Wizzup | uvos: ok, I think I fixed the sphone patch and did some typo fixes | 01:53 |
---|---|---|
Wizzup | will cleanly test soon, maybe not tomorrow, might be travelling | 01:53 |
SuperMarioSF | a little problem... | 07:39 |
SuperMarioSF | seems my d4 all upgraded. | 07:39 |
SuperMarioSF | one of them upgraded to Anrdoid 4.x | 07:39 |
SuperMarioSF | one of them installed LineageOS | 07:40 |
SuperMarioSF | Don't know if this will affect something. | 07:40 |
Wizzup | SuperMarioSF: I think our wiki says if you ought to downgrade (I think so) | 08:30 |
SuperMarioSF66 | oh hi | 09:20 |
SuperMarioSF66 | i' logged in with another pc | 09:20 |
SuperMarioSF66 | I got my first D4 | 09:20 |
SuperMarioSF66 | but it was locked at the moment. | 09:20 |
SuperMarioSF66 | and from postmarketOS guys said this device cannot be OEM unlocked. | 09:21 |
SuperMarioSF66 | I checked and this device already have rooted. | 09:21 |
SuperMarioSF66 | and kernel version at 3.0.8 | 09:22 |
SuperMarioSF66 | seems this phone still have its stock recovery | 09:25 |
uvos__ | all droid 4's (and indeed all verizon devices) have bootloaders that are permanently locked with no way to unlock them | 09:37 |
uvos__ | Wizzup: no downgrading is not nessecary | 09:37 |
SuperMarioSF66 | so what should I do next? | 09:40 |
uvos__ | follow the steps in the wiki as per normal | 09:41 |
SuperMarioSF66 | the installer guide said about flashing with fastboot, is the locked bootloader a problem? | 09:41 |
uvos__ | no | 09:41 |
SuperMarioSF66 | so just flash and everything is ok? | 09:41 |
uvos__ | yes | 09:41 |
SuperMarioSF66 | ... before that, is there a way to do some backup? | 09:42 |
uvos__ | the only prerequisit is that the kernel 3.0.8 is installed on /boot | 09:42 |
uvos__ | but if you have android 4.0 or 4.1 or los this is the case | 09:42 |
uvos__ | SuperMarioSF66: no but there is no reason to do so | 09:42 |
SuperMarioSF66 | I see the bpsw partition will be overwritten, so will this affect baseband? | 09:42 |
uvos__ | there is no way to brick anything and we have all images availble | 09:42 |
uvos__ | no | 09:42 |
uvos__ | bpsw is a misnomer | 09:42 |
uvos__ | its not used for either modem | 09:43 |
SuperMarioSF66 | hmmmm the bpsw partition is a useless one on this phone? | 09:43 |
uvos__ | yes its unused | 09:43 |
SuperMarioSF66 | OK | 09:43 |
uvos__ | except by us ofc ;) | 09:43 |
SuperMarioSF66 | I'm preparing a SD card, wait a moment... | 09:43 |
SuperMarioSF66 | before that I guess I should test my SIM card on stock firmware to make sure everything is still worrking fine. | 09:46 |
uvos__ | you can boot the stock fw fine after too | 09:46 |
uvos__ | but sure | 09:46 |
SuperMarioSF66 | and I forgot to bring my spare SIM cards... | 09:49 |
SuperMarioSF66 | oops | 09:49 |
SuperMarioSF66 | maybe I should test this later. | 09:50 |
uvos__ | installing leste dosent mess with stock android at all | 09:51 |
uvos__ | so it dosent matter | 09:51 |
SuperMarioSF66 | so I just download the latest SD card image and DD it on a card? | 09:52 |
uvos__ | and install the 2 stage bootloader as descirbed on the wiki | 09:52 |
SuperMarioSF66 | OK | 09:53 |
uvos__ | then theres a boot menu with entries you can edit via a text file, by default theres 2 entries, leste from sd and stock android from /system | 09:54 |
uvos__ | but you can have as many entries as you want | 09:54 |
uvos__ | another for los for instance see https://github.com/tmlind/droid4-kexecboot/blob/master/CONFIGURATION | 09:55 |
SuperMarioSF66 | I guess I won't need more than that. Maybe I will consider just disable that stock android entry. | 09:56 |
uvos__ | you can install leste to the partitions of stock android | 09:56 |
uvos__ | but its a manual procedure atm | 09:56 |
SuperMarioSF66 | at least that is an option. | 09:57 |
SuperMarioSF66 | but I would like to have a unified experience and a lots of storage.So currently I will use SD card. easy for me to do full backups. | 09:58 |
SuperMarioSF66 | I just ordered some MLC based SD card, to serve this usage. | 09:58 |
SuperMarioSF66 | So after I installed on my current SD card, there will be a data moving action taken after I received my extra durance MLC cards. | 09:59 |
SuperMarioSF66 | and they are berrer performance one too, A2 standard supported, though I doubt that will work on such a old device. | 10:00 |
uvos__ | the sdcard bus runs at 50Mhz Sdr max | 10:01 |
uvos__ | with 4 bit width | 10:01 |
uvos__ | or 8bit for the internal storage | 10:01 |
uvos__ | you can patch it to run at 100Mhz Sdr | 10:01 |
uvos__ | but this is an overclock (that i do run) | 10:01 |
SuperMarioSF66 | I won't take that risk... | 10:01 |
uvos__ | so yeah it can do A2 speeds in this case | 10:01 |
uvos__ | its not a risk to the device | 10:01 |
uvos__ | its supports this speed | 10:01 |
uvos__ | the problem is it uses 5v signaling at 100mhz | 10:02 |
uvos__ | and sdcard spec is 5v at 50mhz or 1.8v at 100 | 10:02 |
uvos__ | so you might break the sdcard, but the device is fine | 10:02 |
SuperMarioSF66 | so it requied a newer card to respect 5V signaling? | 10:02 |
uvos__ | no | 10:02 |
uvos__ | no sdcards support 5v signaling at 100mhz | 10:02 |
uvos__ | its not supported by the sdcard spec | 10:02 |
uvos__ | it is used by emmcs tho | 10:03 |
uvos__ | this is why the d4's chip can do this mode | 10:03 |
uvos__ | in reality every sdcard i ever tried had no problem doing 5v @100 | 10:03 |
SuperMarioSF66 | so if I can get some eMMC wired there with a special PCB, that will work? | 10:03 |
uvos__ | sure | 10:03 |
uvos__ | but it would be kinda strange for a sdcard to support 5v @50 and 1.8v @100 but break at 5v @100 | 10:04 |
uvos__ | so in reallity it allways works anyhow | 10:04 |
SuperMarioSF66 | no worries | 10:04 |
uvos__ | (5v @50 is requried by spec) | 10:04 |
SuperMarioSF66 | I make a backup and I have many spare cards available. even fried one won't a proble,. | 10:04 |
SuperMarioSF66 | *wont be a problem | 10:04 |
uvos__ | anyhow for now yes stock kernel dosent do a2 speed | 10:05 |
SuperMarioSF66 | I guess that card won't work because they were formatted with ext4 to begin with... | 10:05 |
uvos__ | your dding a filesystem | 10:06 |
uvos__ | what it was formated as is immaterial | 10:06 |
SuperMarioSF66 | I mean... stock android won't just work on ext4 | 10:06 |
uvos__ | oh no | 10:07 |
SuperMarioSF66 | for SD card at least | 10:07 |
uvos__ | it cant mount modern ext4 at all | 10:07 |
uvos__ | it needs disabled csums | 10:07 |
SuperMarioSF66 | blazing fast 30+MB/s dding | 10:14 |
SuperMarioSF66 | oh I was wrong, it just a decompression. actual DD speed is 80MB/s+ | 10:14 |
SuperMarioSF | oh I'm back. seems RevolutionIRC client had some bugs and I can't connect to libera.chat properly. | 10:40 |
Wizzup | wb | 10:42 |
SuperMarioSF | Got my sim card | 11:03 |
SuperMarioSF | works just fine, however signal isn't very good on UMTS. | 11:03 |
SuperMarioSF | seems that LTE only works with CDMA side. | 11:04 |
Wizzup | yeah | 11:12 |
uvos__ | LTE is unrelated to CDMA but supports one band | 11:15 |
uvos__ | 13 iirc | 11:15 |
uvos__ | and on leste it dosent work at al | 11:16 |
uvos__ | since we dont power the LTE modem and dont have a driver for it | 11:16 |
SuperMarioSF | OK I'm back, downloading files for flashing. | 11:20 |
SuperMarioSF | Done, I'm in. | 11:53 |
SuperMarioSF | Looks like there is no UI for 3G right now. | 11:53 |
SuperMarioSF | is there a way to enable that in commandline? | 11:54 |
SuperMarioSF | or should I install modemmanager or something to manage? | 11:54 |
uvos__ | theres a ui that might sorta work in -devel | 11:54 |
