**** BEGIN LOGGING AT Tue May 02 03:00:02 2017 May 02 05:02:35 Vajb: unfortunately we will not get any further updates . . . irssi 1.x needs a newer glib than we have on Maemo May 02 05:08:36 Hi all :) I'm looking for bitcoin wallet for n900 May 02 05:22:57 19:25 < vahe> hi all :) I'm looking for bitcoin wallet for n900 May 02 05:22:57 19:55 < DocScrutinizer05> https://talk.maemo.org/showthread.php?t=66756 May 02 05:26:58 sicelo: that wont be a problem since mainlining is done, right? May 02 05:27:19 tho i have a feeling that thing or two are yet to be done May 02 05:39:32 huhm. since yesterday my n900 often doesn't see the sim. tried with some paper between sim and battery, but no luck. any tips? May 02 05:41:04 ok, bending the sim-pins a bit upwards. see how that goes .. May 02 05:43:41 so far so good .. May 02 05:43:47 eh. nope May 02 05:43:55 on boot it's there. 20 secs later it's gone May 02 05:44:07 oh now it's back .. highly unstable then .. May 02 05:57:21 aaahnd it's gone May 02 05:58:23 happens to me too once in awhile May 02 05:59:30 I read some topic that mentions firmware being the issue. highly improbable May 02 05:59:39 this worked fine until last weekend May 02 06:00:42 probably gsm modem is coming loose from the board May 02 06:01:03 there was some spring fix for that, but didn't try it out May 02 06:01:13 hmz May 02 06:02:15 u could try "pnatd AT+CFUN=0" instead of rebooting whole device May 02 06:02:39 it restarts the modem chip or something like that May 02 06:02:56 also u can try to switch to offline and back May 02 06:05:13 atm it's just keeps switching off/on/off/on May 02 06:16:31 The sim icon is kinda misleading May 02 06:16:40 It's the whole modem blinking out of existence May 02 06:21:14 sicelo: thanks May 02 06:24:05 yeah, modem might be getting loose from th board May 02 06:24:39 check if it happens if you dont touch anything (especially kb) May 02 06:28:17 I've killed 3 or 4 N900 that way .. it always starts disconnecting when typing f and j or in between there May 02 06:31:05 hmz, this is my 3rd n900. 1st got fried usb port. 2nd got 'lost' (seriously no idea what happened there hehe) and this one I only had a month or 2 May 02 06:31:21 but. hmm. so do I find a rework station and try heating the modem part? May 02 06:31:32 maybe it was already faulty, and previous user 'fixed' it by baking? May 02 06:31:43 nah, previous user was a noob May 02 06:31:51 and it only started acting up since this weekend May 02 06:31:58 could be selling it because of that May 02 06:32:18 it's in otherwise pristine condition May 02 06:32:21 ie. problem could be happening from time to time and he decided to try to cash it while its possible May 02 06:34:28 for me it happens from time to time. I hope it is just area specific thing... May 02 06:37:16 it happened to me once or twice with worn sim card May 02 06:37:33 so there is a chance that sim card itself is dirty May 02 06:53:34 KotCzarny: wasn't a he. some middle-aged lady May 02 06:53:47 :) May 02 06:53:55 simcard is also new btw May 02 06:54:03 tried cleaning it. no avail May 02 06:54:09 so, rework station? May 02 07:06:31 if you are skilled enough May 02 07:19:49 btw. surest sign of modem fault is 'error communicating with the modem' yellow popup at the boot time May 02 07:41:15 dreamer: there are some 'easy' fixes May 02 07:41:50 as in, make sure it's not the sim connection, and check what dmesg says, and you can, put a small folded paper in between the phone and the battery (if you think the modem connection is the issue) May 02 08:10:10 vahe: fyi, native bitcoin core supports arm nowadays, no need for untrusted binaries ;) May 02 08:10:24 