:59 bencoh: so they both have to be on for FM to work? Oct 17 08:46:10 regarding usb you have to select "pc-suite" if you want to access user data with usb plugged in Oct 17 08:46:34 but the fact that it prompted you after disconnecting is strange Oct 17 08:46:38 bencoh: ok, that makes sense, but the primary issue is that I am only presented this prompt when I disconnect the phone from the charger Oct 17 08:46:46 bencoh: yes, strange :) Oct 17 08:46:55 well, unless ... Oct 17 08:47:05 bencoh: it's possible it was in the background when I plugged in the phone Oct 17 08:47:15 and then jumped to foreground when I disconnected Oct 17 08:47:20 I would have to experiment with it Oct 17 08:53:24 bencoh: well, unless? Oct 17 08:55:43 What I also find strange is that it says "charging only" -- which implies that nothing is mounted. And yet I can't access storage. This was a dumb charger anyway, there's nothing (at least, that I know of! Hello NSA) in the charger that could mount the storage. I just want the juice... Oct 17 09:09:02 is that behaviour repeatable? Oct 17 09:09:28 maybe the charger you use simulates (or has) computer inside? Oct 17 09:09:58 also, your ac-2e seems fake, because it is barrel plug, not microusb Oct 17 09:10:09 so someone either modified or refurbished it Oct 17 09:15:50 i have same issue sometimes. I think it matters of the angle in which u insert charger Oct 17 09:16:08 uh, oh, sounds like a recipe for broken usb port Oct 17 09:16:12 solution is to plug charger to phone and then to mains Oct 17 09:17:17 also if ur phone thinks it is connected to charger or usb host. U can use usb-host-mode to reset the state Oct 17 09:17:55 but in my case i've had that prompt only when inserting charger Oct 17 09:22:07 KotCzarny: Have you seen my AC-2E? Oct 17 09:22:22 nope, but every ac-2e in google has barrel plug Oct 17 09:23:02 KotCzarny: I don't quite understand how it follows from that that it is a fake :) Oct 17 09:23:06 KotCzarny: It is probably just old Oct 17 09:23:14 one of my n900 came with barrel to micro-usb adapter Oct 17 09:23:27 fake as in 'somone had a charger and replaced the cable' Oct 17 09:23:37 Remember, the legislation to require USB charging was only introduced in 2011 Oct 17 09:23:44 KotCzarny: No. Oct 17 09:24:04 KotCzarny: The design is totally consistent with prior Nokia charger designs Oct 17 09:24:12 would be interesting to see what is resistance between data pins on it Oct 17 09:24:30 Yes, I need to start equipping my toolbox, starting with a good multimeter Oct 17 09:24:50 maybe Nokia just had left over stock of those charger bodies and just used them with new cables? Oct 17 09:25:44 Vajb: new cables? It's just a Nokia standard barrel plug. It goes into a Nokia branded adapter block, which has a micro USB connector on the end Oct 17 09:26:22 sure, but when n900 was released standard was there already, and i'm almost 100% none of n900 were shipped with anything weaker than 900-1000mA Oct 17 09:26:35 There's nothing unusual about this, the manufacturers were legally required (in Europe) beginning in 2011 to provide standardized USB charging capability. You're not just going to throw out all of your old stock. Oct 17 09:27:31 KotCzarny: I have a complete original box with all the parts including the operating manual, warranty card, etc. When I get home tonight, I'll look and see if there is a packing list. Oct 17 09:27:58 and there is a chance that charger is noncomformant to it, hence, with updated bme-replacement in cssu it set max charging to 100mA Oct 17 09:28:33 KotCzarny: That is very possible. It was indeed strange that it coincidentally stopped working after the update. Oct 17 09:29:01 did you try your wife's charger? Oct 17 09:29:02 I have to admit it struck me as peculiar that the charger was so small for such an obviously power hungry phone Oct 17 09:29:35 KotCzarny: Yes, she has a generic white USB charger that she uses for charging her iPhone and iPad Oct 17 09:29:57 so at least you can use it as a fallback Oct 17 09:30:01 :) Oct 17 09:30:01 It worked wonderfully, the phone was charged in three hours. Apart from the issue with the USB mode locking my storage, of course :) Oct 17 09:30:12 KotCzarny: Yes, I'm going to purchase a similar charger Oct 17 09:30:23 didn't ianything have it's own istandard... Oct 17 09:30:49 I have seen two N900 with the Nokia barrel to uUSB adaptor, I don't know of a Nokia barrel charger that outputs 1A. Oct 17 09:30:50 Vajb: it's not an Apple charger Oct 17 09:31:01 ah Oct 17 09:31:03 Vajb: also, standards compliance for chargers is required by law Oct 17 09:31:08 Standards only work if companies adhere to them. Oct 17 09:31:10 charger with many