**** BEGIN LOGGING AT Sun Mar 24 02:59:58 2013 Mar 24 04:08:59 sup Mar 24 05:56:17 amospalla: no Mar 24 07:50:29 I want to use mplayer on easy debian but there is no audio. how can i fix it? i can remember somewhere there was a guy about making changes in /etc/asound.conf but i actually don't remember what that was Mar 24 07:58:58 Timmy: well, the main problem is: maemo is using pulseaudio with Nokia proprietary policies (see alsaped which is oddly named since it's actually PAped aiui), PA in turn occupies the underlying ALSA soundcard exclusively. And any app using ALSA:default as audio output is in fact only using a alsa->PA compatibility plugin and thus still goes thru PA with its policies Mar 24 08:01:13 DocScrutinizer05: so you have any idea how i can fix mplayer on easy debian to play audio? Mar 24 08:01:28 changes in /etc/asound.conf *might* enable normal ALSA sound for apps like mplayer, but the downside most certainly is that it will likely fubar all PA based sound Mar 24 08:08:46 http://www.xmarks.com/site/library.maemodocs.nokia.com/fremantle/ seems to suggest we lost huge parts of documentation forever Mar 24 08:25:11 DocScrutinizer05: here it is : http://talk.maemo.org/showpost.php?p=1289207&postcount=2858 Mar 24 08:30:21 that redefines alsa:default as direct access to hw audiocard with no plugin stack at all in between. Which means a) you get no mixing, b) no reate/formate conversions c) it either will fail since PA has opened that audiocard exclusively, or you need to kill PA and thus break audio for whole rest of device which uses PA for all audio Mar 24 08:32:27 also this advice is about easydebian which is a chroot and thus has unclear but for sure no standard audio config at all Mar 24 08:33:21 DocScrutinizer05: you think it is possible to upgrade the mplayer of maemo to the latest version available? Mar 24 08:33:37 sure, why not Mar 24 08:33:55 DocScrutinizer05: that is because this version doen't display the rtl subtitles correctly Mar 24 08:34:11 it's however questionable if it will work OOTB and without massive patching for all plugins Mar 24 08:36:02 mplayer afaik is using an alternative decoder set to what mafw/gstreamer is using, and that alternative decoders need to take into account the particular hw and drivers found in maemo Mar 24 08:37:53 http://wiki.maemo.org/Documentation/Maemo_5_Developer_Guide/Architecture/Multimedia_Domain http://wiki.maemo.org/Documentation/Maemo_5_Developer_Guide/Using_Multimedia_Components/Media_Application_Framework_(MAFW) Mar 24 08:38:30 DocScrutinizer05: i don't know what's wrong with mplayer on easy debian, in the past when i entered those configurations into asound.conf, both of vlc and mplayer were working well. but after flashing the device, only vlc plays the audio Mar 24 08:41:34 particularly in diagrams in first page I linked above, see PulseAudio block and how it exposes ALSA (compatibility leagcy plugin) API upward to userspace but otoh uses ALSA device drivers in kernel (see downwards from PA block) Mar 24 08:42:54 Timmy: *where* are you changing that /etc/alsa.conf? In easybeb chroot or on host system maemo? Mar 24 08:43:22 Also notice that even the "hint" you linked to says you need to kill PA Mar 24 08:43:47 on easy deb and that was not alsa.conf, that was asound.conf Mar 24 08:43:49 which sure as hell is hot will kill audio for every normal app on maemo Mar 24 08:44:31 sending audio through the real alsa will fuck up your speakers Mar 24 08:44:38 and also this debian image is not a new one, this is what i have configured and used before Mar 24 08:44:51 kerio: that too, eventually Mar 24 08:53:22 i want to get the latest version of easy debian from http://qole.org/files/ but i don't know which one is the latest, can someone help me to choose one? Mar 24 09:02:24 DocScrutinizer05, any ETA for non working maemo.org package interface? Mar 24 09:03:32 after 5 months I'd like to push packages to extras-testing & extras... Mar 24 09:03:50 :) Mar 24 09:11:06 Pali: I resigned from coordinating and supervising maemo infra. So no, I don't know Mar 