uvos__ | but it dosent at all for me (and installing it breaks all conectivity for me) | 11:54 |
uvos__ | for others it dose work | 11:54 |
uvos__ | theres a metapackage called hildon-connectivity-mobile or some sutch | 11:55 |
SuperMarioSF | so how can I install that -devel? | 11:55 |
uvos__ | otherwise you have to use qmi-cli | 11:55 |
SuperMarioSF | a package or whole image? | 11:56 |
uvos__ | please read the documentation, its in the wiki under development | 11:57 |
SuperMarioSF | OK | 11:57 |
Wizzup | let me know if doc is hard to find, can try to help, but I'm travelling right now, so might be in and out | 12:04 |
SuperMarioSF | I found the doc and enabled devel repos | 12:23 |
SuperMarioSF | so just a dist-upgrade next? | 12:23 |
SuperMarioSF | btw I can provide some i18n work for Chinese. | 12:25 |
SuperMarioSF | and already working on it on weblate. | 12:25 |
Wizzup | there are also meta pkgs you need, I can't look it up right now | 12:45 |
SuperMarioSF | any hint for that meta package? | 12:45 |
Wizzup | SuperMarioSF: let me look now | 12:54 |
Wizzup | SuperMarioSF: hildon-connectivity-mobile at least | 12:54 |
Wizzup | you will still need to do -some- manual steps after boot, depending on whether you sim has a pin or not | 12:54 |
SuperMarioSF | I removed PIN for my SIM card before using in d4. | 12:55 |
Wizzup | ok | 12:55 |
Wizzup | in that case, install that package and reboot, see how far it gets you :) | 12:55 |
SuperMarioSF | so should I finish the dist-upgrade first (downloaded) | 12:55 |
Wizzup | yeah, sure, do that too | 12:56 |
SuperMarioSF | woah that's a lot packages... | 12:58 |
Wizzup | lots of translations too I bet | 12:59 |
SuperMarioSF | actually, Simplified Chinese doesn't have much translation unfinished... | 13:02 |
SuperMarioSF | the majority of them is just city names. | 13:02 |
Wizzup | :) | 13:05 |
Wizzup | that is only for the core of course | 13:05 |
Wizzup | user contributed packages and such might need more | 13:05 |
SuperMarioSF | OK | 13:05 |
SuperMarioSF | I have purchase PoEdit Pro so that shouldn't be very hard. | 13:06 |
SuperMarioSF | rebooted. | 13:07 |
Wizzup | do you see a sim not present icon, or does the status menu look the same? | 13:08 |
Wizzup | the modem ought to get online'd once you go to internet connections through the tsatus bar | 13:08 |
Wizzup | (where you also search for wifi) | 13:08 |
Wizzup | it might take a bit for the d4 to first make a connection I suppose | 13:08 |
SuperMarioSF | it is booting... | 13:09 |
Wizzup | (btw since ofono mostly talks over dbus, I usually use dbus to debug it's states, but iirc there are also some scripts (ofono-scripts pkg?) | 13:10 |
SuperMarioSF | btw a suggestion, please make some LED on during a power cycle, or that will be diffcult to know my device is still powering down or not. | 13:10 |
uvos__ | this exits on power down | 13:10 |
uvos__ | and works | 13:10 |
SuperMarioSF | Oh, a SIM card icon shown, however looks like it doesn't find my SIM card. | 13:11 |
uvos__ | but only if you shutdown with something other than the terminal | 13:11 |
Wizzup | SuperMarioSF: ok, give it a little bit of time (1-2 mins) | 13:11 |
Wizzup | the modem itself doesn't report on sim status until then | 13:11 |
Wizzup | it's possible you will see the sim missing icon disappear soon | 13:11 |
SuperMarioSF | and desktop wallpaper is gone | 13:11 |
uvos__ | uh | 13:11 |
Wizzup | huh? | 13:12 |
SuperMarioSF | attempt to set a new wallpaper, got error: No space left | 13:12 |
uvos__ | oh | 13:12 |
Wizzup | oh... | 13:12 |
uvos__ | you need to expand the partiton before upgrade | 13:12 |
Wizzup | /etc/expandcard.sh will resiez the sd card | 13:12 |
SuperMarioSF | and sim icon gone. | 13:12 |
SuperMarioSF | oh | 13:12 |
uvos__ | you broke your install now :/ | 13:12 |
SuperMarioSF | I forgot that | 13:12 |
Wizzup | I hope the apt-get dist-upgrade worked fully | 13:12 |
SuperMarioSF | time to reinstall | 13:12 |
Wizzup | uvos__: not sure if it's broken? | 13:12 |
Wizzup | apt usually errors on disk full | 13:12 |
uvos__ | yeah when i did that last weak it broke it | 13:13 |
SuperMarioSF | dist-upgrade and that package successfully completed. | 13:13 |
Wizzup | and then you can resume it manually | 13:13 |
uvos__ | but maybe he is lucky | 13:13 |
uvos__ | [13:12] <Wizzup> apt usually errors on disk full | 13:13 |
uvos__ | for some reason that is farily unreliable | 13:13 |
SuperMarioSF | resizing | 13:14 |
uvos__ | i have this problem on other machines too | 13:14 |
SuperMarioSF | OK, done, reboot? | 13:14 |
uvos__ | yeah hopefully it will be ok | 13:14 |
SuperMarioSF | no worries, I can just reinstall from start. | 13:14 |
uvos__ | we could maybe make the image a couple 100mb larger | 13:15 |
uvos__ | so that devel fits | 13:15 |
Wizzup | yeah. | 13:16 |
SuperMarioSF | the interesting thing is there is no error reported during upgrade and install. | 13:16 |
Wizzup | we try to make it work for very small cards | 13:16 |
Wizzup | SuperMarioSF: yeah probably you were just on the edge | 13:16 |
SuperMarioSF | the space is barely enough | 13:16 |
SuperMarioSF | Ok | 13:16 |
SuperMarioSF | everything is back | 13:16 |
SuperMarioSF | SIM icon show no card. | 13:16 |
Wizzup | give it a minute again :) | 13:17 |
SuperMarioSF | and icon is gone... | 13:18 |
Wizzup | we probe and power the modem relatively late in the boot process, and the modem needs time to start up itself | 13:18 |
Wizzup | ok, that means your sim is found, likely | 13:18 |
SuperMarioSF | nothing left except battery icon | 13:18 |
SuperMarioSF | so what's next? | 13:18 |
uvos__ | wait | 13:18 |
uvos__ | should come up with a signal | 13:19 |
Wizzup | uvos__: no, it will only online the modem in the network search dialog | 13:19 |
SuperMarioSF | is there are debug logs to see here? | 13:19 |
uvos__ | right true | 13:19 |
Wizzup | you can install ofono-scripts and run stuff like /usr/share/ofono/scripts/list-modems | 13:19 |
Wizzup | or use dbus if you know it | 13:19 |
SuperMarioSF | oh | 13:20 |
SuperMarioSF | singal shown | 13:20 |
Wizzup | but if you just go to the place where in maemo you usually connect to a network, just opening the dialog will make it online the modem | 13:20 |
SuperMarioSF | uhh... there is a problem | 13:20 |
uvos__ | right heres where mine never works beyond this point | 13:20 |
SuperMarioSF | connection selection was grey out | 13:20 |
SuperMarioSF | everything cannot be selected. | 13:20 |
uvos__ | right same here :) | 13:20 |
uvos__ | wait a bit | 13:20 |
uvos__ | its wating for ofono to provision the net | 13:21 |
uvos__ | but it never happens here | 13:21 |
SuperMarioSF | OK | 13:21 |
Wizzup | and for the modem to become online | 13:21 |
SuperMarioSF | I can see signal but it is weak (1 bar) | 13:21 |
SuperMarioSF | started from 3 bars | 13:21 |
Wizzup | that's good, 2g or 3g? | 13:21 |
Wizzup | that means the modem is online | 13:21 |
uvos__ | the signal bars are very conservative on d4 | 13:21 |
SuperMarioSF | no technology were displayed. only bars. | 13:21 |
Wizzup | so -probably- you can even make a phone call now with sphone | 13:22 |
Wizzup | SuperMarioSF: huh, that's ie | 13:22 |
Wizzup | that's weird | 13:22 |
SuperMarioSF | sphone stucked at dialing | 13:23 |
SuperMarioSF | and it closed itself | 13:23 |
Wizzup | stuck how? | 13:23 |
SuperMarioSF | stuck, not means freeze. | 13:23 |
Wizzup | so I am wondering if this is actually using some tech that we have not mapped :) | 13:23 |
Wizzup | if you see bars but no icon | 13:23 |
uvos__ | check what ofono thinks tech is | 13:23 |
uvos__ | in list-modems | 13:23 |