vahe: but without spv , i wouldnt really run it on n900 May 02 08:10:34 there is https://github.com/bitcoin/bitcoin/pull/9483 though May 02 08:12:01 buZz: Yes, I already understood , thanks :D May 02 08:12:08 :) yw May 02 08:12:11 :) May 02 08:13:10 buZz: I was just looking for a wallet for your phone May 02 08:13:30 but didn't think it was so hard for this phone May 02 08:13:51 well its older than bitcoin :P May 02 08:14:06 took a while for ppl to realize what bitcoin was etc May 02 08:14:46 yup :) May 02 08:16:11 there is https://multibit.org/ which is just java May 02 08:16:23 it could work, although you might struggle with interface a bit May 02 08:18:56 I'm on my computer use electrum May 02 08:19:10 Wizzup: already did the folder paper thing. also bended the pins a bit up. pretty sure it's not the connection to the sim May 02 08:19:32 buZz: https://electrum.org/ May 02 08:20:27 vahe: i think someone did make a maemo port of that at some time May 02 08:20:37 but doubt it kept up with recent changes May 02 08:23:10 dreamer: the paper helps not just with the sim connection May 02 08:32:16 buZz: https://bpaste.net/show/12eadef92c96 May 02 08:32:32 :D May 02 08:41:00 well yes :P May 02 09:00:58 Wizzup: well, it didn't help at all :) May 02 09:04:28 dreamer: aw May 02 10:35:10 ~tell dreamer about springfix May 02 10:36:11 "paper between sim and battery" doesn't help anything May 02 10:36:40 "bending the sim-pins a bit upwards" is risky and doesn't help either May 02 10:36:44 Vajb: irssi needs your testing and voting :) - http://maemo.org/packages/package_instance/view/fremantle_extras-testing_free_armel/irssi/0.8.21-1/ May 02 10:41:00 when the SIM symbol is initially there but vanishes after ~20s just to return later on, it's highly likely the problem is faintly related to TX, so the modem detects problems as soon as trying to register to a cell it found. Similar effect known to happen on mountain summits when phones have clear and good reception but the station is too far away to allow login. May 02 10:41:54 another possible issue is very old battery which causes brownouts during TX bursts, ruining the TX signal to a degree where all connections fail May 02 10:42:22 but that's a condition the modem should detect and report separately May 02 10:47:59 the only real effect of paper cushion for SIM is: it supports the PCB slightly, so the PCB can bend less when pressing keys on kbd May 02 10:48:43 dreamer: ^^^ May 02 10:50:31 rework station is a good approach, when the operator is experienced May 02 10:51:04 ~usbfix May 02 10:51:04 it has been said that usbfix is http://talk.maemo.org/showthread.php?t=75920 - and **NEVER** use epoxy (unless you want to seal your device for underwater), or https://www.youtube.com/watch?v=fYz5nIHH0iY#t=1866, you will basically need two irons: a small good one (or better hot-air reflow) and a 60+ Watt May 02 10:51:10 while you're at it May 02 10:52:11 sicelo: i don't have account there. Failed o activation. I reported it here and forgot the whole thing. May 02 10:52:28 in acti... May 02 10:53:23 when you're "baking" the device, you must remove all heat sensitive items before: camera, domesheet (*carefully* peel off and store on a "wax" paper), plastic frame on uSD board May 02 10:53:26 ) May 02 10:54:35 better use a professional rework station and reflow the modem chipset only. You need to remove tha can prior to that May 02 10:55:26 but before you do that, clean the antenna contacts of GSM antenna May 02 10:56:11 and make sure you got a battery with low ESR aka impedance, colloquially "a new fresh battery" May 02 11:01:59 * dreamer never had the usb issue May 02 11:02:07 