24 09:11:48 now I'm uploading ipv6 packages to extras-devel..., but still I cannot login to maemo.org/packages, so I cannot promote anything... Mar 24 09:12:03 DocScrutinizer05: what? since when? Mar 24 09:12:09 DocScrutinizer05, who is responsible for it now? Mar 24 09:12:56 nobody? Mar 24 09:13:00 HiFo? Mar 24 09:13:44 DocScrutinizer05: last you know - who had acces to those VM? Mar 24 09:13:52 *access Mar 24 09:13:56 which ones Mar 24 09:15:11 DocScrutinizer05: I guess it is garage, though not sure Mar 24 09:15:24 abd autobuilder Mar 24 09:15:27 *and Mar 24 09:18:13 well, last I know got officially established been Woody14619a:maintainer of garage, jacekowski:maintainer_ins_spe of builder. But builder_the_service_ is spread across several VM Mar 24 09:19:11 Pali:I guess we should pester them than Mar 24 09:19:17 Woody14619a: ping Mar 24 09:20:46 DocScrutinizer05, access to website components maemo.org/packages/view// Mar 24 09:20:50 and fix logging Mar 24 09:21:34 * freemangordon feels like he is back in the kindergarden Mar 24 09:22:34 * DocScrutinizer05 feels like having escaped from such kindergarden finally Mar 24 09:24:42 Pali: we either have to find the guy who is responsible and pester him or find the login and fix it. Mar 24 09:26:41 Pali: BTW it seems noone is responsible for fixing CSSU repo, too Mar 24 09:27:10 that's merlin1991:maintainer_of_repo Mar 24 09:27:44 DocScrutinizer05: last thing I know is that repo is roken, so he can't upload Mar 24 09:27:51 broken even Mar 24 09:28:13 so? Mar 24 09:28:29 who's going to fix it? Mar 24 09:28:36 DocScrutinizer05: exactly :) Mar 24 09:28:37 maintainer, no? Mar 24 09:28:48 no, as I assume maintainer has no root access Mar 24 09:29:07 so what? Mar 24 09:29:22 so he can't fix it Mar 24 09:29:32 you got a sysop for that, to request for stuff getting done that needs root access Mar 24 09:29:50 ok, so I neeed to pester merlin1991, got it :) Mar 24 09:29:56 freemangordon: i assume he's got write access to the repos Mar 24 09:30:44 kerio: afaik there is some problem with permissions. Mar 24 09:31:04 freemangordon: see? you got it instantly. Others seem not willing to get it Mar 24 09:31:10 that's why I resigned Mar 24 09:31:36 DocScrutinizer05: then keep talking with me only :P Mar 24 09:32:54 DocScrutinizer05: i guess I should check on wiki for maintainers, no? Mar 24 09:33:36 I mustn't hand out root perms to everybody on maemo.org. We got a sysop for that who also takes responsibility for keeping the system integrity Mar 24 09:33:45 sure Mar 24 09:33:56 some maintainers don't accept that fact Mar 24 09:34:11 my point is - in order to find that sysop, i have to check on wiki, ain;t? Mar 24 09:34:21 sysop is warfare Mar 24 09:34:28 aah, ok Mar 24 09:35:12 but you shouldn't pester sysop about maintainer stuff. For example warfare has no idea about repos or autobuilder, basically Mar 24 09:35:23 got it Mar 24 09:35:36 I am just trying to figure out who is who Mar 24 09:35:37 it needs a request from maintainer who should know what needs to get done, and sysop will do it for them Mar 24 09:35:46 :nod: Mar 24 09:35:47 DocScrutinizer05: so it should be "i need `$command` to be run as root on $machine for $reason"? Mar 24 09:35:59 kerio: maybe not that formal Mar 24 09:36:01 yes, exactly Mar 24 09:36:08 it's hard to be unformal when you have to expand variables Mar 24 09:38:06 jacekowski and woody bitched at me about "do you honestly suggest I shall put up such a request for everything I need to get done, and then wayit 30min or even an hour for it to get processed?" then ragequit Mar 24 09:38:50 well, if they have such an amount of requests maybe they do need more permissions Mar 24 09:38:55 but not *all* permissions! Mar 24 09:39:32 kerio: I thouhgt same, and tried to use a policy based on cooperation and common sense. This didn't fly Mar 24 09:39:39 DocScrutinizer05: I don;t know what that conversation was, but I could imagine. Having a formal request for every nit is a nightmare and contraproductive, I agree Mar 24 09:40:33 but I guess a temp rights elevation is