plugs? Oct 17 09:31:22 Vajb: charger with one USB port :) Oct 17 09:31:29 swb: ac-4e (or ac-4u) for example Oct 17 09:31:41 890mA with barrel plug Oct 17 09:31:41 ah of course Oct 17 09:31:42 sixwheeledbeast: yes, the charger is small and lightweight. It did strike me as strange. Oct 17 09:32:01 (from my n800) Oct 17 09:32:14 sixwheeledbeast: standards compliance mandated by law is different :) Oct 17 09:32:32 in any event, the charger works both for her iPhone and my N900 :) Oct 17 09:32:44 (afk) Oct 17 09:33:01 i suspect that barrel-to-usb adapter might have something on data pins Oct 17 09:33:31 i can try and measure Oct 17 09:33:45 i have multimeter plus said adapter Oct 17 09:34:56 you might try to open it too if you arent attached to it Oct 17 09:36:04 oh yes, wonderfull way to split my fingers too =) Oct 17 09:37:04 for science! Oct 17 09:37:06 ;) Oct 17 09:37:12 i'll go by measure first Oct 17 09:37:31 i just need to search for schematics for the plug Oct 17 09:38:06 i vaguely recall that furthest were data + and -,but... Oct 17 09:38:50 http://pinouts.ru/PortableDevices/micro_usb_pinout.shtml Oct 17 09:38:59 or just http://pinouts.ru/visual/micro_usb.jpg Oct 17 10:05:13 Vajb: how do I reset usb mode from the console? In case I have this hangup business again. Or do I just use mount? Oct 17 10:05:15 (afk) Oct 17 10:53:42 USB host detected when UNplugging happens sometimes. It's a race. Just plug+znplug again Oct 17 10:58:59 rhombus: you should have mentioned the barrel->usb adapter earlier :-P Yes they came with N900, and they are deprecated since they can't do the needed 900mA and _may_ catch fire X-P Oct 17 10:59:34 oh, so he has barrel adapter? Oct 17 10:59:36 o.O Oct 17 11:00:05 where did he mention that? Oct 17 11:00:46 but I can see how KotCzarny gave you no chance to mention the edapter, being too busy to suggest nonsense to you Oct 17 11:01:32 it was vajb mentioning it Oct 17 11:01:46 heh you two Oct 17 11:02:58 I mentioned yesterday that I have an AC-2E Oct 17 11:03:05 which I believe came only with a barrel connector Oct 17 11:05:16 [2017-10-17 Tue 11:25:44] Vajb: new cables? It's just a Nokia standard barrel plug. It goes into a Nokia branded adapter block, which has a micro USB connector on the end Oct 17 11:06:01 must have been flooded by output from other channels, oh well Oct 17 11:06:07 still, one mystery less Oct 17 11:06:19 for FM: you need wired headset plugged in to have an antenna Oct 17 11:06:49 KotCzarny: or flooded by own posts Oct 17 11:10:36 and I can only hiess how miuch noise came from users that are on ignore here Oct 17 11:10:41 guess* Oct 17 11:10:55 DocScrutinizer05: I had a wired headset plugged in :) Oct 17 11:11:14 then please re-iterate the problem Oct 17 11:11:24 I failed to get it Oct 17 11:11:47 rhombus: http://maemo.org/packages/view/fmradio/ Oct 17 11:11:52 it's in extras Oct 17 11:12:14 DocScrutinizer05: I installed C FM radio. When I start the app, it turns Bluetooth on Oct 17 11:12:28 normal Oct 17 11:12:29 I can hear radio Oct 17 11:12:37 fine Oct 17 11:12:38 If I turn Bluetooth off, then the radio goes silent Oct 17 11:12:43 normal Oct 17 11:12:53 If I turn Bluetooth back on... no effect. Radio is still silent. Oct 17 11:12:59 normal Oct 17 11:13:05 Closing and reopening the app does not help Oct 17 11:13:14 the only thing that makes radio audible again is rebooting the phone Oct 17 11:13:34 Ok, so you are saying that the FM receiver only works if Bluetooth is on. Oct 17 11:13:34 yeah, not exactly normal but somewhat expected Oct 17 11:13:54 you're not supposed to shut down BT when FM Oct 17 11:13:58 That's unfortunate. Bluetooth consumes power Oct 17 11:14:01 they share one chip Oct 17 11:14:10 no, not realy Oct 17 11:14:27 Ok, well, if I know that, then I will leave it on next time Oct 17 11:14:44 But it shouldn't require a reboot of the phone to restore normal operation :) Oct 17 11:15:06 the audio routing is a tad fragile Oct 17 11:15:23 Was this always the case, or does it have to do with CSSU? Oct 17 11:15:48 when you reboot with wired headset, or plug in wired headset while BT headset enabled, you may see weird results Oct 17 11:16:11 same for shutting down BT while wired headset plugged in Oct 17 11:16:38 it's a sw flaw Oct 17 11:16:58 always was like that Oct 17 11:18:23 Ah. (Just fyi, I was not using a BT headset :) ) Oct 17 11:18:30 (afk) Oct 17 11:19:04 but usually that's no problem. The real flaw is that it shows "BT enabled" when you only start radio, since what it does is power up the chip but not exactly enable BT afaik Oct 17 11:21:37 bt is enabled in the real sense :-) Oct 17 11:21:49 possible Oct 17 11:22:02 I never *really* inverstigated Oct 17 