SuperMarioSF | and there is a problem | 13:24 |
SuperMarioSF | WiFi cannot be connected... | 13:24 |
Wizzup | slow down :D | 13:24 |
uvos__ | right if mobile network dosent work the applet breaks wifi | 13:24 |
Wizzup | uvos__: not for me :/ | 13:24 |
uvos__ | because it hangs icd forever | 13:24 |
Wizzup | icd doesn't hang, the dialog waits for something | 13:25 |
SuperMarioSF | if there is a way to let me get a usb0 interface via USB? | 13:25 |
uvos__ | right the dialog hangs | 13:25 |
Wizzup | SuperMarioSF: just plug in usb and it'll show up | 13:25 |
Wizzup | https://leste.maemo.org/Status/USB_Peripheral | 13:25 |
SuperMarioSF | seems my Windows machine need a driver. | 13:25 |
Wizzup | I don't know how it works on windows | 13:25 |
Wizzup | sorry | 13:25 |
SuperMarioSF | I'm backing to home, and I will attempt connect that to my Mikrotik router. | 13:26 |
Wizzup | do you have a linux device? | 13:26 |
SuperMarioSF | I do know how to make it work after that. | 13:26 |
uvos__ | probubly no one here runs windows on pcs :P | 13:26 |
Wizzup | so I'm wondering what the technology here will be | 13:26 |
uvos__ | if you want wifi to work again, just unstiall the mobile icd applet | 13:26 |
Wizzup | uvos__: list-modems doesn't show it fwiw | 13:26 |
uvos__ | (for now) | 13:26 |
SuperMarioSF | I have linux based device at home, but I don't want to setup the connection manually, so I will use a router instead. ( I know how to do that) | 13:26 |
Wizzup | ok | 13:26 |
uvos__ | Wizzup: you can see it in ofono-scipts for sure | 13:27 |
uvos__ | Wizzup: i gues maybe not that one | 13:27 |
SuperMarioSF | oh wait | 13:27 |
Wizzup | uvos__: yeah I was searching | 13:27 |
SuperMarioSF | wifi just came back on itself | 13:27 |
uvos__ | right autoconnect works | 13:27 |
uvos__ | same here | 13:27 |
Wizzup | do we have a ticket for this behaviour? | 13:28 |
uvos__ | pretty sure i made one | 13:28 |
SuperMarioSF | btw there are some suggested package during install hildon-connectivity-mobile, so maybe something missing can be in there? | 13:29 |
Wizzup | not likely | 13:29 |
uvos__ | ofono has some suggestions | 13:29 |
uvos__ | there not nesseary | 13:29 |
Wizzup | btw I would install this .deb https://packages.debian.org/stretch/armhf/mdbus2/download for debugging purposes later | 13:29 |
uvos__ | ists probubly those | 13:29 |
Wizzup | I find it the easiest way to browse around dbus interfaces on cli | 13:30 |
SuperMarioSF | so how to list modems? no auto-completion for list-modem, for ofono, only ofonod shown. | 13:30 |
Wizzup | sudo apt-get install ofono-scripts | 13:30 |
uvos__ | its in /usr/share/ofono/scritps | 13:30 |
uvos__ | and not in PATH | 13:30 |
SuperMarioSF | OK | 13:30 |
uvos__ | and yes you need that package | 13:30 |
SuperMarioSF | Status = roaming | 13:31 |
SuperMarioSF | this doesn't seems right. | 13:31 |
Wizzup | does it list a network name anywhere? | 13:32 |
Wizzup | not sure if this will eventually be relevant (note: not the same device, careful, etc) | 13:33 |
Wizzup | https://leste.maemo.org/Motorola_Droid_Bionic#Unlocking_the_modem_for_usage_in_Europe | 13:33 |
uvos__ | both of those are set on d4 | 13:33 |
Wizzup | ok | 13:33 |
uvos__ | thats where those came from (diffing d4) | 13:33 |
Wizzup | xt910 but yeah | 13:33 |
SuperMarioSF | https://paste.mozilla.org/pMzmi5mj/raw | 13:34 |
SuperMarioSF | here is the output. sensive data masked with X | 13:34 |
SuperMarioSF | this link will be expired in 1 hour. | 13:35 |
Wizzup | ok yeah so maybe your missing icon is related to the disk full | 13:36 |
Wizzup | maybe the icon cache is broken or so | 13:36 |
Wizzup | since: | 13:37 |
Wizzup | [ org.ofono.RadioSettings ] | 13:37 |
Wizzup | TechnologyPreference = any | 13:37 |
Wizzup | AvailableTechnologies = gsm umts | 13:37 |
SuperMarioSF | btw, this phone start with a roaming state when I got it, and at that time there is even no SIM inserted. | 13:37 |
SuperMarioSF | OK, so I need to rebuild the cache? | 13:37 |
Wizzup | either that or reinstall because of the disk full issue to be sure | 13:37 |
Wizzup | wrt roaming I have to check what my phone says | 13:37 |
SuperMarioSF | so what directory I should point update-icon-caches to? | 13:38 |
SuperMarioSF | -> /usr/share/icons/* ? | 13:39 |
Wizzup | sorry, don't know on top of my head | 13:40 |
Wizzup | but if this is broken, other stuff could be too, like uvos says | 13:40 |
SuperMarioSF | so my best bet is just reinstall everything. | 13:40 |
SuperMarioSF | will do that. | 13:40 |
Wizzup | I think so. sorry. | 13:40 |
SuperMarioSF | no big deal, I have a bazing fast SD card and a proper UHS-II supported card reader. | 13:41 |
Wizzup | I need to get to the airport, back a bit later. | 13:41 |
SuperMarioSF | about 3D printing d4 back cover, that cover is so thin... I understand why you said that isn't a good idea... | 13:46 |
SuperMarioSF | reflashed, booting. | 13:47 |
SuperMarioSF | ok i'm back | 14:11 |
SuperMarioSF | re-upgrading | 14:11 |
SuperMarioSF | and this time, no signal. | 14:33 |
SuperMarioSF | SIM card detected, no signal bar displaied, try to call via sphone, sphone said cannot connect to ofono | 14:33 |
SuperMarioSF | rebooting again to try again. | 14:33 |
Wizzup | did you install the meta pkg? | 14:33 |
SuperMarioSF | yes, I had installed it. | 14:34 |
SuperMarioSF | otherwise there should no phone app. | 14:34 |
Wizzup | right | 14:34 |
uvos__ | no wrong | 14:36 |
uvos__ | i think sphone is in the main metapackage | 14:36 |
uvos__ | or? | 14:36 |
Wizzup | not sure, trying to install the mobile connectivity meta pkg will help us figure this out I think | 14:37 |
uvos__ | cannot connect to ofono == ofono not running | 14:37 |
uvos__ | it is in hildon-connectivity-mobile Wizzup was right | 14:38 |
SuperMarioSF | this time it doesn't even show about roaming | 14:42 |
SuperMarioSF | list-modem shown some info | 14:42 |
SuperMarioSF | my phone number and subscriber identity things are present. | 14:43 |
SuperMarioSF | online=0Â powered=1Â lockdown=0Â emergency=0 | 14:43 |
Wizzup | yes, so did you go to the network search again to online it? | 14:44 |
Wizzup | (this is a kludge until we figure out what should online it) | 14:44 |
SuperMarioSF | network search stucked at grey out as usual. | 14:45 |
SuperMarioSF | oh | 14:45 |
SuperMarioSF | signal shown | 14:45 |
SuperMarioSF | and... same again | 14:45 |
SuperMarioSF | no technology shown | 14:45 |
SuperMarioSF | only a single bar | 14:45 |
uvos__ | maybe try forceing GSM via scripts | 14:45 |
SuperMarioSF | how to? I did't find an obvious one for that | 14:48 |
SuperMarioSF | btw show cell tower info works | 14:48 |
uvos__ | set-tech-preference | 14:50 |
SuperMarioSF | oh | 14:51 |
SuperMarioSF | I got something by forcing it to UMTS | 14:51 |
SuperMarioSF | shown 3G now | 14:51 |
SuperMarioSF | and GSM usually don't work in my area anyways. | 14:51 |
SuperMarioSF | OK, phone calling works confirmed. | 14:52 |
SuperMarioSF | how can I setup APN | 14:53 |
SuperMarioSF | for modem? | 14:53 |
SuperMarioSF | receiving call works. | 14:55 |
SuperMarioSF | and voice is okay | 14:55 |
SuperMarioSF | the "no icon" state maybe just GSM itself. | 14:56 |
uvos__ | no | 14:56 |
uvos__ | not ususally | 14:56 |
SuperMarioSF | on that state I ran get-serving-cell-info, active access techonogy shown me GSM. | 14:58 |
SuperMarioSF | so for internet, should I run create-internet-context ? | 14:58 |
uvos__ | it should work automagicly - it dosent for me | 14:59 |
uvos__ | i use qmicli | 14:59 |
uvos__ | automagicly == you should be able to select mobile data as a connection in the connection dialog | 14:59 |