always possible if needed Mar 24 09:42:38 freemangordon: no, it's not. It just requires more discipline and higher competence form maintainers. Like: instead of requesting "please add +x to /local/bin/xyz" maintainer needs to ask for "please add my account to group+we of /local/bin" Mar 24 09:42:57 kerio: lol Mar 24 09:43:02 (no it's not been about nightmare) Mar 24 09:43:04 ? Mar 24 09:43:22 "i need `$command` to be run as root on $machine for $reason"? Mar 24 09:44:04 Macer: what? Mar 24 09:44:30 jut thought that was a bit funny :) Mar 24 09:44:36 Macer: you obviously never worked in a corporate environment Mar 24 09:44:54 ok? Mar 24 09:45:19 it wasn't what he said. it was how he worded it. Mar 24 09:45:25 usually you don't even have admin permissions of your very own local PC there Mar 24 09:45:55 DocScrutinizer05: and I guess FOSS community is a differnt beast Mar 24 09:45:56 natural languages would be better with variable substitution Mar 24 09:46:09 kerio: that it would ;) Mar 24 09:46:25 now in a corporate environment you have very few problems with accountability. For a FOSS projecr there basically is no accountability Mar 24 09:46:32 kerio: lead with "for" heh Mar 24 09:47:06 so the less you can allow everybody to have root permissions on everything Mar 24 09:47:31 DocScrutinizer05: is there a reason so many would require root to begin with? Mar 24 09:47:48 i figure most things would be able to be grouped Mar 24 09:48:05 I thought you could when you educate those who got those permissions about what they may and what they MUST NOT do with them. But that been an epic fail and I'm responsible for that failure Mar 24 09:48:43 heh. oh nooooo. you're playing with fire giving more than a handful of people that sort of control over the system heh Mar 24 09:48:59 especially if one were to go rougue on you Mar 24 09:49:02 heh it been ~4 people heh Mar 24 09:49:22 and they had no "control over the system" Mar 24 09:49:46 almost makes you wish you could make like 5 clones of yourself huh? Mar 24 09:49:49 they had just locally elevated permissions Mar 24 09:50:02 Macer: he already did, notice the name Mar 24 09:50:03 kerio, do you know if there is in maemo extras some package which enabling sudo for user user? Mar 24 09:50:07 lol Mar 24 09:50:20 Pali: there's sudser, but it works awfully Mar 24 09:50:29 what it doing? Mar 24 09:50:44 setting up a password and adding passwordful or passwordless sudo Mar 24 09:50:47 i miss lcuk :-/ used to be comedy hour every time he was here Mar 24 09:50:53 Pali: edit /etc/sudoers.d/* Mar 24 09:51:07 DocScrutinizer05, I done it 2 years ago :D Mar 24 09:51:08 add your own plicy, run update-sudoers Mar 24 09:51:18 but I'm asking if there is package which doing it Mar 24 09:51:33 there are several, one worse than the other Mar 24 09:51:39 now trying to fix sudo/root problem in maemo Mar 24 09:51:41 iirc sudser left my sudoers in an inconsistent state Mar 24 09:51:50 which broke... everything, really Mar 24 09:52:01 sudser is BS Mar 24 09:52:04 afaik Mar 24 09:52:08 Pali: ssh root@localhost Mar 24 09:52:27 kerio, that working too Mar 24 09:52:29 Pali: sudo/root problem? which problem? Mar 24 09:52:43 DocScrutinizer05, sudo gainroot Mar 24 09:52:49 nobody ever reported about a problem Mar 24 09:53:14 we talked about it more times Mar 24 09:53:43 yes, and you always refused to deliver any sound rationale why it needs any fixing Mar 24 09:54:37 and particularly how that fixing could be done in a non-intrusive non-breaking-compatibility way Mar 24 09:54:37 sudo gainroot works as intended Mar 24 09:54:42 if rd mode is enabled, give a root shell Mar 24 09:55:31 but if you have rootsh installed without rd mode it enable it too Mar 24 09:55:43 ah yeah, rootsh is horrendous Mar 24 09:55:43 no, it doesn't Mar 24 09:56:00 err, sorry, missed the rootsh part Mar 24 09:56:20 whatever rootsh does, you CAN NOT fix it since it already exists Mar 24 09:56:35 ah indeed Mar 24 09:56:46 I will look into packages which using rootsh Mar 24 09:56:57 SO WHAT? Mar 24 09:57:26 and fix using stupid rootsh calls