11:22:11 it is Oct 17 11:22:26 pycage evidently also got inside info (from nokia) Oct 17 11:22:38 not good in the light of recent bt exploits Oct 17 11:23:42 you turn off the radio is you want to be safe :-) Oct 17 11:23:42 what's pycage? Oct 17 11:24:16 and how is it related? Oct 17 11:24:27 Martin Grimme, iirc Oct 17 11:24:35 hm? Oct 17 11:25:38 i believe he was the first to develop fm radio application for N900 Oct 17 11:27:13 tl;dr .. the radio & bt are on same chip. hence the need for bt to be on when fm radio on Oct 17 11:27:52 that's what I said, yes Oct 17 11:31:08 rhombus: you seem to have a very early device given you got that adapter. You should find out about your hw version (2010, 2101...) in case you need it Oct 17 11:35:19 rhombus: flasher can tell you, and did tell you Oct 17 11:35:40 though I bet there are other ways to find out about hw rev Oct 17 11:41:21 anyway Neo900 has FM RX on FM TX chip Oct 17 11:43:28 which seems more natural as well Oct 17 11:45:48 DocScrutinizer05: makes sense Oct 17 11:46:51 I didn't save flasher's output, unfortunately Oct 17 11:47:57 I have to stay focused, so I'm going to sign off for now, but perhaps I will be on in the evening. Thanks for the help thus far. Oct 17 11:48:36 http://paste.opensuse.org/98288505 Oct 17 14:30:57 hey DocScrutinizer05 Oct 17 14:31:01 * timeless has an unfortunate problem Oct 17 14:31:14 i made the mistake of using a non low numbered uid for a system service Oct 17 14:31:29 it turns out that systemd gives special protection to lower numbered uids Oct 17 14:31:41 remove systemd, problem solved Oct 17 14:31:45 yeah, thanks Oct 17 14:32:04 so, i'd like to transition from the non lower numbered uid to a low numbered uid Oct 17 14:32:17 but i don't want the system to have much downtime if i can avoid it Oct 17 14:32:50 assuming that task has a distinct uid+gid pair that's only used for itself, and that all files are 066x Oct 17 14:33:31 can i cheat by changing the uid to the new low number, leaving the service as a member of the old high number gid, and then chown all files to the low numbered gid? Oct 17 14:33:44 (restarting before the chown) Oct 17 14:34:16 on a normal system it could work, but who knows what systemd would do.. Oct 17 14:34:36 and then similarly change the service to default to the corresponding low numbered gid (but still being a member of the high numbered group), restart, and then chgrp ... ? Oct 17 14:34:54 does that seem like the "simplest" approach? Oct 17 14:35:15 perhaps not "simplest" but a way that's easy to understand and should limit downtime for my system Oct 17 18:09:35 hello, has somebody checked if n900 is affected by current security problem found in wpa implementation handling? Oct 17 18:10:27 Maemo on N900 does not use wpa_supplicant, but instead use two daemons, wlancond and eapd, first implements wpa part and is open source, second implements eap part and is closed Oct 17 18:10:54 word is, if they kept to specs, it's vulnerable Oct 17 18:11:14 fix is backward compatible with spec Oct 17 18:11:30 just to refuse reusing same nonce Oct 17 18:11:41 here is source of wlancond: https://github.com/community-ssu/osso-wlan Oct 17 18:13:53 basically wifi on Maemo is handled differently as in whole Linux world, Nokia wrote own software for it... Oct 17 18:14:24 as did cisco. and they are vulnerable Oct 17 18:14:40 same with *bsd folks Oct 17 18:35:52 Pali: we discussed it yesterday. It wasn't clear where WPA gets handled, yesterday conculsion was it must be in eapd or osso-wlan-security Oct 17 18:37:29 eapd = osso-wlan-security Oct 17 18:37:54 wlancond = osso-wlan Oct 17 18:40:26 see what jonwil said Oct 17 18:40:47 maybe replace that stuff by wpa-supplicant et al then? Oct 17 18:42:09 my personal take on it: should be easy enough to just replace osso-wlan-security by wpa-supplicant Oct 17 18:43:23 probably now (after 7(?) years?) wpa-supplicant is better and more versatile than osso-wlan-security, unlike back when Nokia made osso-wlan-security Oct 17 18:44:34 I seem to recall wpa-supplicant was just a pile of segfaulting Sh* back when Oct 17 18:45:42 which is most likely _one_ of the reasons why Nokia cooked their own Oct 17 18:47:02 also WOW @ https://github.com/community-ssu/osso-wlan - I thought that was a blob too Oct 17 18:48:33 HMMMM https://github.com/community-ssu/osso-wlan/blob/master/wpa.c Oct 17 18:50:34 gosh!!! I *HATE* how github (in konqueror) eats one i5 core for 6s to render any scrolling Oct 17 18:50:48 JS crap Oct 17 18:57:50 you know I also hate stuff like value = key_mgmt != 0 || (encryption & WLANCOND_ENCRYPT_METHOD_MASK) == WLANCOND_WEP; Oct 17 19:01:22 immediately fllowed by the other extreme: http://paste.opensuse.org/68253426 Oct 17 19:09:10 hmm, no real crypto stuff in https://github.com/community-ssu/osso-wlan/blob/master/wpa.c Oct 17 19:11:21 neither protocol stuff Oct 17 19:11:32 aiui Oct 17 19:36:33 crypto is done in kernel/hardware Oct 17 19:36:47 you should just provide aes or wep keys to kernel Oct 17 19:37:02 just protocol is in userspace Oct 17 19:45:15 good to have you on board Pali :) Oct 17 19:45:52 i have been trying to see if the tools for testing are released yet .. doesn't seem to be the case. or i missed something? Oct 17 19:47:33 https://www.krackattacks.com/#tools Oct 17 20:12:02 so easiest would be to wait until check tool is released Oct 17 20:12:23 and then decide if some steps are needed or not Oct 17 20:12:35 or locale which part is responsible for wpa Oct 17 20:12:58 I would bet it is this file: https://github.com/community-ssu/osso-wlan/blob/master/dbus-handler.c Oct 17 20:13:38 there is a function named clear_wpa_keys() Oct 17 20:37:21 Greetings Oct 17 20:38:35 DocScrutinizer05: Question-Machine is back Oct 17 20:42:06 ohhai Oct 17 20:42:37 just in time :-) a 5 minutes back I wasn't on kbd Oct 17 20:43:10 (( just protocol is in userspace(( where exactly? Oct 17 20:43:20 s/((/))/ Oct 17 20:43:21 DocScrutinizer05 meant: )) just protocol is in userspace(( where exactly? Oct 17 20:43:37 or somesuch Oct 17 20:44:19 ~+uptime Oct 17 20:44:19 - Uptime for purl - Oct 17 20:44:20 Now: 32d 23h 34m 3s running infobot 1.5.4 (SVN) -- linux Oct 17 20:44:21 1: 59d 8h 41m 19s running infobot 1.5.4 (SVN) -- linux, ended Sun Nov 14 18:39:57 2010 Oct 17 20:44:21 2: 57d 3h 9m 23s running infobot 1.5.4 (SVN) -- linux, ended Fri Jun 26 20:39:27 2009 Oct 17 20:44:21 3: 36d 20h 47m 14s running infobot 1.5.4 (SVN) -- linux, ended Tue Aug 4 17:38:59 2009 Oct 17 20:44:26 WOW Oct 17 20:45:43 DocScrutinizer05: what is that? Oct 17 20:46:07 nevermind, unrelated Oct 17 20:46:15 ~botsnack Oct 17 20:46:15 :), DocScrutinizer05 Oct 17 20:46:37 ~botsnack Oct 17 20:46:38 aw, gee, rhombus Oct 17 20:46:54 ~poettering Oct 17 20:46:55 'sth is poettering' means it acts invasive, possessive, destructive, and generally in an egocentric exacerbating negative way. ``this cancer is extremely poettering'', or you look here for Linus' notion on what's poettering: http://lkml.iu.edu/hypermail/linux/kernel/1404.0/01331.html, or http://lkml.iu.edu/hypermail/linux/kernel/1404.0/01488.html, or see ~systemd cabal Oct 17 20:47:27 Probably something Swedish. Oct 17 20:47:40 So... trying to decide where to start here Oct 17 20:47:41 ~infobot Oct 17 20:47:41 infobot is, like, A program on the IRC that helps users, ask it to do something by putting a ~ and then say a command! Oct 17 20:47:50 ~fish Oct 17 20:47:50 fish is, like, FISHFISHFISH! DO THE FISH DANCE! "Give a man a fish and you'll feed him a day. Teach him how to fish and he'll feed himself for the rest of his life." This is so appropriate, instead of asking us to tell you exactly what to do, why not read some docs, then come back and ask specific questions which aren't covered?, or http://www.p-nand-q.com/maldives/rbr/fish/sf25.jpg Oct 17 20:48:10 wow, good guess. Oct 17 20:48:56 So... is there any hope that there will be a patch for the WPA2 vulnerability on maemo? Oct 17 20:49:09 :-D Oct 17 20:49:44 (( just protocol is in userspace)) where exactly? Oct 17 20:50:35 http://mg.pov.lt/maemo-irclog/latest.log.html#t2017-10-17T02:44:49 Oct 17 20:50:38 wlancond and eapd are only wlan related daemons Oct 17 20:51:22 Pali: lemme rephrase - where in kernel is WPA encryption? Oct 17 20:51:28 eapd seems do TLS+PEAP related crypto Oct 17 20:52:19 net/wireless/lib80211_crypt_*.c Oct 17 20:52:32 lib80211_crypt_ccmp.c, lib80211_crypt_tkip.c, lib80211_crypt_wep.c Oct 17 20:52:42 so seems in kernel is aes + wep Oct 17 20:52:49 isn;t that done in HW? Oct 17 20:52:55 and IIRC hw itself support aes Oct 17 20:53:37 at least I remember parameters of some wifi drivers to enable HW accel of aes Oct 17 20:54:19 Pali: sounded wierd, but indeed https://github.com/community-ssu/osso-wlan/blob/master/dbus-handler.c#L2067 Oct 17 20:55:27 ((and IIRC hw itself support aes)) that's exactly what I seem to recall too Oct 17 20:55:39 didn't find a hint though, last 2 days Oct 17 20:55:54 anyway what we need is protocol Oct 17 20:56:45 yes, need to figure out where is implemented that 4 way handshake Oct 17 20:57:10 https://github.com/community-ssu/osso-wlan/blob/master/dbus-handler.c#L2067 for (i=0;i<4;i++) { Oct 17 20:57:25 looks sort if interesting Oct 17 20:58:00 though there are 4 keys in WEP Oct 17 20:58:02 iirc Oct 17 20:58:17 tbh I got 0 clue about WPA Oct 17 20:58:54 I just know "nunce" from SIP challenge/response