SuperMarioSF | that diaglog still in that grey out state, and I can't select wifi. | 15:00 |
uvos__ | then its broken just like on my device | 15:00 |
uvos__ | afaik i know it only works for wizzup | 15:00 |
uvos__ | maybe buzz too | 15:00 |
SuperMarioSF | one line says "No connections available", and clearly there some WiFi shown up there just can't click | 15:00 |
SuperMarioSF | wait | 15:01 |
SuperMarioSF | I ran enable-gprs, disable-gprs and enable-gprs again, that diaglog isn't grey out now. | 15:02 |
SuperMarioSF | but there is still no connection. | 15:02 |
Wizzup | uvos__: SuperMarioSF: try to install libicd-network-dummy | 15:02 |
Wizzup | this will work around your problem for now | 15:02 |
Wizzup | (and set up the gconf rule as the apt install says) | 15:02 |
SuperMarioSF | let me connect wifi very quick | 15:03 |
SuperMarioSF | seems disable-gprs can give your connection selection back | 15:03 |
Wizzup | I don't know, our sw doesn't assume other users control ofono typically | 15:04 |
Wizzup | in any case enable-gprs is not the right way to do it, but it could be a way to test | 15:04 |
SuperMarioSF | what privillage level should I run with that gconf command? | 15:05 |
SuperMarioSF | I ran with user privillage, no errors reported. | 15:06 |
SuperMarioSF | OK the menu won't just lockup now. | 15:07 |
SuperMarioSF | at least I can receive calls. | 15:08 |
SuperMarioSF | SMS works confirmed. | 15:11 |
SuperMarioSF | I'm having some meals, later. | 15:14 |
Wizzup | SuperMarioSF: great, some stuff is working then. | 15:44 |
SuperMarioSF | only 3G data not working at the moment | 15:45 |
Wizzup | Let me grab a coffee here and then I'll try to help | 15:45 |
SuperMarioSF | I'm not nearby a computer at the moment, but I will at home in about 2hrs. | 15:46 |
SuperMarioSF | if things can be done on phone itself then I'm ready. (WiFi available) | 15:46 |
SuperMarioSF | ok im back from wifi | 15:52 |
Wizzup | SuperMarioSF: so I am pretty sure gprs data itself will work (will help momentarily with testing that), -but- the provisionin might not | 15:53 |
SuperMarioSF | so the problem is let OS know how to use that network? | 15:54 |
Wizzup | can you install mdbus2, the pkg I linked before? | 15:55 |
SuperMarioSF | sorry I can't get that link because history isn't on my phone. can you post again? | 15:56 |
Wizzup | sure | 15:56 |
Wizzup | https://packages.debian.org/stretch/armhf/mdbus2/download | 15:56 |
uvos__ | theres also allways https://maedevu.maemo.org/irc-this-week.txt | 15:58 |
SuperMarioSF | ok package downloaded. do i need gdebi to resolve deps for that package? | 15:59 |
uvos__ | it has really limited depends | 16:01 |
uvos__ | if you even have to install something it should be easy to do by hand | 16:01 |
SuperMarioSF | ok so I just dpkg -i | 16:01 |
SuperMarioSF | done. no deps required. | 16:01 |
Wizzup | right | 16:03 |
Wizzup | SuperMarioSF: so mdbus2 -s org.ofono should show you various paths | 16:03 |
Wizzup | hopefully at least /motmdm_0 | 16:03 |
Wizzup | perhaps also /motmdm_0/context1 | 16:04 |
SuperMarioSF | yup. shown me 2 contexts, and an operator (460001, China Unicom) | 16:04 |
Wizzup | on the context path, you can run, uh, let me see | 16:04 |
Wizzup | like this: | 16:05 |
Wizzup | mdbus2 -s org.ofono /motmdm_0/context1 org.ofono.ConnectionContext.GetProperties | 16:05 |
Wizzup | this might tell if you if the apn and such are configured correctly | 16:05 |
SuperMarioSF | -> ({'Name': <'Internet'>, 'Active': <false>, 'Type': <'internet'>, 'Protocol': <'ip'>, 'AccessPointName': <'3gnet'>, 'Username': <'uninet'>, 'Password': <''>, 'AuthenticationMethod': <'none'>, 'Settings': <@a{sv} {}>, 'IPv6.Settings': <@a{sv} {}>},) | 16:05 |
Wizzup | try this: | 16:05 |
Wizzup | mdbus2 -s org.ofono /motmdm_0/context1 org.ofono.ConnectionContext.SetProperty Active true | 16:05 |
SuperMarioSF | looks good to me. (3gnet or 3gwap is ghe correct one) | 16:06 |
Wizzup | (and then use get properties in a bit to see if it's active after a bit) | 16:06 |
SuperMarioSF | returned: () | 16:06 |
Wizzup | that means it's done | 16:06 |
Wizzup | try the GetProperties again | 16:06 |
SuperMarioSF | -> ({'Name': <'Internet'>, 'Active': <true>, 'Type': <'internet'>, 'Protocol': <'ip'>, 'AccessPointName': <'3gnet'>, 'Username': <'uninet'>, 'Password': <''>, 'AuthenticationMethod': <'none'>, 'Settings': <{'Interface': <'wwan3'>, 'Method': <'static'>, 'Address': <'10.115.89.109'>, 'Netmask': <'255.0.0.0'>, 'Gateway': <'10.115.89.110'>, 'DomainNameServers': <['218.104.111.114']>}>, 'IPv6.Settings': <@a{sv} {}>},) | 16:07 |
Wizzup | so this means that the modem side of it works, now, try: sudo service icd2 restart | 16:07 |
SuperMarioSF | we got IP | 16:07 |
Wizzup | (this will reset your wifi) | 16:07 |
Wizzup | icd2 is what normally activates the context and then reads the ip and such | 16:07 |
Wizzup | at this point you might see your connection in the connection list where there is also wireless | 16:07 |
Wizzup | and yes some of this ui stuff is quite clumsy atm :) | 16:08 |
SuperMarioSF | I just see normal wifi and a dummy network | 16:08 |
SuperMarioSF | compared to my other phone, no other network available. | 16:09 |
Wizzup | hm, ok | 16:09 |
Wizzup | so the next step is to see if it's configured at al | 16:10 |
Wizzup | can you try to go to settings -> internet connections -> connections | 16:10 |
Wizzup | and then see if you see any non-wifi connection there? | 16:10 |
SuperMarioSF | the only non-wifi is that dummy network. | 16:10 |
Wizzup | ok, so it's not configured at all | 16:10 |
Wizzup | ok, for this I don't have an immediate fix. | 16:11 |
Wizzup | one way would be to configure it manually, but there's no telling if that will work | 16:11 |
Wizzup | others might have suggestions on how to just set up the ip manually, but I'm more interesting in figuring out the problem in our stack | 16:12 |
SuperMarioSF | I can do that if you prefer to dig depper. | 16:12 |
Wizzup | http://dpaste.com/BTEWVJ4TC | 16:12 |
Wizzup | this is the gconf entry for my KPN 3g data | 16:12 |
Wizzup | it was generated automatically | 16:12 |
Wizzup | (the ips get updated in there) | 16:13 |
Wizzup | the part after IAP/ is just a uuid | 16:13 |
Wizzup | so ideally you'd have something like that, but clearly it is not there | 16:13 |
SuperMarioSF | so i should replace the data with my own? | 16:13 |
Wizzup | not sure if it will work | 16:13 |
Wizzup | I am wondering, what is the second context that you are seeing? | 16:14 |
Wizzup | you said there were two | 16:14 |
SuperMarioSF | usually we have two profile in China. one is 3gnet for internet access. one is 3gwap for ISP proxied access. | 16:15 |
SuperMarioSF | same applied to all 3 ISPs in China. | 16:15 |
Wizzup | ok | 16:16 |
Wizzup | uvos__: do you have more than one context? | 16:16 |
SuperMarioSF | btw, usually the XXwap one also act as the network that MMS works on. | 16:17 |
Wizzup | SuperMarioSF: if you know how to write gconf data (it's painful, you have to do it per value and specify the type etc) it might be worth a shot to create an entry like mine, but it's not a great way | 16:17 |
uvos__ | Wizzup: my d4 is 100km away | 16:17 |
Wizzup | uvos__: :) | 16:17 |
uvos__ | buut | 16:18 |
SuperMarioSF | maybe I should make a script to do that automatically | 16:18 |
Wizzup | you went outside without a leste device in your pocket? shame :D | 16:18 |
uvos__ | sorry :( | 16:18 |
Wizzup | SuperMarioSF: well, we have code to do it automatically, we need to fix our code :) | 16:18 |