Mar 24 09:57:31 you can't fix a privilege escalation done in XY0.1 by publishing XY0.2 Mar 24 09:57:47 then we can add password for user and root Mar 24 09:57:58 unless you forbid downgrading like in HARM Mar 24 09:58:30 >implying local privilege escalation is even remotely (sic!) fixable in fremantle Mar 24 09:58:37 DocScrutinizer05, look why package XYZ need calling gainroot, fix to use directly sudo and add sudoers.d config file Mar 24 09:58:59 sure, sane thing Mar 24 09:59:14 then it is possible to ask password before enabling root shell (via gainroot or other script) Mar 24 09:59:22 will not make rootsh-v${current} vanish Mar 24 10:00:16 you can publish a new package "sane-rootsh" that conflicts with rootsh even Mar 24 10:00:32 but messing around with original rootsh pkg won't help anything Mar 24 10:02:56 even if you did, there are still a dozen other fsckdup pkgs like sudser and easyroot or whatsthename that do even worse shit than rootsh pkg does Mar 24 10:12:34 so honestly, I thin best approach is to create a new project/package "secure sane sudo (formerly know as rootsh)" and make it conflict in dependencies with rootsh and and all other known-borked root access packages Mar 24 10:12:41 think* Mar 24 10:13:12 users are free to install it, and see what it gives them Mar 24 10:13:48 they will find all the broken apps that depend on rootsh getting removed as conflicting dependencies. Mar 24 10:14:08 unless those apps have been updated to not depend on rootsh anymore Mar 24 10:14:35 clean easy-to-ggrok straight approach Mar 24 10:15:34 I'd be willing to support it any way me/council/techstaff can (and may) do Mar 24 10:15:45 like helping with promotion Mar 24 10:16:40 grep across whole repo for particular keystrings like "sudo" in and program text Mar 24 10:16:56 any* Mar 24 10:20:18 <_rd> Is there a good way to track which fremantle packages see updates in extras? Mar 24 10:20:40 define "see" Mar 24 10:22:51 <_rd> Well...on my n900 I just see updates on cssu, modrana, and navit. Navit I do myself. Mar 24 10:23:34 <_rd> I am wondering, if there are any ohter interesting areas work goes in, which I simply do not see...because I have not installed the apps. Mar 24 10:23:58 <_rd> Something like a log of uploaded packages would be nice Mar 24 10:24:47 <_rd> I could download Packages files and diff them every few days, but there might be a more direct way Mar 24 10:25:01 so you're asking for a list of maemo(-devel?) packages with mtime newer err... 1 week? Mar 24 10:26:40 _rd: you don't need to download them every few days. Since HAM is doing that for you, only when there are changes, usually once a day, per default Mar 24 10:28:03 so all you'd need to do is safe away old packages.gz and if there appears a new version on your device. you do a diff to old version you saved and then save new version to the backup location Mar 24 10:28:17 s/safe/save/ Mar 24 10:28:17 DocScrutinizer05 meant: so all you'd need to do is save away old packages.gz and if there appears a new version on your device. you do a diff to old version you saved and then save new version to the backup location Mar 24 10:28:49 <_rd> yes, but HAM only tells me on updated packages which changed. Mar 24 10:28:57 <_rd> sorry.... Mar 24 10:28:58 so what? Mar 24 10:29:18 <_rd> HAM only tells me on installed packages which updated Mar 24 10:29:22 did I say you should ask HAM about new packages, or did I suggest to do a diff? Mar 24 10:29:49 <_rd> I would like to see a list of all updates, also for packages which I have not (yet?) installed. Mar 24 10:30:01 <_rd> ....aah. Mar 24 10:30:04 _rd: http://maemo.org/packages/repository/latest/fremantle_extras-devel_free_armel/all/ Mar 24 10:30:07 so all you'd need to do is save away old packages.gz and if there appears a new version on your device. you do a diff to old version you saved and then save new version to the backup location Mar 24 10:30:21 <_rd> you say I could use the Package file, which is already on the n900. Mar 24 10:30:23 <_rd> understood. Mar 24 10:30:39 <_rd> If I do that from a PC, downloading