Oct 17 21:00:23 I didn't look into complete set of sec patches as e.g. found in UBNT updates either Oct 17 21:01:02 it's not just wpa-supplicant response#3 0 key. There are more CVE to fix Oct 17 21:01:21 just ^^^ is the worst one Oct 17 21:02:58 posting old shite, but one out of 200 may have missed it: https://www.krackattacks.com/ https://papers.mathyvanhoef.com/ccs2017.pdf Oct 17 21:10:41 rhombus: a few notes on another topic: USB charger voltage between 4.75V and 6V is fine. Often chargers use ~5.3V to compensate for voltage drop on cable Oct 17 21:10:56 has anyone installed slackware ARM onto nokia n900 and/or anyone know of a guide to do so? Oct 17 21:11:04 rhombus: charger should be capable of at least 800mA Oct 17 21:11:05 DocScrutinizer05: Ah, thanks Oct 17 21:11:41 rhombus: the Nokia barrel->usb adapter is *deprecated*, it's dangerous, starts a fire sometimes Oct 17 21:11:46 specifically, booted it from SD card via U boot Oct 17 21:12:25 regardless I got one of them in use since 5 years 24/7/365 Oct 17 21:12:28 also does anyone know of a way to pause the u-boot text when launching a kernel? like sendign a pause break key? Oct 17 21:13:22 pkill9: no, and no Oct 17 21:13:59 first possible just never heard of, 2nd not possible afaik Oct 17 21:14:08 kk Oct 17 21:14:41 closest thing I could think of: gentoo on N900 Oct 17 21:14:58 hmm yeah i might try that first Oct 17 21:15:16 then work out how to get slack working Oct 17 21:15:38 afaik slackware is ultra-conservative so I doubt it has any support for special stuff on N900 hw platform Oct 17 21:15:52 DocScrutinizer05: I'm a fan of gentoo :) Oct 17 21:16:15 I tried to start ssh server on my N900, but it fails because it can't load the host keys Oct 17 21:16:19 afaik gentoo was the first non-maemo distro on N900 Oct 17 21:16:41 ugh! Oct 17 21:16:57 DocScrutinizer05: are all the phone functions supported on gentoo? Oct 17 21:17:06 no Oct 17 21:17:20 damn Oct 17 21:17:25 DocScrutinizer05: could slack be configured to support extra hw stuff? Oct 17 21:17:30 Ok -- so to my ssh problem Oct 17 21:17:35 only maemo has complete phone support Oct 17 21:17:47 I'm not fussed about the phone stuff really Oct 17 21:17:55 I wanna use it as a tablet Oct 17 21:17:59 SHR has poor audio, otherwise sorta works Oct 17 21:18:08 I'm kinda sick of smartphones tbh lol Oct 17 21:18:19 I feel with you Oct 17 21:18:46 I opt for renaming then to "fartphone" Oct 17 21:18:47 I just figured something out Oct 17 21:18:58 Or buttphone Oct 17 21:19:10 spyphone Oct 17 21:19:38 re sshd I can't recall any problems with hostkeys Oct 17 21:19:51 DocScrutinizer05: It helps if one does it as root Oct 17 21:19:53 :D Oct 17 21:19:53 should get auto-generated on installation, I'd assume Oct 17 21:20:07 ooh yeah Oct 17 21:20:16 definitely ;-P Oct 17 21:21:39 pkill9: ((could slack be configured to support extra hw stuff?)) for sure Oct 17 21:21:45 kk cool :) Oct 17 21:22:05 I can't seem to ping the phone. Is ICMP turned off by default? Oct 17 21:22:19 possibly, yes Oct 17 21:22:22 I can ping my laptop *from* the phone Oct 17 21:22:52 though: Oct 17 21:23:00 23:22:33 jr@saturn:~>ping IroN900 Oct 17 21:23:02 PING IroN900 (192.168.4.41) 56(84) bytes of data. Oct 17 21:23:03 64 bytes from IroN900 (192.168.4.41): icmp_seq=1 ttl=64 time=78.1 ms Oct 17 21:23:15 "The password for user cannot be changed." Oct 17 21:23:30 there's no password for user Oct 17 21:23:43 How do I log into the phone via ssh, then? Only root logins? Oct 17 21:24:01 I'm not sure how to log in via user account - I always go straight to root Oct 17 21:25:02 I bet it *can* get done. Anyway you're supposed to use ssh keys and no passowrds ;-P Oct 17 21:25:28 DocScrutinizer05: I haven't used ssh much the past few years, so I'm out of practice Oct 17 21:25:58 man ssh-copy-id Oct 17 21:28:35 23:28:07 jr@saturn:~>cat `which iron900` Oct 17 21:28:36 ssh -q -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null -l root IroN900 $@ Oct 17 21:30:00 23:28:19 jr@saturn:~>grep -i iron900 /etc/hosts Oct 17 21:30:02 192.168.4.41 IroN900 iron900 Oct 17 21:32:35 http://paste.opensuse.org/51594250 Oct 17 21:34:48 rhombus: the dir and file permissions and owner mist be exactly like this, otherwise login fails with pretty unhelpful error message Oct 17 21:35:27 afaik ssh-copy-id takes care about all this Oct 17 21:36:11 also see ssh-keygen Oct 17 21:37:07 though I'm sure you got a keypair already. do you? Oct 17 21:37:34 on your PC Oct 17 21:37:53 DocScrutinizer05: I am just discovering that Oct 17 21:38:15 Warning: Permanently added '192.168.3.24' (RSA) to the list of known hosts. Oct 17 21:38:15 Authentication failed. Oct 17 21:39:16 see ^^^ Oct 17 21:39:28 