**** BEGIN LOGGING AT Sat Nov 12 03:00:00 2016
Nov 12 07:13:56 <brolin_empey> I was using Microsoft Edge on Windows 10 to listen to Internet radio (di.fm) but I discovered that Edge stops playing audio when Windows turns off the computer’s displays due to an inactivity timeout.  When the displays are back on, Edge continues playing audio.  Is this behaviour caused by a change in Edge?  Or by something else, such as Flash Player or Windows?  Now I am using Internet Explorer 11 to see if it behaves the same as Edge.
Nov 12 07:17:35 <brolin_empey> No, I do not use Edge nor IE as my primary Web browser.
Nov 12 07:18:27 <joga> oh, this wasn't ##windows...
Nov 12 07:18:32 <joga> maybe ask there?
Nov 12 07:21:00 <brolin_empey> ##windows moves very quickly.  Anyway, I will soon have an answer from my own experimentation.
Nov 12 07:22:38 <joga> hdmi audio?
Nov 12 07:23:57 <brolin_empey> joga: No, analog line-level output from the audio controller integrated on the motherboard.
Nov 12 07:24:21 <auenf> brolin_empey, surface?
Nov 12 07:26:00 <brolin_empey> auenf: No, Asus P5E-VM DO LGA775 (Core 2 era) desktop motherboard.
Nov 12 07:26:37 <auenf> had an issue with a surface dropping sound from a hangouts in chrome when the screen saver came on
Nov 12 07:26:54 <joga> win10 does all sorts of wacky things
Nov 12 07:27:21 <auenf> ie, in a hangouts, screen saver was at 1min, when screen saver came on, hangouts became silent
Nov 12 07:27:28 <auenf> wake screen, and hangouts resumed
Nov 12 07:28:34 <brolin_empey> auenf: My screen saver is automatic power management of the displays.
Nov 12 07:30:15 <auenf> this surface didnt have a real screensaver, think it was just blank screen or something like what you're saying
Nov 12 07:32:03 <KotCzarny> maybe its aggressive power saving mode that simply suspends the whole os
Nov 12 07:33:02 <auenf> it'd be win10 killing audio when monitors sleep or when screen saver starts
Nov 12 07:33:43 <auenf> sounds like whats happening to edge, i assume ie will be the same, and i think chrome would be the same also
Nov 12 07:34:19 <brolin_empey> KotCzarny: No, my computers are always running but the OS automatically turns off the displays after an inactivity timeout.
Nov 12 09:07:47 <brolin_empey> I wonder why the format of the message identifiers generated by Mozilla Thunderbird was changed a few months ago.
Nov 12 15:41:59 <freemangordon> Pali: https://github.com/community-ssu/iphbd/commit/cddcdd817f82056dd51b82d148dd1ad7ea6cebce
Nov 12 15:44:08 <Pali> ok
Nov 12 17:09:19 <DocScrutinizer05> HDMI and similar shit sucks big time, when you can't hotplug stuff since it's not getting re-initialized
Nov 12 17:11:29 <DocScrutinizer05> Pali: do you know about usage of N900 IHF_EN signal (from GAIA to speaker amplifier)? I'm poarticularly interested who's operating that GPIO and when and how
Nov 12 17:11:59 <bencoh> huhu
Nov 12 17:12:22 <DocScrutinizer05> GAIA/TWL4030 pin N14 "GPIO7/PWM1"
Nov 12 17:12:23 <bencoh> DocScrutinizer05: at least it's not a mic? :p
Nov 12 17:12:48 <DocScrutinizer05> bencoh: nope, it enables the amp
Nov 12 17:13:17 <bencoh> DocScrutinizer05: yeah I know :p
Nov 12 17:13:58 <Pali> grepping kernel source code...
Nov 12 17:14:07 <Pali> what is number of that gpio?