uvos__ | but its in my network at home, maybe i can ssh in :P | 16:18 |
Wizzup | SuperMarioSF: so, there might log entries in /var/log/maemo/icd2.log that could be helpful | 16:18 |
SuperMarioSF | ok let me connect to wifi | 16:19 |
uvos__ | heh i can | 16:19 |
uvos__ | Wizzup: can i check what you need over ssh only? | 16:20 |
SuperMarioSF | and I'm unable. ICD UI constantly say network error. (and that was a public open wifi) | 16:20 |
Wizzup | SuperMarioSF: there might be something like this: SIM %s seen for the very first time, provisioning. | 16:20 |
Wizzup | uvos__: yes, via mdbus2 | 16:20 |
Wizzup | SuperMarioSF: hm, wifi doesn't fail for me often, but usually on retry it works for sure | 16:20 |
uvos__ | im gona have to enable the modem via scripts (cant interact with icd gui ofc | 16:21 |
uvos__ | ) | 16:21 |
Wizzup | I can give you mdbus2 commands for it | 16:21 |
uvos__ | ok | 16:21 |
Wizzup | SuperMarioSF: actually all the logs that are relevant will be prefixed with [OFONO] | 16:21 |
sicelo | RE: space issues ... maybe we really should auto-run expandcard on first boot | 16:22 |
SuperMarioSF | searched that file, doesn't contain either 'SIM' nor 'seen'. something went wrong. | 16:22 |
Wizzup | sicelo: hm, yeah.. | 16:22 |
Wizzup | SuperMarioSF: what about 'ofono' ? | 16:22 |
uvos__ | ./list-modems "Technology = gsm" | 16:22 |
uvos__ | Wizzup: it dose show there | 16:22 |
uvos__ | for me at least | 16:22 |
Wizzup | uvos__: so just run this: mdbus2 -s org.ofono | 16:22 |
Wizzup | uvos__: and then see how many /motmdm_0/context* stuff you see | 16:23 |
SuperMarioSF | nope, not even for 'ofono# | 16:23 |
SuperMarioSF | * 'ofono' | 16:23 |
Wizzup | SuperMarioSF: what does dpkg -L libicd-network-ofono say? | 16:23 |
Wizzup | SuperMarioSF: (fwiw in case you did a case sensitive search, it's OFONO in the logs) | 16:23 |
SuperMarioSF | listed its file in that package. | 16:23 |
uvos__ | Wizzup: zero | 16:24 |
Wizzup | SuperMarioSF: hm, ok. | 16:24 |
SuperMarioSF | upper case tried, no result. | 16:24 |
uvos__ | just /motmdm_0/operator | 16:24 |
Wizzup | I suppose libicd-network-ofono has to be loaded on your system, because it onlines the modem for you | 16:24 |
Wizzup | uvos__: hm, with the modem online and pin provided etc? | 16:24 |
uvos__ | yeah | 16:24 |
Wizzup | uvos__: ok that's weird | 16:24 |
Wizzup | uvos__: we can debug that later I suppoes, normally ofono auto creates a context | 16:25 |
uvos__ | ( both done via scripts, i can try to reboot and try via icd over dbus if you can give me commands) | 16:25 |
Wizzup | nah won't matter I think | 16:25 |
SuperMarioSF | so ICD doesn't care about usb0 interface? | 16:25 |
Wizzup | nah | 16:26 |
SuperMarioSF | ok | 16:26 |
SuperMarioSF | I can setup a port forward and make the phone SSH on internet or ZeroTier. | 16:26 |
SuperMarioSF | if you need to debug it will be there. | 16:26 |
uvos__ | Wizzup: so i restarted ofono | 16:28 |
uvos__ | and now i have one context | 16:28 |
uvos__ | /motmdm_0/context1 | 16:28 |
Wizzup | ok | 16:29 |
Wizzup | maybe share this: | 16:29 |
Wizzup | mdbus2 -s org.ofono /motmdm_0/context1 org.ofono.ConnectionContext.GetProperties | 16:29 |
uvos__ | ({'Name': <'Internet'>, 'Active': <false>, 'Type': <'internet'>, 'Protocol': <'ip'>, 'AccessPointName': <''>, 'Username': <''>, 'Password': <''>, 'AuthenticationMethod': <'chap'>, 'Settings': <@a{sv} {}>, 'IPv6.Settings': <@a{sv} {}>},) | 16:29 |
Wizzup | that looks very unprovisioned | 16:29 |
uvos__ | y | 16:30 |
Wizzup | maybe we need to check, if, like with buzz, your isp is not known somehow | 16:30 |
uvos__ | ok - at least android is able to autoconfigure | 16:31 |
uvos__ | but idk about how it works :) | 16:31 |
Wizzup | I kind of know how we do it at least | 16:31 |
uvos__ | also qmicli has no issue | 16:31 |
Wizzup | mhm, not sure how that works | 16:31 |
Wizzup | maybe it asks the sim or modem somehow | 16:31 |
Wizzup | looks like the privisioning code needs some more love | 16:32 |
uvos__ | qmicli -d /dev/cdc-wdm1 --device-open-qmi '--wds-start-network=apn=internet' --wds-follow-network | 16:32 |
Wizzup | yeah ok you provide the apmn | 16:32 |
uvos__ | looks like i tell it the apn at least ;P | 16:32 |
sicelo | or uses a generic apn name. in many places, just 'internet' as apn tends to work | 16:32 |
Wizzup | s/apmn/apn/ | 16:32 |
Wizzup | for the record, I think this does the same: | 16:32 |
Wizzup | mdbus2 -s org.ofono /motmdm_0/context1 org.ofono.ConnectionContext.SetProperty AccessPointName "'internet'" | 16:32 |
uvos__ | doing that yelds ({'Name': <'Internet'>, 'Active': <false>, 'Type': <'internet'>, 'Protocol': <'ip'>, 'AccessPointName': <'internet'>, 'Username': <''>, 'Password': <''>, 'AuthenticationMethod': <'chap'>, 'Settings': <@a{sv} {}>, 'IPv6.Settings': <@a{sv} {}>},) | 16:33 |
uvos__ | perhaps unsuprising | 16:33 |
Wizzup | yeah | 16:34 |
Wizzup | this might work: | 16:34 |
Wizzup | net'"^C | 16:34 |
Wizzup | err | 16:34 |
Wizzup | mdbus2 -s org.ofono /motmdm_0/context1 org.ofono.ConnectionContext.SetProperty Active true | 16:34 |
Wizzup | or it will block until it times out unless you kick ofono in the meantime :P | 16:34 |
uvos__ | returned sucess instantly | 16:35 |
Wizzup | uvos__: does it show as active now? | 16:35 |
uvos__ | yeah i got a ip and everything | 16:35 |
Wizzup | SuperMarioSF: btw, super glad to hear that that calls worked for you, that cost us a lot of time to get mostly working | 16:35 |
Wizzup | uvos__: ok, so I guess for you also the gprs provisioning in our ui is the problem | 16:35 |
SuperMarioSF | yeah, call and sms is a tricky thing to work with. | 16:35 |
uvos__ | would be a neat fix - maybe the d4 would be closer than 100km if that would work ;P | 16:36 |
uvos__ | *neat to fix | 16:36 |
Wizzup | SuperMarioSF: just to be clear there are still some problems with our ofono | 16:36 |
SuperMarioSF | I was playing some full feature LTE modem device before. Setup for a call with only serial ttys are painful... | 16:36 |
Wizzup | like I sometimes repeatedly 'receive' the same sms on my d4 | 16:36 |
uvos__ | SuperMarioSF: on droid 4 its relatviely easy - the d4 implements everything in hw | 16:36 |
uvos__ | like you can issue a call and shutdown linux | 16:36 |
uvos__ | call will continue | 16:37 |
Wizzup | will be harder to hang up :P | 16:37 |
uvos__ | still yeah was plenty work | 16:37 |
uvos__ | Wizzup: well the other party has to hang up then :P | 16:37 |
SuperMarioSF | or wait until battery dies... | 16:38 |
sicelo | the modem will power the speakers? | 16:38 |
uvos__ | cpcap contains all the hw to drive the spekers, and you gave the channels to the modem, so yes | 16:39 |
uvos__ | if you shutdown linux but dont change cpcap reg state | 16:39 |
uvos__ | the call would work | 16:39 |
uvos__ | cpcap also handles the some of the audio procesing, other processing is handled by the modem | 16:39 |
uvos__ | the cpu dose nothing | 16:39 |
Wizzup | bbl | 16:41 |
SuperMarioSF | reminds me about some early Windows Mobile phones. They don't have 'Airplane mode', they have 'Phone Power Down' mode, and that just disable modem alright. | 16:42 |
SuperMarioSF | maybe if Windows Mobile freezes the call should just go on. | 16:42 |
sicelo | otoh, it scares me ... because now it means the modem/cpcap have a life of their own. such setup really absolutely needs kill switches | 16:43 |
uvos__ | sicelo: no | 16:43 |
uvos__ | sicelo: since the modem can not talk to cpcap at all | 16:43 |
uvos__ | sicelo: it only has analog audio connections to it | 16:43 |
uvos__ | sicelo: so it cant give itself the speaker or the mic | 16:43 |