https://www.startpage.com/do/metasearch.pl?query=ssh+essentials Oct 17 21:40:21 DocScrutinizer05: yes and no Oct 17 21:40:23 pick the one you like most, they all are useful and correct Oct 17 21:40:51 DocScrutinizer05: I can generate this on my laptop, yes? Oct 17 21:41:15 ooh, and of course you need the password you gave on installing sshd. for user "root" Oct 17 21:41:21 yes, sure Oct 17 21:43:02 you may ommit "passphrase" on ssh-keygen. Some think that's no good idea, others thing passphrase to unlock local ssh private key is BS Oct 17 21:43:04 DocScrutinizer05: what do I need a password for if I'm using keys? Oct 17 21:43:20 to install the keys :-) Oct 17 21:44:07 on PC run ssh-keygen and ssh-copy-id Oct 17 21:44:26 Ok, i have generated a key pair with ssh-keygen Oct 17 21:44:58 but I have to change the permissions on the N900 first before this will work at all, right? Oct 17 21:45:28 no, you just run ssh-copy-id or somesuch Oct 17 21:45:51 on my laptop Oct 17 21:46:12 Usage: /usr/bin/ssh-copy-id [-h|-?|-n] [-i [identity_file]] [-p port] [[-o ] ...] [user@]hostname Oct 17 21:46:22 how can ssh-copy-id work if normal ssh logins don't? Oct 17 21:46:36 it uses your password Oct 17 21:46:48 normal login using password needs to work for that Oct 17 21:46:59 DocScrutinizer05: well, it doesn't Oct 17 21:47:09 I get an error message Oct 17 21:47:12 ssh -q -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null -l root IroN900 $@ Oct 17 21:47:37 DocScrutinizer05: on the N900 Oct 17 21:47:41 so i have to restart ssh Oct 17 21:47:53 in your case: ssh -q -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null -l root 192.168.3.24' Oct 17 21:48:46 do on phone: `root; passwd` Oct 17 21:49:15 then Oct 17 21:49:17 in your case: ssh -q -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null -l root 192.168.3.24' Oct 17 21:50:00 on PC Oct 17 21:50:23 then ssh-copy-id Oct 17 21:50:26 on PC Oct 17 21:50:44 but the ssh server has to be running on the phone first for this to work Oct 17 21:50:46 :) Oct 17 21:50:50 yes Oct 17 21:51:04 do I have to fix any permissions on the phone? Oct 17 21:51:49 /etc/init.d/ssh start Oct 17 21:51:54 no Oct 17 21:54:05 you're in a xterm and root on phone? Oct 17 21:54:32 yes Oct 17 21:54:38 just changed the password Oct 17 21:55:03 then run /etc/init.d/ssh start Oct 17 21:55:18 (should have been done during boot, but...) Oct 17 21:55:56 then on PC, run: ssh -q -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null -l root 192.168.3.24 Oct 17 21:57:09 you may omit the "-o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null" part if your phone's IP never changes Oct 17 21:57:46 you're already past the "Warning: Permanently added '192.168.3.24' (RSA) to the list of known hosts." Oct 17 22:01:04 so what does a ssh -l root 192.168.3.24 give you now? Oct 17 22:01:11 DocScrutinizer05: ok, what was the ssh -q command for? Oct 17 22:01:26 * DocScrutinizer05 invokes ssh --help Oct 17 22:01:56 -q Quiet mode. Causes most warning and diagnostic messages to be suppressed. Only fatal errors are displayed. Oct 17 22:03:04 so it suppresses the "Warning: Permanently added '192.168.3.24' (RSA) to the list of known hosts. Oct 17 22:04:01 ok, well, nothing happen Oct 17 22:04:04 and "-o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null" suppress the quedtion "do you want to add...?" and avoid filling your known_hists file with litter Oct 17 22:04:06 happened Oct 17 22:04:23 it waited for ages and then died and gave me my local prompt again Oct 17 22:04:24 how, nothing? Oct 17 22:04:38 hmm, which command? Oct 17 22:04:47 ssh -q -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null -l root 192.168.3.24 Oct 17 22:05:04 try `ssh -l root 192.168.3.24` Oct 17 22:05:52 nothing is happening Oct 17 22:05:57 o.O Oct 17 22:05:59 it's like it's timing out Oct 17 22:06:08 weeeiiird Oct 17 22:06:35 try `ssh -l root 192.168.3.24 date` Oct 17 22:06:59 I just got a password prompt Oct 17 22:07:03 try `ssh -vv -l root 192.168.3.24 date` and paste the diagnostic output Oct 17 22:07:09 after I pinged my laptop from the phone Oct 17 22:07:19 yes, that's exactly what shall happen Oct 17 22:07:39 not after pinhing though Oct 17 22:07:40 I just entered the password, it's taking forever Oct 17 22:07:47 ok, I'm in the phone Oct 17 22:08:00 Is that the reverse lookup timeout causing that? Oct 17 22:08:08 you WLAN might be... hmmm, there are often issues with lag Oct 17 22:08:16 yes, prolly Oct 17 22:08:39 even likely Oct 17 22:08:44 my WLAN has probably been compromised by hackers Oct 17 22:09:02 ohnoes! no kidding Oct 17 22:09:22 Ok, so I am in the phone. That means I can run ssh-copy-id from the laptop now and expect it to