Nov 12 17:14:21 <DocScrutinizer05> it's one of the nasty legacy signals we might need to keep compatible to N900 in new devices, if we want audio (closed blobs) to work
Nov 12 17:14:32 <DocScrutinizer05> Pali: GPIO7
Nov 12 17:14:49 <bencoh> gpio7 on gaia, not on omap3
Nov 12 17:14:53 <DocScrutinizer05> yes
Nov 12 17:15:09 <Pali> speaker-amplifier-gpios = <&twl_gpio 7 GPIO_ACTIVE_HIGH>;
Nov 12 17:15:15 <DocScrutinizer05> exactly
Nov 12 17:15:15 <Pali> could be this?
Nov 12 17:15:23 <DocScrutinizer05> :-)
Nov 12 17:15:43 <DocScrutinizer05> where does it show up?
Nov 12 17:15:48 <bencoh> dts
Nov 12 17:15:51 <DocScrutinizer05> where is it uised
Nov 12 17:15:51 <Pali> sound: n900-audio { compatible = "nokia,n900-audio"; ... speaker-amplifier-gpios = <&twl_gpio 7 GPIO_ACTIVE_HIGH>; };
Nov 12 17:15:54 <Pali> this is in DT
Nov 12 17:16:10 <Pali> going to look at that speaker-amplifier-gpios property
Nov 12 17:16:29 <DocScrutinizer05> in DT? well, that's great for new kernel, but questionable if it works with old audio blobs
Nov 12 17:17:28 <DocScrutinizer05> I assume they (Nokia) has some very special crap in there, to realtime enable amp for e.g. keybd and touchscreen audio click feedback
Nov 12 17:17:59 <bencoh> ?
Nov 12 17:18:17 <DocScrutinizer05> dunno if it's handled via generic ALSA or inside some dang closed blob
Nov 12 17:18:36 <Pali> https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/sound/soc/omap/rx51.c#n465
Nov 12 17:18:43 <bencoh> honestly I wouldnt care much anyway....
Nov 12 17:18:46 <Pali> here is read that gpio number from DT ^^^
Nov 12 17:19:54 <DocScrutinizer05> yes, that looks pretty good
Nov 12 17:20:34 <DocScrutinizer05> so everything in 'higher tiers' should use that symbolic resource rather than direct access to GPIO7, right?
Nov 12 17:20:47 <Pali> yes
Nov 12 17:20:47 <DocScrutinizer05> even in maemo kernel and OS?
Nov 12 17:20:58 <Pali> going to check maemo kernel
Nov 12 17:21:08 <DocScrutinizer05> thanks a megaton! :-)
Nov 12 17:21:46 <DocScrutinizer05> realtime audio output is tricky, _very_ tricky
Nov 12 17:22:09 <brolin_empey> Heh: It is nice that http://info.cern.ch/ , the first (public) Web site, is back online but it would be nicer if it was actually accessible with a Web browser from 1995 or earlier, such as the Win32 version of Netscape Navigator 1.22, the earliest Win32 (as opposed to Win16 or Win32s) version of Netscape/Mozilla I have found, I have running on Windows 10 for x86-64.
Nov 12 17:22:11 <Pali> maemo kernel has hardcoded it into that sound/soc/omap/rx51.c file :-(
Nov 12 17:22:20 <Pali> #define RX51_SPEAKER_AMP_TWL_GPIO	(192 + 7)
Nov 12 17:22:25 <DocScrutinizer05> you want amp normaly shut down, but you also want minimal overhead/delay when output of a feedback click or whatever
Nov 12 17:22:50 <Pali> \/* REVISIT: TWL4030 GPIO base in RX51. Now statically defined to 192 */
Nov 12 17:23:03 <DocScrutinizer05> ((maemo kernel has hardcoded it)) thought as much
Nov 12 17:23:55 <Pali> but looks like gpio number is not exported to userspace
Nov 12 17:24:09 <DocScrutinizer05> so access is up to any arbitrary 'userland' app, or are they all supposed to use a kernel function which we could patch to change the GPIO assignment?
Nov 12 17:24:22 <Pali> and even not outside of that file
Nov 12 17:24:39 <Pali> so only kernel and onlt that sound/soc/omap/rx51.c can access it
Nov 12 17:24:49 <brolin_empey> Yes, I know that I can use https://www.jwz.org/hacks/http10proxy.pl .