directly might be more convenient. Mar 24 10:30:49 yeah Mar 24 10:30:54 obviously Mar 24 10:31:31 <_rd> kerio: many thanks, that is what I was looking for. Mar 24 10:31:37 <_rd> ....actually it is much better :-) Mar 24 10:32:21 doesn't have packages that are not in "user" probably Mar 24 10:33:15 also seems to not show any date when those "latest" actually got updated Mar 24 10:34:09 do you know of any doc or text to guide the installation of rescueos ? Mar 24 10:34:26 amospalla: short answer: don't do it Mar 24 10:34:41 it's unsupported, and not meant to work Mar 24 10:35:15 NIN101: have you considered putting something that can charge from a usb host instead of the wallcharger? Mar 24 10:36:17 kerio: oh, it was on rescueos url, am I supposed to use flasher everytime I use it, so it's not meant to be copied in any way into n900 Mar 24 10:36:45 that's the gist of it, yes Mar 24 10:37:03 uboot *can* boot it on-device, and it works fine Mar 24 10:37:16 but configuring uboot isn't that easy Mar 24 10:37:37 somebody who has time should write wiki page about uboot bootmenu :-) Mar 24 10:40:11 kerio: that's why I was asking, about installation on the device itself, but I prefer not to deal with boot, I do not want to break anything Mar 24 10:40:13 thanks Mar 24 10:40:33 it's ok if I can boot it with cable Mar 24 10:40:47 in fact, it is a great idea make it that way Mar 24 10:40:55 ~rescueOS Mar 24 10:40:55 from memory, rescue-os is http://206.253.166.96/N900/rescueOS/ Mar 24 10:42:10 honestly rescueOS is a poor choice for any OS to get actually installed on eMMC or uSD to boot it per uBoot from there. It's simply not meant to get used this way Mar 24 10:42:41 * kerio found it damn useful more than once Mar 24 10:42:54 yes, they were all situations where i was literally sitting at a computer Mar 24 10:42:57 there are quite a number of better choices Mar 24 10:42:59 like most of the time in my life Mar 24 10:43:23 but plugging in a cable and typing a command is still more hassle than selecting a menu entry :) Mar 24 10:44:38 kerio: yeah, considered, not done yet... Mar 24 10:44:59 ideal choice was a auxiliary maemo installed to a separate partition, with support to mount the main OS partitions to /original-maemo/* Mar 24 10:45:01 DocScrutinizer05: are there "official" charge scripts? Mar 24 10:45:12 it is not something I'm usually using, it's ok dealing with cable when I have to use that Mar 24 10:45:38 DocScrutinizer05: that's not ideal at all, having to load h-d and all the other crap when you just need a terminal to run a command consumes time Mar 24 10:45:40 bootmenu did the job sometimes, but it is not the same Mar 24 10:45:43 kerio: well, on maemo.cloud-7.de and ShadowJK's server it is as official as it gets Mar 24 10:45:46 and yes, rescueOS is not meant to be preinstalled somewhere, only booted when needed. Mar 24 10:46:38 with bootmenu I have some kind of rescue shell, but have to do it through cable network and ssh, because using the device itself does not work correctly Mar 24 10:46:54 and I doubt it does charge Mar 24 10:48:23 http://maemo.cloud-7.de/maemo5/patches_n_tools/charge21_500mA-USBlimit__UNTESTED.sh should not conflict with server USBhost charging Mar 24 10:48:30 for the limit to 500 Mar 24 10:49:16 NIN101: you could also try the bq2415x-charger and bq27x00-battery modules Mar 24 10:49:26 for automagical everything Mar 24 10:49:44 DocScrutinizer05, correct me if I'm wrong, but isnt max current for usb2.0 only 100mA until enumeration is done? Mar 24 10:49:59 Pali: per spec yes Mar 24 10:50:33 but those PCs that enforce that also fail to power any fan or mugwarmer or kbd-light Mar 24 10:50:46 yeah I pondered about including the "UNTESTED" 500ma script, but I can't risk to test anything of this kind as I have only one N900 and I am not very experienced at this topic in general. Mar 24 10:50:49 well, maybe kbd-light doesn't need >100 Mar 24 10:51:19 1) make initrd with kernel-power modules Mar 24 10:51:25 NIN101: it's kinda tested meanwhile. And for sure it's not dangerous in any kind Mar 24 