work, if slowly Oct 17 22:09:33 yes Oct 17 22:09:53 once you're in the phone, is it still slow? Oct 17 22:10:58 DocScrutinizer05: yes, pretty slow Oct 17 22:11:06 I mean, it works, but there's significant lag Oct 17 22:11:14 then it's not rDNS, it's wlan lag Oct 17 22:11:40 change PSM in phone's WLAN settings to 'none' Oct 17 22:11:55 what's your AP brand/make? Oct 17 22:12:17 large amount of dropped packets on ping? Oct 17 22:12:26 I need to update my router. I'm running OpenBSD 6.0 It has no 802.11N support. N support was added in 6.1. Besides, I have to upgrade to 6.2 with the new WPA2 hole anyway Oct 17 22:12:52 Even G wouldn't be that slow Oct 17 22:12:53 no packet loss on the ping from the phone Oct 17 22:13:02 yeah, you need to fix router settings to make PSM work Oct 17 22:13:15 average round trip 29 ms Oct 17 22:13:25 max 1035 ms Oct 17 22:13:42 What's PSM? Oct 17 22:13:44 :) Oct 17 22:14:24 phone's WLAN goes to sleep for a second or more, in PSM, and just wakes up for a few milliseconds to losten to next router beacon that has a flah telling "please stay awake" which may be 1 or 0 Oct 17 22:14:46 to listen to* Oct 17 22:14:57 yay, I figured out how to get hildon-status-menu into GDB. You need to stop hildon-status-menu with dsmetool then you need to start it again using maemo-summoner Oct 17 22:15:25 DocScrutinizer05: where do I set that? Oct 17 22:15:27 so there are two parameters you want to fix on router;: beacon freqiency and another TTdunnowhat that must be 1 Oct 17 22:15:44 PSM, I mean Oct 17 22:16:15 on phone you - for now - go to WLAN connections, edit, clik OK until on last requester there's "advanced..." Oct 17 22:16:31 and there you set PSM to "none" Oct 17 22:17:25 power saving mode Oct 17 22:17:49 it says it will eat my children and steal all my money Oct 17 22:18:14 settings, internet connections, , edit, ok, ok, ok... Oct 17 22:18:38 damn, I accidentally disconnected the WLAN and now my ssh connection is frozen Oct 17 22:19:35 " ~. Oct 17 22:19:44 I've turned psm off Oct 17 22:19:58 I'm trying to log in again Oct 17 22:20:49 \n ~. Oct 17 22:21:05 better Oct 17 22:21:08 I didn't know that. Oct 17 22:22:20 still highly variable Oct 17 22:22:30 It is probably the router Oct 17 22:22:30 AP problem Oct 17 22:22:53 beacon frequency and TTwhat? Oct 17 22:23:04 ywell, tbh it's "variable" even for me Oct 17 22:23:11 yes Oct 17 22:23:28 and correct implementation of PSM in AP Oct 17 22:23:32 I'll have to revisit this another day, I can't be staying up until all hours again Oct 17 22:23:39 I have PSM off on the AP Oct 17 22:23:46 oooh Oct 17 22:23:58 wait, that's probably another PSM Oct 17 22:24:44 ok, the minute I stop pinging my laptop from the phone, the connection becomes excruciatingly slow Oct 17 22:25:07 yes, typical problem with PSM and AP Oct 17 22:25:29 not much that can be done on phone except disabling PSM for the connection Oct 17 22:25:33 I don't have power save mode on on the AP Oct 17 22:25:54 It's an OpenBSD router that I put together myself, there's none of that nonesense configured Oct 17 22:25:57 nonsense Oct 17 22:26:14 on AP the meaning of "power save mode" will probably be quite different Oct 17 22:26:43 you need a beacon timing of 10 per second Oct 17 22:26:47 ok Oct 17 22:26:48 to startz with Oct 17 22:26:53 ok Oct 17 22:27:05 the other parameter is "1" by default Oct 17 22:27:31 it tells "send stay-awake-flag with N-th eakon only" Oct 17 22:28:29 also a reboot on phone *might* help to reset WLAN, the PSM setting is a tad ... special on that thing Oct 17 22:29:08 ~psm Oct 17 22:29:08 psm is, like, Personal Security Manager for Mozilla, at http://docs.iplanet.com/docs/manuals/psm/psm-mozilla/index.html. You don't need to restart mozilla as noted on that URL. It also consumes 100% CPU unfortunately... lame piece of software. no longer works as of mozilla 0.7, or needed for viewing of some login sites, i.e. yahoo. install by apt-get install mozilla-psm Oct 17 22:29:13 meh Oct 17 22:30:47 you should probably get an update for your AP/router Oct 17 22:31:29 but first make sure to have beacon timing at 10Hz or more Oct 17 22:31:57 some routers have it at 1HZ or lower per default Oct 17 22:32:38 DocScrutinizer05: I'll have to see where that is set, as I said, it's a homespun OpenBSD box Oct 17 22:33:10 yeah, for now your pest option is to run a ping from a second terminal to phone Oct 17 22:33:22 that will keep connection responsive Oct 17 22:33:29 best* Oct 17 22:34:27 Interesting. I'm logged in via sftp now, and it's much faster. Oct 17 22:34:34 or maybe really reboot the phone. your sshd also wasn't started despite it should, right? Oct 17 22:35:44 did you find a working solution for