Nov 12 17:24:50 <DocScrutinizer05> GOOD!
Nov 12 17:24:51 <Pali> userland nad other kernel modules needs to use abstract kernel functions
Nov 12 17:24:52 <DocScrutinizer05> :-))
Nov 12 17:25:01 <DocScrutinizer05> excellent
Nov 12 17:25:06 <DocScrutinizer05> many many thanks!
Nov 12 18:49:54 <freemangordon> Pali: how hard will be to make iphbd out-of tree dkms module?
Nov 12 18:50:02 <freemangordon> I've never played with dkms
Nov 12 18:50:26 <Pali> freemangordon: if you have working kernel module (e.g. can be compiled), dkms packaging is easy
Nov 12 18:50:47 <Pali> I can prepare dkms packaging, this is not a problem
Nov 12 18:50:59 <Pali> I already have tens of dkms packages on my laptop
Nov 12 18:51:26 <freemangordon> Pali: iphb.ko is open-sourced, right?
Nov 12 18:51:29 <Pali> yes
Nov 12 18:51:34 <Pali> in maemo kernel are sources
Nov 12 18:51:35 <freemangordon> so it shouldn't be a problem
Nov 12 18:51:55 <freemangordon> ok, I'll look into it
Nov 12 18:52:01 <Pali> we just needs to decide which sources to get... from maemo? from harmattan?
Nov 12 18:52:08 <Pali> or are any other new versions for new kernels?
Nov 12 18:52:16 <Pali> I can prepare dkms package
Nov 12 18:52:27 <freemangordon> i guess they are identical
Nov 12 18:52:45 <Pali> ok, I will look at it
Nov 12 18:52:50 <freemangordon> will check if there are any differences between maemo and harmattan
Nov 12 18:52:53 <freemangordon> ok
Nov 12 18:53:02 <Pali> and prepare dkms package
Nov 12 19:40:12 <Pali> DocScrutinizer05: do you know if there is some way on Maemo to get country name from mcc code?
Nov 12 19:41:08 <Pali> I just found file /etc/operator_settings which contains some information for particular mcc (haha, there is mms settings), but not country name
Nov 12 19:41:44 <Pali> next another mapping file /usr/share/operator-wizard/mcc_mapping but country name is again some unreadable format
Nov 12 19:42:25 <Pali> I would like to have country name in iso3166 (= two characters)
Nov 12 19:58:38 <freemangordon> Pali: define "unreadable format"
Nov 12 19:58:49 <freemangordon> is it 3-digit code?
Nov 12 19:59:06 <Pali> freemangordon: format in /usr/share/operator-wizard/mcc_mapping or /etc/operator_settings files
Nov 12 19:59:44 <Pali> freemangordon: I want to get 2-char country name (e.g US) from 3-digit mcc code
Nov 12 20:00:42 <freemangordon> ok, will look at it later
Nov 12 20:02:24 <Pali> on wikipedia is list: https://en.wikipedia.org/wiki/Mobile_country_code
Nov 12 20:02:44 <Pali> and something similar should provide maemo api or some file...
Nov 12 20:31:31 <freemangordon> omg, yet another standard :(
Nov 12 20:32:34 <freemangordon> I was hoping on 3-digit country codes (like 100 for Bulgaria), but no...
Nov 12 20:37:57 <Maxdamantus> /usr/share/clock/wdb
Nov 12 20:38:09 <Maxdamantus> looks like you can cross-reference against that.
Nov 12 20:38:44 <Maxdamantus> dunno if it has all mcc entries.