10:51:29 2) modprobe bq2415x-charger Mar 24 10:51:41 kerio: bwahahaha Mar 24 10:52:32 it can't be too hard to port the bq24k module to 2.6.37 Mar 24 10:54:29 Pali: what happens to the bq24k module if there's no usb gadget module to do the detection? Mar 24 10:54:43 does it do nothing at all? or does it draw the 100mA? Mar 24 10:54:53 <_rd> which browsers do you use on the n900? Built-in microb? Mar 24 10:55:01 kerio: wtf would anybody want to do that, for rescueOS? Mar 24 10:55:24 kerio, it will be in none/off state Mar 24 10:55:58 in maemo 2.6.28 kernel is glue between musb and bq27x00_battery which is stored in board-rx51-peripherals.c file Mar 24 10:56:42 s/bq27x00_battery/bq2415x_charger/ Mar 24 10:56:43 Pali meant: in maemo 2.6.28 kernel is glue between musb and bq2415x_charger which is stored in board-rx51-peripherals.c file Mar 24 10:57:07 in upstream kernel is (or will be) direct calls between isp1704_charger (which doing charger detection) and bq2415x_charger Mar 24 11:01:45 <_rd> which browsers do you use on the n900? Built-in microb? Mar 24 11:06:09 _rd: select one they all have there pros and cons, I use Opera 11 many use MicroB with Flash tweak. Mar 24 11:11:04 <_rd> you think of tweakflashver as "Flash tweak" ? Mar 24 11:12:18 Pali: those direct calls will fuckup hostmode and most likely even USB at large Mar 24 11:12:24 yep, hopefully alopex will rule the browser space soon... Mar 24 11:13:30 Pali: since, to detect charger, 170x needs to reconfigure the datalines, and that will brwak adat transmissions Mar 24 11:13:35 break* Mar 24 11:13:46 break data* even Mar 24 11:14:12 DocScrutinizer05, I have patches for musb and that 1704 driver to (semi)-fix that in upstream Mar 24 11:14:25 but there is bigger problem: in upstream host mode not working Mar 24 11:14:57 there is code for force host mode, but it cause instant kernel panic Mar 24 11:15:37 well, you could argue a botch liike N900 hostmode has no place in upstream anyway Mar 24 11:15:55 so I rather fix charger detection which do not break peripheral usb mode for upstream kernel Mar 24 11:16:02 for sure it's colliding with any sane USB code in upstream Mar 24 11:16:28 hmm, good luck with that Mar 24 11:16:32 it can't get done Mar 24 11:16:45 DocScrutinizer05, our hostmode code in n900 kernel is adding code for choosing speed via testmode musb param (or how it is called) Mar 24 11:16:59 access to that register is in upstream kernel via debugfs Mar 24 11:17:17 but cat /sys/debug/.../that_file cause kernel panic on n900 Mar 24 11:17:31 *shrug* Mar 24 11:17:50 but needed code for force host mode for n900 is in upstream Mar 24 11:18:01 I'm honestly deprecating to try and upstream H-E-N hostmode kernel patches Mar 24 11:18:08 api is different, not using /proc, but code is here Mar 24 11:18:26 it's absolutely nonsensical Mar 24 11:19:57 and you *can not* fix the conflict between charger detection and data-transmission in kernel space, since 1707 has no idea if there is a data transmission happening resp a session established Mar 24 11:20:56 so, to "fix2 that, you'd need to add metric tons of useless cruft to kernel drivers, to make assumptions about state of USB Mar 24 11:22:04 I explained all that during hostmode development, and then again several times after that to KP devels Mar 24 11:28:09 paul fertser (coder of original final version of hostmode kernel) discussed that with me when we developed those patches. He's a die-hard FOSS and upstream and debian-only guy, and he agreed that N900 hostmode *must not* go to upstream, since it's a terrible botch that can't get done right and also it's useless in upstream since there never agin will be a similar hardware that could re-use that botch Mar 24 11:29:25 so bringing it upstream is a PITA and a bloat at best, but most probably can't get done at all, in a compliant way Mar 24 11:30:33 or let me put it this way: IF it would've been possible and reasonable to bring H-E-N kernel upstream, it would've been there since very beginning Mar 24 11:31:30 huh, after 4 months without upgrade, apt-get upgrade wants to