your charging? Oct 17 22:37:02 ((when requester "assrage, PCsuite, charge?" pops up on unplugging USB, just plug and unplug again)) Oct 17 22:37:15 sorry, scratch that, I'm putting a file and it's excruciatingly slow Oct 17 22:37:25 ok, getting better Oct 17 22:37:28 that's weird Oct 17 22:37:33 1.4 MB/s now Oct 17 22:37:48 connection should stay alive during data transfer Oct 17 22:38:00 yes Oct 17 22:38:11 well, that's _almost_ max speed via sftp Oct 17 22:38:20 Oh, and I don't have a specific charger yet, but I am going to get one that's similar to my wife's Oct 17 22:38:44 mind you, it needs to decrypt on a CPU that's like a 300MHz P-II Oct 17 22:39:05 ok, then I can live with that :) Oct 17 22:40:15 DocScrutinizer05: the problem yesterday was that by the time I noticed the USB problem, I was nowhere near a charger or cable Oct 17 22:40:36 yeah Oct 17 22:41:12 the phone fell asleep and the sftp transfer stalled Oct 17 22:41:15 I had to wake it up Oct 17 22:41:18 it was slow waking up Oct 17 22:41:21 o.O Oct 17 22:41:54 I'm doing ssh and sftp all the time while device is next room sleeping on charger Oct 17 22:42:17 Well, I'm sure that there's lots I have to configure on the phone Oct 17 22:42:35 not really Oct 17 22:42:46 most thinks should "just work" Oct 17 22:43:15 but phone getting irresponsive on large amounts of data to write to flash is a known annoying nasty effect Oct 17 22:43:39 you're easily entering swap hell with copy of large files Oct 17 22:43:43 DocScrutinizer05: ok, then that is the explanation, I'm uploading 85 MB podcast files to the phone Oct 17 22:44:53 should work, just take a while Oct 17 22:46:44 never forget about partitions/volumes: you only got a 2GN minus X on your ~user dir, root has no homedir at all where you could store large files, since it's on rootfs. and the true mass storage is ~user/MyDocs Oct 17 22:46:55 2GB Oct 17 22:47:14 That's where I was pushing the files to Oct 17 22:47:29 ~user/MyDocs Oct 17 22:47:31 I hope you don't store those 85MB to root's home? Oct 17 22:47:35 no Oct 17 22:47:36 :) Oct 17 22:47:40 aah ok :-) Oct 17 22:47:42 ~user/MyDocs Oct 17 22:47:46 good Oct 17 22:48:00 is that on the microSD? Oct 17 22:49:23 ok man, I gotta go to bed. That's a bit of progress for today. I'm starting to really enjoy the phone Oct 17 22:49:33 listened to the radio on the train today... some podcasts Oct 17 22:49:41 it's working nicely Oct 17 22:50:04 I'll get a new charger shortly, fix my WLAN, etc. Oct 17 22:50:10 Chat soon Oct 17 22:59:00 Looks like we have another person who doesn't heed the warnings and has just installed all the stuff I uploaded to cssu-devel and is now wondering why his device bootloops... Oct 17 23:19:55 I am not surprised. Won't be the first or the last Oct 17 23:20:42 My guess is one of the recent clones by Freemangordon is what's causing the boot loop (either his cloning or my compile/build) Oct 17 23:21:04 I need my N900 too much to try installing any of that stuff to narrow down the culprit Oct 18 00:01:39 ~ping Oct 18 00:01:39 1 packet transmitted, 1 packet received, 0.0% packet loss Oct 18 00:02:02 ~ping Oct 18 00:02:02 ~pong Oct 18 00:38:15 well, DO NOT USE CSSU-DEVEL AS REPO! (unless you're a developer of a package you want to test there) IT **IS** BROKEN Oct 18 00:39:35 we should upload a package to cssu-devel that displays this text, the erases the complete device Oct 18 00:39:49 then* Oct 18 00:40:05 I'm tempted to actually do Oct 18 01:25:17 probably the root problem is CSSU-devel working like a repo at all. We should change each single package there in a way so you have to edit catalogs and add a new entry for each. Like, instead of "free" use "$packagename" Oct 18 01:25:34 * DocScrutinizer05 ust wonders where from users have the idea to add cssu-devel as catalog (and then, worse, do indiscriminate updates or install packages they don't know not thoroughly test status of). Wouldn't know where to find the repo settings needed for cssu-devel even if he planned to use it Oct 18 01:25:49 * DocScrutinizer05 blames some tmo post Oct 18 01:33:59 indeed replacing category(?) name ("free") by ${packagename} via a script on server sounds like an excellent idea to me. Serves exactly the purpose cssu-devel is meant for: allowing test of package installation from a repo, but only when the person installing it has explicitly opted to do so by filling in the "credentials" manually - not grouping packages into one catch-all procedure. And when some smartass makes a tool for automating Oct 18 01:34:01 that, we can change to ${packagename}_$(date) Oct 18 01:36:01 dynamically, I.E date of the day it gets accessed **** ENDING LOGGING AT Wed Oct 18 03:00:00 2017