but hmm. will see if I can use a rework station one of these days May 02 11:12:01 usbfix is baby easy as long as USB is not broken yet. Once it's broken you will be sorry you didn't reinforce it beforehand May 02 11:17:13 actually for USB fix you don't need anything but a screwdriver or knife to scratch away the green solderstop varnish, a decent 60W soldering iron and good quality brand 60/40 flux core solder May 02 11:20:21 if you can solder two wires together and are sufficiently skilled to erase a writing from paper with a razorblade/knife, you can do USB fix May 02 11:21:55 reflow fixing the "no SIM" modem defect is a whole different story May 02 11:24:41 also make sure it's not actually a problem of your network providers. E.g. in some parts of USA I could see how N900 slowly ceases working at all, simply due to discontinued support for any of the frequency bands it could use May 02 11:25:56 a "fails after 20s" could also mean "the device thinks there's a band/network suitable for logging in, but when trying to log in the carrier rejects the login" May 02 11:28:11 weird new extensions to the OverTheAir protocol might actually even cause lockups or freezes in the modem's firmware, so yes, it may also be a firmware issue, alas not fixable on N900 side, only by using a different SIM for a different provider, or the 'rogue' provider fixing their protocol extensions May 02 11:29:49 generally I strongly suggest installing syslogd and having a look into the logs May 02 11:32:23 to rule out the "incompatible carrier" issues you could use any other Nokia phone using BB5 modem of roughly same age. They are a zillion May 02 11:33:48 https://en.wikipedia.org/wiki/List_of_Nokia_products ^F "BB5" May 02 11:44:23 ideally you check for similar modem firmware, but I don't know how to accomplish that May 02 11:45:05 going for release date in *#0000# might be the as close as it gets here May 02 11:57:19 DocScrutinizer05: lol @ springs and glue .. May 02 11:58:32 but ehm. the cell connection just keeps goin on/off/on/off May 02 11:58:36 was so at home and now at work May 02 11:58:49 can't really discern any when/how/what would be the cause May 02 11:59:05 change provider? May 02 11:59:12 ie. test with different network? May 02 11:59:31 borrow from someone or just buy cheapest prepaid May 02 12:11:24 so, that would be if it's the towers dropping certain bands? May 02 12:11:36 there are only 2 kinds of towers in .nl May 02 12:11:43 those by kpn and those by t-mobile May 02 12:11:59 any prepaid will also use the kpn network (where I'm on). so it wouldn't matter May 02 13:10:25 modem on its way out. mine did the same thing for a few weeks, until eventually it completely stopped working (second N900. thankfully my daily is still fine) May 02 14:35:27 dreamer: most likely your modem can receive but fails to transmit May 02 14:36:15 since logging in to a network involves transmitting, the login will eventually fail May 02 14:36:50 the modem "resets" and retries May 02 14:41:11 possible reasons are (in no particular order): chip balls coming off the PCB; IMEI blacklisted at provider; antenna contacts dirty; SIM contacts dirty; battery weak/worn; huge distance to base station; firmware issues, possibly caused by new protocol details on carrier May 02 14:45:24 DocScrutinizer05: well. it's mostly just off (simcard with red stripe). then sometimes it can connect again May 02 14:45:54 no, it only thinks it can connect? May 02 14:48:16 anyway, the problem is known, probably usually related to chip solder joints snapping open, and there's no surefire fix for it May 02 14:49:14 the initial cause is too high force pushing down