uvos__ | sicelo: on xt860 thats the end of it | 16:44 |
SuperMarioSF | so I guess if hacker got your modem hacked, the only thing they can do is rickroll you. (x) | 16:44 |
uvos__ | on d4 the lte modem DOSE have access to cpacap and could control its registers | 16:44 |
uvos__ | but it dosent have a audio interface | 16:45 |
uvos__ | so the only way something nefarious could happen is if the lte modem and the qcom modem comunicated over the internet | 16:45 |
uvos__ | and the lte modem gave the qcom modem the mic and speakers | 16:45 |
uvos__ | fiarly absurd thret model | 16:45 |
uvos__ | also the lte modem is not powered in leste | 16:46 |
uvos__ | ofc all modems have gps and the tower knows roughtly where the modem is anyhow | 16:47 |
uvos__ | so the modem alone can be nafarious enough | 16:47 |
sicelo | especially when it can be doing things with cpcap behind linux' back | 16:48 |
SuperMarioSF | no wonder I know some people who make DIY phones doesn't have celluar network at all, they use LoRA instead. | 16:48 |
uvos__ | sicelo: as state this is impossible on leste | 16:48 |
uvos__ | *stated | 16:48 |
uvos__ | unless you use a gpio hack to power up the lte modem | 16:49 |
uvos__ | (from linux) | 16:49 |
uvos__ | if your concerned about that run xt860 i gues | 16:49 |
SuperMarioSF | if there is a way to bridge some joint point to do that? | 16:49 |
sicelo | i don't know how much one can trust cpcap to begin with, as "cpcap contains all the hw to drive the spekers, and you gave the channels to the modem, so yes" | 16:50 |
uvos__ | if you expect cpcap to have some kind of hw level surveilnce then you might as well assume omap4 also has soemthing like that | 16:51 |
uvos__ | omap4 also has pelty of microcode you could hide something in | 16:52 |
sicelo | i know | 16:52 |
sicelo | i didn't know there were additional places to hide it :p | 16:53 |
buZz | Wizzup: i do feel isp autoconfig needs some hard work | 17:48 |
buZz | my sim was roaming in .be and .lu recently, finally saw 3G again, but couldnt get data working over it :( | 17:49 |
sicelo | i forgot - are we using mobile-broadband-something-info? if we are, maybe you can add your connection profiles to it. they're very accepting upstream. at least i added mine without issues | 17:50 |
uvos | maybe just figure out what android dose and use that (maybe only as fallback) | 18:04 |
uvos | its guaneteed to work everywhere since operators have great incentive | 18:04 |
norayr | > uvos__ wrote: | 18:07 |
norayr | > then theres a boot menu with entries you can edit via a text file, by default theres 2 entries, leste from sd and stock android from /system | 18:07 |
norayr | > but you can have as many entries as you want | 18:07 |
norayr | So i didn't understand can we have an entry which just gets charge? or we should use console maemo option in hope it'll be enough power to boot until charge started? | 18:07 |
norayr | we moved we should move again soon. so i did not unpack most of the boxes and i don't know where is my generic battery charger. But i was abloe to disassemble another charger for canon batteries | 18:08 |
uvos | if you boot with usb plugged in it the nromal leste entry should enter charge mode | 18:08 |
uvos | it dosent at least for sicleo | 18:08 |
uvos | but yeah | 18:08 |
uvos | other than that | 18:08 |
norayr | and keep droid's battery several minutes with those contacts | 18:09 |
uvos | the stock android option is the quickest | 18:09 |
uvos | and will enter androids charge mode | 18:09 |
norayr | and it was enough for it to boot to console leste and start taking charge. but i was not able to use ih for several days before | 18:09 |
norayr | now it gets charge | 18:10 |
uvos | so really the d4 should never get into the state where it cant boot all the way to leste anyhow | 18:10 |
uvos | since we have leste shutdown early to prevent it | 18:11 |
uvos | on low bat | 18:11 |
uvos | so unless you messed with the values or you battery is in very poor state or you leave the device unantended for multiple weaks while off and empty | 18:11 |
uvos | it should not happen | 18:11 |
norayr | SuperMarioSF: congrats, having droid is fun and you already have fun with it | 18:12 |
uvos | booting to console leste should be about the same as booting to charge mode | 18:12 |
uvos | except the backlight will be on | 18:12 |
uvos | so its worse | 18:12 |
norayr | oh | 18:12 |
uvos | otherwise its the same (kernel and sysinit userspace have to boot before we can start charging) | 18:13 |
norayr | in my experience booting to leste console mode is faster and therefore more chances to start booting | 18:13 |
uvos | charge mode skips more parts | 18:13 |
norayr | it takes 0.6 amper in this state | 18:13 |
uvos | so its impossible for it to be faster | 18:13 |
uvos | they both boot the same kernel and the sysinit runlevel | 18:14 |
norayr | androids charge mode didn't help me... | 18:14 |
uvos | and then console mode boots another runlevel | 18:14 |
uvos | if androids charge mode dosent work | 18:14 |
uvos | this strongly suggests android dosent like the battery | 18:14 |
norayr | it was drawing 0.1-0.2 amper for whole night, then | 18:14 |
norayr | it was warm in the morning | 18:15 |
norayr | but was not able to boot. | 18:15 |
norayr | even a bit | 18:15 |
norayr | yes android doesnt like most of my batteries | 18:16 |
norayr | but droid3 is so beautiful | 18:17 |
SuperMarioSF | OK I'm back | 18:17 |
SuperMarioSF | I'm at home | 18:17 |
SuperMarioSF | and ready for port forwarding | 18:17 |
norayr | alas its maemo boots and shuts down severl minutes later | 18:17 |
norayr | but even near sony xperia it looks sooo cool | 18:18 |
uvos | norayr: yeah so if android dosent charge your batteries while in charge mode that suggests hw failure or at least that android thinks charging is unsafe | 18:18 |
norayr | unlike droid4 | 18:18 |
norayr | it did chalge the battery it came with | 18:18 |
uvos | it checks way more parameters than we do | 18:18 |
norayr | but that battery was in bad shape | 18:18 |
uvos | it checks temperature deltaT and deltaV | 18:18 |
uvos | to ensure its safe | 18:18 |
norayr | very interesting | 18:18 |
uvos | (at least) | 18:19 |
SuperMarioSF | btw I have a Sony Ericsson SK17i | 18:19 |
SuperMarioSF | very tiny yet powerful phone | 18:19 |
SuperMarioSF | ok my tmate session created. | 18:28 |
buZz | SuperMarioSF: need some help with d4 gprs? | 18:28 |
SuperMarioSF | yup | 18:29 |
buZz | whats up | 18:29 |
SuperMarioSF | that why I am creating tmate session | 18:29 |
buZz | whats a tmate? | 18:29 |
SuperMarioSF | a terminal share service, via SSH or web terminal | 18:29 |
buZz | ah, multiplayer ssh? screen -x | 18:29 |
SuperMarioSF | yup | 18:30 |
buZz | 'connecting to gprs' for me on d4 causes all other networks to disconnect, btw | 18:30 |
SuperMarioSF | no worries, the link is over usb0 | 18:31 |
SuperMarioSF | here the links: https://paste.mozilla.org/YLO2Vc0D | 18:31 |
SuperMarioSF | use link there to access my d4 | 18:33 |
SuperMarioSF | that page will expire in 1hour | 18:33 |
sicelo | btw, RE: provisioning - at least with networks in my country, i've had to manually add apn. android has never auto-provisioned them | 18:34 |
sicelo | of course, there's an USSD code one can dial, where operator sends the phone a configuration message. i think ofono knows how to handle those | 18:34 |
SuperMarioSF | the link provided are Web based terminal and SSH connection, for both read-only and read-write types. | 18:35 |
sicelo | so for my networks, i guess mobile-broadband-provider-info is more useful than built-in mechanisms | 18:35 |