Nov 12 20:39:16 <freemangordon> it is relatively new (in cssu at least) so it should be ok
Nov 12 20:43:41 <Maxdamantus> cat /usr/share/operator-wizard/mcc_mapping | sed 's/^[^\t]\+\t//' | while read x; do grep -q '\<'"$x"'\>' /usr/share/clock/wdb || echo "$x"; done
Nov 12 20:43:44 <Maxdamantus> qtn_clk_country_hong_kong
Nov 12 20:43:47 <Maxdamantus> qtn_clk_country_n_mariana_i
Nov 12 20:45:14 <freemangordon> Pali: https://pypi.python.org/pypi/mobile-codes
Nov 12 20:45:40 <Pali> freemangordon: list is on wikipedia, in worst case I can grab it
Nov 12 20:45:49 <Pali> I was thinking about maemo-only method
Nov 12 20:46:02 <Pali> e.g. using only sources available on maemo/n900
Nov 12 20:47:33 <Pali> Maxdamantus: I need function which get on its input 3 digit mcc code and on output I need 2char ISO / IEC 3166 alpha2 country code
Nov 12 20:47:38 <freemangordon> cat /usr/share/operator-wizard/mcc_mapping
Nov 12 20:47:41 <freemangordon> oops
Nov 12 20:47:58 <Pali> I'm going to investigate more...
Nov 12 20:48:58 <Maxdamantus> Pali: yeah, so you look up that qt thing in mcc_mapping, then look for an entry mentioning it in clock/wdb
Nov 12 20:49:06 <Maxdamantus> seems hacky though.
Nov 12 20:49:14 <freemangordon> why?
Nov 12 20:49:17 <Maxdamantus> (and apparently is missing two things)
Nov 12 20:49:20 <freemangordon> seems ok
Nov 12 20:49:47 <Pali> ah right, I miss information that clock/wdb contains also country code
Nov 12 20:49:48 <Maxdamantus> Because clock/wdb doesn't seem to intentionally list countries.
Nov 12 20:50:04 <Pali> it is 3rd column
Nov 12 20:50:08 <freemangordon> yes
Nov 12 20:50:27 <Maxdamantus> it lists cities (presumably associated with sets of timezones), which happens to include most countries.
Nov 12 20:50:29 <freemangordon> also, there are coordinates you can use to call libcityinfo
Nov 12 20:51:17 <Pali> this is for implementing wl1251-cal, regdomain from mcc to pass into kernel
Nov 12 20:51:46 <freemangordon> Pali: where are you going to get mcc from?
Nov 12 20:51:55 <Pali> dbus/csd
Nov 12 20:52:01 <freemangordon> ok
Nov 12 20:52:10 * freemangordon is afk
Nov 12 20:52:29 <Pali> if we want to use wl1251-cal also on non-maemo platform (e.g. debian/devuan) we can implement ofono
Nov 12 20:52:41 <Pali> ofono provides also mcc via dbus, but different call
Nov 12 20:56:46 <KotCzarny> its something that almost definitely there should be a lib for that..
Nov 12 20:57:14 <Pali> KotCzarny: do you know such lib/function in maemo?
Nov 12 20:58:29 <KotCzarny> https://github.com/musalbas/mcc-mnc-table/blob/master/mcc-mnc-table.csv
Nov 12 20:59:35 <KotCzarny> http://mcc-mnc.com/
Nov 12 21:00:04 <KotCzarny> i guess it would be trivial to make a package out of this and even make it auto update
Nov 12 21:17:04 <KotCzarny> (no-js version is unpaginated)
Nov 12 21:17:19 <Maxdamantus> I don't think there's anything else that maps from these mccs.
Nov 12 21:17:37 <Maxdamantus> $ echo -n 'find . -type f'; cat usr/share/operator-wizard/mcc_mapping | sed 's/\t.*$//' | while read x; do echo -n ' | xargs -d\\n fgrep -l '"$x"; done; echo
Nov 12 21:17:56 <Maxdamantus> running the command produced by that doesn't seem to find anything useful.
Nov 12 21:18:10 <Maxdamantus> I guess there could theoretically be something compressed or something.
Nov 12 21:18:55 <KotCzarny> page with html is ~23kB
Nov 12 21:19:15 <KotCzarny> so i guess whole table of strings would be ~1/3..1/2 of that
Nov 12 21:19:38 <Maxdamantus> There could of course also be something that encodes the numbers in binary.