keys on hw keyboard May 02 14:53:01 could also happen from pushing/squeezing the complete device while in pocket for example May 02 15:08:39 DocScrutinizer05: no it actually connects. I can use the cell functions when it is connected May 02 15:08:53 hmm ok May 02 15:09:08 so, my guess is reflowing modem May 02 15:09:20 still that doesn't help tracking down the issue much May 02 15:09:21 but .. no idea when I'll have access to a reflow gun May 02 15:10:07 check your syslog! May 02 15:10:21 though odds are it's the usual issue May 02 15:10:38 McSAAB? May 02 15:10:50 hm? May 02 15:11:06 I see bunch of messages by McSAAB in dmesg May 02 15:11:26 this is the interface to modem May 02 15:11:41 'CMT reset line change detected' and then bunch of stuff May 02 15:13:46 yes, that means the modem received a reset May 02 15:14:02 or rather, sent one May 02 15:14:17 iow it rebooted May 02 15:14:37 so that indicates the modem doesn't connect to the pcb properly? May 02 15:14:42 or can it reboot to other reasons? May 02 15:14:57 of course it can reboot for a lot of reasons May 02 15:15:12 none of them expected though May 02 15:17:12 for reference: http://paste.ubuntu.com/24499019 May 02 15:19:15 IroN900:~# wc -l /var/log/syslog May 02 15:19:16 137722 /var/log/syslog May 02 15:20:11 * dreamer really needs to fix ssh into the n900 May 02 15:20:18 for this kind of stuff hehe May 02 15:20:48 usually after a restart it can connect to the cell btw. can see my provider name and 3G reception etc. May 02 16:43:25 Hello. I investigated a bit about gpio keys in Devuan on the n900. After loading the kernel module gpio-keys, a new node appears on /dev/input/event3. I tried with evtest event3 and got the following: May 02 16:43:29 root@devuan:/dev/input# evtest event3 May 02 16:43:29 Input driver version is 1.0.1 May 02 16:43:29 Input device ID: bus 0x19 vendor 0x1 product 0x1 version 0x100 May 02 16:43:29 Input device name: "gpio_keys" May 02 16:43:30 Supported events: May 02 16:43:31 Event type 0 (EV_SYN) May 02 16:43:32 Event type 1 (EV_KEY) May 02 16:43:33 Event code 152 (KEY_SCREENLOCK) May 02 16:43:34 Event code 212 (KEY_CAMERA) May 02 16:43:35 Event code 528 (?) May 02 16:43:36 Event type 5 (EV_SW) May 02 16:43:37 Event code 9 (SW_CAMERA_LENS_COVER) May 02 16:43:38 Event code 10 (SW_KEYPAD_SLIDE) May 02 16:43:39 Event code 11 (SW_FRONT_PROXIMITY) May 02 16:43:40 Properties: May 02 16:43:41 Testing ... (interrupt to exit) May 02 16:44:26 slide it baby, slid May 02 16:44:32 The side slide switch reacts as KEY_SCREENLOCK. Let's forget for now the (?) event (quite strange). May 02 16:45:04 and proximity is the photoresistor near the screen May 02 16:45:25 Ok. May 02 16:46:19 Now these gpio keys produce strange symbols on console. I'd expect something listening to them and reporting events to the console. But how to find what it is and make it behave properly? May 02 16:47:45 you get what you have there May 02 16:48:06 ie. check input.h for key aliases and friends May 02 16:48:10 KEY_CAMERA KEY_SCREENLOCK etc May 02 16:49:00 Input.h in the kernel module? May 02 16:49:36 in kernel tree May 02 16:49:49 Ok. I'll look for it. May 02 16:49:54 SW_* ones might be in some n900 specific one though May 02 16:50:17 Ok. May 02 16:50:51 those are the keys generated, think of it as a keyboard with multimedia keys May 02 17:14:04 Enrico_Menotti: didn't I point you to a website a few days ago? http://compsoc.man.ac.uk/~shep/pedals.html May 02 17:14:09 also see May 02 17:14:20 ~vi-keymap May 02 17:14:27 dang what been it May 02 17:14:50 ~listvalues vi*awesome May 02 17:14:52 Factoid search of 'vi*awesome' by value (3): vi-kbd ;; ironman ;; #maemo