* sicelo installs a trojan on SuperMarioSF's phone :-D | 18:36 | |
SuperMarioSF | oh, I am monitoring the session of course. | 18:36 |
SuperMarioSF | since this session is provided to help you guys pinpointing the issue, if something went wrong, I can just reinstall, or replace with a new phone if needed. | 18:37 |
sicelo | what problem are you troubleshooting btw? | 18:38 |
SuperMarioSF | things working are: calling, SMS | 18:38 |
SuperMarioSF | problem is: ICD unable to provision connection profile properly | 18:38 |
SuperMarioSF | ofono knows what APN my ISP provides. | 18:39 |
SuperMarioSF | mdbus2 was installed. | 18:40 |
SuperMarioSF | the password are defaults, so therefor this session is inside a tmate, not a internet facing SSH connection. | 18:41 |
SuperMarioSF | to prevent restarting ICD cause connection termination, the SSH link to d4 is over USB RNDIS | 18:45 |
SuperMarioSF | and that USB connected to my router. | 18:45 |
SuperMarioSF | everything just works. | 18:46 |
SuperMarioSF | oh btw a question | 18:47 |
SuperMarioSF | how can I type ":" and ";" with d4 hardware keyboard? | 18:47 |
uvos | ok+: | 18:48 |
SuperMarioSF | I almost can't quit vim because I have no idea how to type them. | 18:48 |
uvos | see wiki | 18:48 |
SuperMarioSF | oh i missed that one | 18:48 |
SuperMarioSF | since no one actually joined the SSH session, I'm closing it. | 19:15 |
Wizzup | SuperMarioSF: sorry, I was literally running to make my connecting flight | 19:20 |
Wizzup | we can do this remote debug another day? | 19:20 |
SuperMarioSF | sure | 19:21 |
SuperMarioSF | when you have time just mail me if i'm not online. | 19:21 |
Wizzup | alright | 19:21 |
SuperMarioSF | usually I'm online at 02:00 ~ 16:00 UTC | 19:22 |
Wizzup | for me that is 0800 - 0100 utc | 19:22 |
SuperMarioSF | plenty overlap, that's ok | 19:23 |
SuperMarioSF | if you can do debug work tomorrow, I will leave my d4 at home connected to router, so I can setup connection remotely for you. | 19:24 |
SuperMarioSF | it will be connected via usb0 so restart icd won't be an issue. | 19:24 |
SuperMarioSF | and I just test, it can even survive a reboot. | 19:25 |
Wizzup | alright, sweet | 19:25 |
SuperMarioSF | hmmmm not excatly | 19:26 |
Wizzup | reboot can take a few mins | 19:26 |
SuperMarioSF | seems reboot caused my router not picking the device. | 19:26 |
SuperMarioSF | however I can reset the USB on router so maybe it will work again. | 19:26 |
SuperMarioSF | I'm testing this | 19:26 |
SuperMarioSF | there are some known issue with RouterOS based devices on RNDIS interface. that is one of it. | 19:31 |
SuperMarioSF | usually need to reset USB port once to get it working. | 19:31 |
SuperMarioSF | sometimes this may get even worse, it won't pick up any RNDIS interface, and you have to bring a real Android to enable RNDIS on it once, then it will happily accept any RNDIS devices afterwards. | 19:32 |
SuperMarioSF | oops | 19:33 |
SuperMarioSF | I have some bad news | 19:33 |
SuperMarioSF | seems it can't reset itself via USB hub | 19:33 |
SuperMarioSF | a powered hub | 19:33 |
SuperMarioSF | Time to add more things to make it happen. | 19:34 |
SuperMarioSF | I'm adding a smart switch. | 19:34 |
SuperMarioSF | OK it worked, finally | 19:40 |
Wizzup | you could run the gadget in local.d as well | 19:40 |
SuperMarioSF | I guess the problem is without a power reset on USB, gadget won't be working properly after reboot. | 19:41 |
SuperMarioSF | so I add a smart switch to solve the issue, and it worked, | 19:41 |
Wizzup | ok | 19:42 |
SuperMarioSF | just send USB bus reset on router, following with a power cycle on Hub's power, everything will be back online. | 19:42 |
SuperMarioSF | btw maybe customizing device vendor info will be nice. currently it was Foo Inc Bar Device. | 19:43 |
SuperMarioSF | better be: Maemo Leste Fundation Droid 4 Device. ;P | 19:44 |
SuperMarioSF | * typo: Foundation | 19:45 |
Wizzup | SuperMarioSF: actually I set that to foo and bar :) | 19:47 |
Wizzup | https://github.com/maemo-leste/hildon-usb-gadgets/blob/master/src/gadget-helper.c | 19:47 |
Wizzup | https://github.com/maemo-leste/hildon-usb-gadgets/blob/master/src/mass-storage.h | 19:48 |
SuperMarioSF | hmmmm 🤔 | 19:48 |
Wizzup | we probably need to make that something sensible :D | 19:48 |
SuperMarioSF | what if this can be managed via config file? | 19:50 |
Wizzup | fine by me | 19:53 |
Wizzup | some glib ini reading or so, we do this in many places? | 19:53 |
Wizzup | bbl | 19:54 |
freemangordon1 | maybe extend and use https://github.com/maemo-leste/osso-product-info | 20:03 |
uvos | i mean implemantation wise thats about as terrible as it comes | 20:28 |
uvos | a hardcoded char array of stuff | 20:29 |
uvos | anyhow an effort should be made to get this from more standart places | 20:30 |
uvos | the gadget name could be related to of name of the device for instance | 20:30 |
sicelo | What's an example standard place? | 20:34 |
uvos | /proc/device-tree/model | 20:34 |
sicelo | I think pmOS uses what they call deviceinfo | 20:34 |
sicelo | Their own thing, that is | 20:35 |
uvos | im sure deviceinfo gathers information from places like /proc/device-tree | 20:35 |
uvos | (instead of a hardcoded list) | 20:35 |
sicelo | You assume :-) | 20:36 |
uvos | anyhow yeah joineing forces with pmos is also an option | 20:36 |
uvos | thats preferable to resurecting some maemo dianosaur hardcoded list | 20:36 |
sicelo | Deviceinfo is a static file, hardcoded per device | 20:36 |
uvos | if so then at least its a file ;) | 20:36 |
uvos | also ok deviceinfo makes sense beacuse a primary user is pmbootstrap | 20:39 |
uvos | obv you cant gather info in pmboostrap, it needs to be told what to build | 20:40 |
uvos | so osso-product-info and Deviceinfo have different contstraints | 20:40 |
uvos | and deviceinfo is sane while osso-product-info is very not | 20:40 |
* sicelo doesn't see the difference so much | 20:45 | |
sicelo | pmOS also aims to support 200+ devices, so deviceinfo is geared towards that | 20:46 |
Wizzup | uvos: yeah I wrote it mostly for debug and test purposes, when it wasn't even working in kernel | 21:51 |
uvos | Wizzup: i wasent talking about hildon-usb-gadgets | 21:54 |
uvos | Wizzup: i get that | 21:54 |
uvos | Wizzup: i ment osso-product-info | 21:54 |
Wizzup | ah | 21:54 |
Wizzup | yeah we want to remove/deprecate tat | 21:55 |
Wizzup | s/tat/that/ | 21:55 |
freemangordon1 | and use what instead? | 21:55 |
freemangordon1 | not that I disagree, I am just not aware of the replacement | 21:56 |
uvos | try and get away with using just the info that the kernel provides | 21:56 |
Wizzup | I mean the interface is not bad per se | 21:56 |
uvos | if not possible do soemthing like pmos | 21:56 |
uvos | (ie with text files) | 21:56 |
freemangordon1 | how is that different? | 21:56 |
uvos | ? | 21:56 |
freemangordon1 | I mean -we can easily change osso-product-info to use text files | 21:56 |
uvos | how ist that not different to a hardcoded c array | 21:56 |
freemangordon1 | it is the same - it should be maintained by us | 21:57 |
Wizzup | well the interface stays the same | 21:57 |
freemangordon1 | either files or C array | 21:57 |
uvos | well using pmos implentation gets us the values for lots of devices, otherwise sure | 21:57 |
freemangordon1 | I see | 21:57 |
Wizzup | most not mainline though, but yeah | 21:57 |
buZz | is hildon-usb-gadgets a tool for configuring usb-gadget modes? | 21:57 |
Wizzup | arg vivatel still just fails to up for my dutch sim in BG :) | 21:58 |
Wizzup | buZz: yes | 21:58 |
buZz | nice | 21:58 |