Nov 12 21:20:09 <KotCzarny> you can even use bitfields :P
Nov 12 21:20:26 <Maxdamantus> I'm talking about looking for information already in Maemo.
Nov 12 21:20:46 <KotCzarny> information in maemo vs updated information?
Nov 12 21:25:19 <KotCzarny> in debian there is package named 'mobile-broadband-provider-info '
Nov 12 21:25:25 <KotCzarny> could it be it?
Nov 12 21:27:23 <KotCzarny> first appeared in 2010-02-27
Nov 12 21:27:33 <KotCzarny> um. 2009-07-09
Nov 12 21:29:58 <KotCzarny> yup. has country codes (two letters) and mmc/mnc (among other things)
Nov 12 21:31:42 <KotCzarny> https://tracker.debian.org/pkg/mobile-broadband-provider-info
Nov 12 22:09:51 <Pali> KotCzarny: I hardcoded Maemo data into code: https://github.com/community-ssu/wl1251-cal/blob/master/wl1251-cal.c#L524
Nov 12 22:09:55 <Pali> it is not so big...
Nov 12 22:11:01 <Pali> first time I used GNU join program :D
Nov 12 22:15:37 <Pali> hm... I cannot push new version of wl1251-cal to cssu-devel
Nov 12 22:15:43 <Pali> "The RSA host key for maemo.merlin1991.at has changed"
Nov 12 22:16:02 <Pali> merlin1991, freemangordon, RSA key was really changed?
Nov 12 22:21:50 <freemangordon> Pali: yes, merlin1991 has changed the server some time ago
Nov 12 22:21:57 <Pali> ok
Nov 12 22:22:09 <Pali> I see that we have couple of packages non updated in cssu
Nov 12 22:22:20 <Pali> cssu-devel
Nov 12 22:22:26 <Pali> but new versions are on github
Nov 12 22:22:32 <Pali> going to build them now
Nov 12 22:54:29 <Pali> new packages pushed to cssu-devel
Nov 12 22:54:43 <Pali> see tmo thread about changes
Nov 12 22:55:06 <Pali> http://talk.maemo.org/showthread.php?p=1518605#post1518605
Nov 12 23:59:44 <Pali> qt4 is still compiling...
Nov 12 23:59:51 <Pali> more then hour...
Nov 13 00:04:17 <freemangordon> Pali: on how many cores?
Nov 13 00:04:37 <Pali> dpkg-buildpackage
Nov 13 00:04:41 <Pali> no idea
Nov 13 00:04:51 <Pali> scratchbox in qemu-system 
Nov 13 00:10:10 <parazyd> qemu is very slow with compiling
Nov 13 00:10:16 <parazyd> especially when non-native
Nov 13 00:12:05 <freemangordon> gcc is native in SB afaik
Nov 13 00:12:20 <freemangordon> it is just that qt is huge
Nov 13 00:12:24 <Pali> qemu-system-x86_64
Nov 13 00:12:44 <Pali> in which is running old ubuntu with chrotted scratchbox
Nov 13 00:12:53 <Pali> in which is running arm cross compiler
Nov 13 00:13:05 <freemangordon> Pali: did you check if mce in cssu-devel matches github?
Nov 13 00:13:21 <Pali> yes
Nov 13 00:13:24 <Pali> matches
Nov 13 00:13:47 <freemangordon> ok. as I don;t remember if I pushed it
Nov 13 00:13:54 <freemangordon> seems like yes
Nov 13 00:14:40 <Pali> freemangordon: can you build new qt4 with thumb?
Nov 13 00:14:52 <Pali> my non-thumb version for cssu-devel is still compiling...
Nov 13 00:22:05 <freemangordon> Pali: tomorrow 
Nov 13 00:22:08 <Pali> ok
Nov 13 00:22:20 <freemangordon> but what to do with it after building?
Nov 13 01:16:30 <Pali> new version disable ssl2 which should have been disabled for years!
**** ENDING LOGGING AT Sun Nov 13 02:59:59 2016