awesome-keymap. May 02 17:15:01 ~awesome-keymap May 02 17:15:01 i heard awesome-keymap is http://wiki.maemo.org/Remapping_keyboard/user_vis_awesome_kbdmapping May 02 17:15:34 DocScrutinizer05 To be honest I don't remember that first website. Yes, I know the awesome kbdmapping - I will look at that again. Ty. May 02 17:16:16 Meanwhile I found where the gpio key definitions are (at least, some trace of them): https://lkml.org/lkml/2013/10/22/509. May 02 17:17:32 IIRC you first need to determine the scancodes, then assign keysyms to them, and finally make your apps react on the keysyms May 02 17:18:05 wait May 02 17:18:39 IIRC you first need to determine the scancodes, then assign keycodes to scancodes, then assign keysyms to keycodes, and finally make your apps react on the keysyms ??? May 02 17:19:32 Now I see the first link refers to udev, which I don't have. May 02 17:19:47 input in linux is a MESS May 02 17:19:53 :) May 02 17:20:12 😎 May 02 17:20:19 :D May 02 17:20:24 udev only executes other commandline tools basically May 02 17:20:33 Ah. May 02 17:21:25 afaik the only thing udev does on its own is creating generic 'files' with permissions and maybe also mount or symlink sth there May 02 17:21:32 😎 May 02 17:22:45 also it can change owner/permission of files May 02 17:23:18 but for stuff like keymapping it for sure resorts to other commandline tools, I'd think May 02 17:25:41 oh fsck! actually udev handles scancode mapping internally May 02 17:27:00 KEYBOARD_KEY_70004=f13 ((/etc/udev/hwdb.d/90-custom-keyboard.hwdb)) but AIUI this is already accomplished in your gpio driver which doesn't provide scancodes but actually keycodes May 02 17:29:49 at least your "evtest event3" results suggest that May 02 17:32:46 so for lockswitch I'd assume you get a "Event type 1 (EV_KEY) - Event code 152 (KEY_SCREENLOCK)" event on dev/input3 May 02 17:33:24 in short, gpio input layer says 'there is scancode 152' and kernel says 'ok got KEY_SCREENLOCK' May 02 17:33:55 so your apps should react on KEY_* and SW_* keys May 02 17:35:19 xev would see something similar to http://paste.ubuntu.com/24499762 May 02 17:37:52 compare to http://paste.ubuntu.com/24499769 which is my spacebar key and mapped to char " " May 02 17:40:26 I guess "KeyPress event" is equivalent to "Event type 1 (EV_KEY)" May 02 17:40:57 I admit I am a bit confused. DocScrutinizer05 , how did you get exactly those pastes? And if I understand, they are related to the X environment, right? Now I'm on console - where is the mapping from KEY_SCREENLOCK to any action? May 02 17:41:21 there is none for console May 02 17:42:10 Enrico_Menotti: you can remap those keys with mainline kernel May 02 17:42:11 except in the shell and mintty or whatever you're running May 02 17:43:31 I'll add what I get from evtest event3 if I move the slide switch: May 02 17:43:32 Event: time 1493745937.054557, type 1 (EV_KEY), code 152 (KEY_SCREENLOCK), value 1 May 02 17:43:33 Event: time 1493745937.054557, -------------- EV_SYN ------------ May 02 17:43:33 Event: time 1493745937.228474, type 1 (EV_KEY), code 152 (KEY_SCREENLOCK), value 0 May 02 17:43:33 Event: time 1493745937.228474, -------------- EV_SYN ------------ May 02 17:44:24 when I'm on console, pressing that keycode 172 (keysym 0x1008ff14, XF86AudioPlay) on my PC's keyboard doesn't do shit May 02 17:44:56 172? May 02 17:45:25 yes, that's the jeycode for that multimedia "Play" button May 02 17:45:35 Ok. May 02 17:46:10 lockswitch on N900 with GPIOswitch obviously delivers code 152 (KEY_SCREENLOCK) May 02 17:47:07 plus a EV_SYN event whatever that is May 02 17:47:31 Ok, so this code appears somewhat at /dev/input/event3. Now