buZz | Wizzup: i have the feeling the APN settings might not be properly configured in 'weird' situations, like roaming | 21:59 |
Wizzup | + https://github.com/maemo-leste/hildon-status-bar-usb | 21:59 |
buZz | nice | 21:59 |
Wizzup | buZz: this also fails on my n900 and other phones | 21:59 |
Wizzup | it's probably unrelated to leste specifically | 21:59 |
buZz | now we just need to write a PC suite? :D | 21:59 |
Wizzup | other operators here work | 21:59 |
buZz | hehe | 21:59 |
Wizzup | pc suite is pretty easy, the modem access of it not so much | 21:59 |
buZz | unless we abuse usbgadget for it? | 22:00 |
buZz | i think that can emulate a serial port anyway, we could just 1:1 pass it through with a small tool | 22:00 |
freemangordon1 | Wizzup: you are in BG? | 22:00 |
buZz | (i dont know what pc suite all did) | 22:00 |
uvos | why do we want modem access thrugh pc suite? | 22:01 |
buZz | no clue | 22:01 |
buZz | i do love the idea of connecting a droid4 to my laptop for 'internet access' with USB | 22:01 |
uvos | well you dont need direct access to the modem for that | 22:02 |
buZz | but usb ethernet gadget thingy can be used for that, i think | 22:02 |
uvos | usbnet and some routing | 22:02 |
buZz | yeah | 22:02 |
freemangordon1 | uvos: oh, seems you have fixed the auto-relock thingie (at least judging by mce changelog) | 22:02 |
buZz | probably even better without | 22:02 |
Wizzup | freemangordon1: yup, just now | 22:02 |
uvos | freemangordon1: yes but its a stop gap | 22:02 |
buZz | Wizzup: thats always such a weird idea in my head, if we are 'maemo 7' why are we just trying to be 1:1 'maemo 5' | 22:02 |
Wizzup | buZz: hm? | 22:03 |
uvos | i still need to rewirte the input layer to use libinput | 22:03 |
buZz | i mean, many things in UI are done away with 'thats how fremantle did it' | 22:03 |
buZz | etc | 22:03 |
Wizzup | buZz: this isn't about UI | 22:03 |
Wizzup | it's about modem func | 22:03 |
Wizzup | and the sim + this operator is just weird | 22:03 |
buZz | 100% of my sims have that experience | 22:03 |
freemangordon | never had any issue abroad | 22:04 |
uvos | so that it can share device share configureation with xorg/sdl/wayland (think callibration, ignored devices, what is a gamepad/mouse/ts/etc) | 22:04 |
freemangordon | besides most of the places now have disable 3G | 22:04 |
buZz | i'm gonna buy a second gprs capable sim just for testing mosh powerusage btw, gonna look for a provider that allows me just a 1 month contract | 22:04 |
Wizzup | freemangordon: this is about kpn (dutch) with vivatel/vivacom specificallu | 22:04 |
Wizzup | buZz: cool | 22:04 |
freemangordon | Wizzup: yeah | 22:04 |
buZz | shouldnt cost me >10 eu , i -expect- | 22:04 |
Wizzup | uvos: libinput as opposed to, /dev/input/ ? | 22:04 |
buZz | maybe i can put a usb charger connected to that d4 with a sonoff so i can turn charging on/off remotely | 22:05 |
uvos | Wizzup: right atm it just opens all the event devices itself and reads them | 22:05 |
Wizzup | I mean, it's a pretty sane way, no? :) | 22:05 |
uvos | no | 22:05 |
Wizzup | what's lacking? | 22:06 |
Wizzup | just curious | 22:06 |
uvos | because not all input devices should be used, and callibration should be applied and mce should not have its own heuristics of what is a mouse/keyboard/gamepad etc | 22:06 |
uvos | libinput uses udev hints and has a huge list of exceptions that makes it work reliably | 22:06 |
uvos | mce cant compeat | 22:06 |
uvos | also we configure xorg via libinput (callibration - ignoreing some stuff) | 22:06 |
uvos | mce needs its own special config to do the same thing - again. | 22:07 |
Wizzup | ah ok, that makes sense, so it's for properties and such | 22:07 |
uvos | yes | 22:07 |
buZz | you can have some intelligence about which type of /dev/input/event devices to try | 22:08 |
buZz | like by bus | 22:08 |
buZz | or something | 22:08 |
uvos | righ libinput dose all these things | 22:09 |
uvos | and has a large exception table to fix where things go wrong | 22:09 |
uvos | and can be configured via udev rules | 22:09 |
uvos | and that then applies to xorg/wayland/sdl etc | 22:09 |
uvos | all in one place | 22:09 |
sicelo | uvos: re - modem via 'pc suite' ... usbnet + routing is difficult for the average user, while handling (AT) modems is very easy even in OSes such as Windows | 22:17 |
uvos | sicelo: we cant really assume every phones modem is goning to have a at interface | 22:18 |
uvos | sicelo: so this means implementing a modem emulator | 22:18 |
uvos | android presents a usbnic, that seams best | 22:18 |
uvos | and is easy to deal with on windows too | 22:19 |
sicelo | hence i added AT in () ... the point is - this stuff is seamless in almost any OS, unlike messing with usb-net. have you ever seen how RNDIS stuff works in Windows? | 22:19 |
sicelo | anyway ... | 22:19 |
uvos | sure its a bit of a mess | 22:19 |
uvos | still better than routing out the raw modem imo | 22:19 |
uvos | btw mbm can do this iirc | 22:20 |
uvos | on d4 | 22:20 |
uvos | cpcap can directly route the modem out to the usb port | 22:20 |
sicelo | cpcap :-) | 22:20 |
uvos | not a fan of the chip eh? :P | 22:21 |
sicelo | buZz: tbh, leste was never really going to be maemo '7' in that sense. it was simply 7 because 6 was already taken by Harmattan. Otherwise, it was always going to be based largely on Fremantle. at least in the beginning | 22:23 |
buZz | sicelo: right, sure, i can understand it from starting point, just not from dailydriver + 'we have the wheel' standpoint, always ;) | 22:29 |
sicelo | maybe it'll end up like iOS ;-) | 22:30 |
buZz | sicelo: taking screenshots of programs so you dont see the recovery from swap? | 22:38 |
sicelo | that too | 22:39 |
buZz | :) | 22:40 |
buZz | i kinda like the live thumbnails in app switcher | 22:40 |
buZz | or whatever its called | 22:40 |
sicelo | others don't | 22:40 |
sicelo | the nature of people :-) | 22:41 |
buZz | i have no idea, i barely used smartphones ever since wince/symbian wasnt a thing anymore ;) | 22:41 |
uvos | i like it, i just dont think its worth it resource wise. | 22:41 |
buZz | the live thumbs arent worth it? | 22:41 |
uvos | no | 22:41 |
buZz | does it eat a lot of ram? | 22:41 |
uvos | they force a extra full screen buffer of eatch app at all times | 22:41 |
buZz | or resources? i havent noticed much cpu load from it | 22:41 |
uvos | (even when not in the switcher) | 22:41 |
buZz | but didnt look at power stats specifically | 22:41 |
buZz | gee | 22:42 |
uvos | also they eat a lot of ram bandwith | 22:42 |
uvos | = a lot of power | 22:42 |
buZz | maybe we can change that behaviour to use caches more efficiently? | 22:42 |
buZz | s/caches/buffers/ | 22:42 |
uvos | not really | 22:43 |
uvos | you could suspend compositing when not in the switcher | 22:44 |
uvos | but this requires lots of rearchitectureing in hildon | 22:44 |
sicelo | buZz: sometimes i miss S80 (found in the Communicators, 9500/9300). Apparently that's where Hildon traces its roots from | 22:48 |
buZz | i think we could do 'better' for the future ppl that use leste once its 'gamma' | 22:50 |
sicelo | sure :-) | 22:57 |
sicelo | btw, of late, when my d4 manages to get into charge mode, it gets stuck there. only a hard reset allows me to get to a working system | 23:23 |
Wizzup | new kernel problem? | 23:27 |
uvos | no this is the drm hang | 23:45 |
uvos | charging sdl gets into the same drm hang as xorg | 23:46 |
Wizzup | in the ioctl | 23:46 |
Wizzup | ? | 23:46 |
uvos | yeah | 23:46 |
uvos | it uses blocking io so its a little different | 23:46 |
uvos | but yes essentally | 23:46 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!