how is that converted to a signal to the console? May 02 17:47:45 signal? May 02 17:48:29 what kind of "signal" do you hope for? May 02 17:49:02 I used that ambigous term on purpose, since I don't know actually how to define some event as perceived by the console. May 02 17:49:04 tbh the console shouldn't even read from /dev/input3 at all May 02 17:49:18 Yes, in fact. This is what is strange to me. May 02 17:49:55 and when it does, it shouldn't receive a char mapped to those events May 02 17:50:02 so nothing for the console to do May 02 17:50:05 Yes. May 02 17:50:33 od -foobarparam /dev/input3 May 02 17:50:41 With the same kernel, and booting Debian, the slide switch works properly (on console): it wakes up the screen. May 02 17:51:04 that's not done "on console" May 02 17:51:13 Yes, I understand. May 02 17:51:43 But console doesn't output strange symbols. (On Debian.) (At least, I did not recognise any.) May 02 17:51:50 Let me check. May 02 17:51:50 some other process is listening to /dev/onput3 and acts accordingly when seeing "type 1 (EV_KEY), code 152 (KEY_SCREENLOCK), value 1" May 02 17:52:56 "act accordingly" means: blank/unblamk the screen May 02 17:53:31 I'm booting Debian. I want to understand what happens there. May 02 17:53:35 unrelated to whether that's a console or a hildon-chess or whatever showing on display May 02 17:53:58 lsof|grep /dev/input3 May 02 17:54:19 if there's even such thing like /dev/input3 May 02 17:54:30 Ok. A moment. May 02 17:54:47 could even be kernel itself handling that May 02 17:55:21 to start with, check if there's /dev/input3 and lsmod|grep gpio May 02 17:56:16 maemo handles that stuff completely different, via /sys/* May 02 17:56:29 and kevents iirc May 02 17:56:48 check kerecv May 02 17:59:35 hmm nope, maemo's ke-recv handles a lot but not lockswitch http://paste.ubuntu.com/24499859 May 02 18:00:15 Ok, I was completely wrong. Sorry. Also with Debian I get the same strange symbol on console. :/ May 02 18:06:03 yes, I also get ^@ on console for keys that are not known to the console May 02 18:06:18 on my PC May 02 18:07:15 however there are other keys I don't get anything at all May 02 18:10:36 only in evtest though. without evtest running I don't get anything at all for most "special" keys", incl the vol- key that delivers ^[[25~ and the Play key that delivers ^@ in evtest May 02 18:12:56 dreamer: just a thought, but have u tried to change phone modes from settings? I mean dual-3g-gsm. May 02 18:12:57 anyway who cares about special keys in console of N900? ;-P May 02 18:13:54 after all console isn't what you want to run on N900 in the end May 02 18:14:26 DocScrutinizer05 Yes, ok. I was thinking there was something wrong, that's why I started investigating these gpio keys. Now I know the module which handles them, and also where they show up (/dev/input/...). May 02 18:15:15 I'm just curious to understand where these events get mapped to keycodes. May 02 18:50:09 Vajb: either gsm or 3g it still loses connection May 02 18:52:55 I think I'll have to go try and destroy a pen to put the spring in there May 02 18:53:01 or 'bits of spring' May 02 18:53:08 and see if that keeps the modem connected .. May 02 19:52:24 meh May 02 19:53:10 Wizzup: you don't happen to want to trade an n900 do you? ;) (if you don't need stable modem) May 02 21:16:45 dreamer: i'll give you one of mine if you wish May 02 21:16:53 (trade) May 02 21:17:07 parazyd: if the sim and usb work :P May 02 21:17:17 does your usb work? May 02 21:17:20 yes May 02 21:17:31 then i'd trade May 02 21:17:43 will query May 02 21:17:59 cool May 02 21:36:43 sicelo: yay **** ENDING LOGGING AT Wed May 03 03:00:00 2017