upgrade 47 packages (some are hildon*, pulseaudio*), is this expected? Mar 24 11:31:32 and Paul is a rather gifted and experienced kernel developer who knows his shit Mar 24 11:32:00 amospalla: maybe not Mar 24 11:32:12 I'm using cssu thumb Mar 24 11:32:36 so, until there does not appear a new cssu update, I should not touch that Mar 24 11:32:56 guys are messing with repo atm, we seen some hashsum errors sneaking in again, and most likely it's best to _not_ update for a while Mar 24 11:33:24 ok, understood, thanks DocScrutinizer05 Mar 24 11:33:36 meh, cssu-thumb, that's again a completely different story Mar 24 11:33:43 ask freemangordon Mar 24 11:35:41 DocScrutinizer05, charger detection is called when charger is connected, isp driver enabling musb after detection is complete Mar 24 11:36:13 yes, but then again you get layering violations that you can't upstream Mar 24 11:36:23 this is already in upstream Mar 24 11:36:46 yeah, and it's not working, according to what you said Mar 24 11:37:04 not working is other part Mar 24 11:37:25 because musb driver started using omap mailbox api instead some old Mar 24 11:37:34 which still using isp driver Mar 24 11:37:52 but musb or isp can be upgraded to use new "notify" api Mar 24 11:39:36 I created patch which add optional hotplug/removable file_storeage gadget to g_nokia.ko Mar 24 11:39:57 so you do not want to switch and load/unload g_file_storage and g_nokia Mar 24 11:40:17 you can export what you want (only file_storage or only usb network or both) Mar 24 11:40:29 but usb maintainer rejected my patch Mar 24 11:40:48 because only *nokia* can device what to do with their g_nokia.ko driver in upstream Mar 24 11:41:46 see? Mar 24 11:41:49 and he did not understand that nokia does not care about their linux driver in upstream anymore and also they are not working on n900 anymore Mar 24 11:42:21 so I do not know if something can be patched for nokia n900 support... Mar 24 11:48:11 Pali: I honestly don't feel like going thru all that again. I explained we discussed *all* that during _several_ weeks when developing H-E-N hostmode kernel patches. And the guy who been most die-hard radical upsteam/debian-only and almost refused to touch maemo at all _because_ it wasn't completely upstream compliant, this guy concluded that his own patches can't go upstream and there is not even any rationale why they should. We found Mar 24 11:48:12 layering problems of all sorts, and there were other more general argumants as well that suggested hostmode botch MUST NOT go upstream and CAN NOT go upstream, since on one edge you always run into "doesn't work with standard kernels" or "doesn't make sense at all" Mar 24 11:48:42 I'm honestly not going to discuss all this again with you, now that you think you can do better than we did 3 or 4 years ago Mar 24 11:49:37 DocScrutinizer05, upstreaming our host mode patches from 2.6.28 kernel are not needed Mar 24 11:49:52 I think that all needed code is already in upstream kernel Mar 24 11:50:08 think what you like Mar 24 13:01:36 196M free on MyDocs :o Mar 24 13:34:38 hi all Mar 24 13:34:52 someone know what is the username of jacekowski on TMO ? Mar 24 13:35:43 I have to put credits to him for the help he provide during my last package build on autobuilder .... Mar 24 13:36:16 is there any program/application for linux to manage the n900? something similar to the nokia pc suite. the only thing i want to do is to send and receive sms through my computer Mar 24 13:36:56 Timmy: there is some command line utility, to do that Mar 24 13:37:15 like what? Mar 24 13:38:25 first result on google "n900 sms send": http://talk.maemo.org/showthread.php?t=69661 Mar 24 13:40:44 you could try gammu, no idea if it works with N900 but have used it with few other Nokias and huawei dongle successfylly Mar 24 13:45:05 Pali: I talked with Paul and it seems he's still interested in upstreaming hostmode patches. He might give a helping hand, now that I convinced him that *kernel* on N900 is working properly, even the upstream one Mar 24 13:45:56 DocScrutinizer05, nice to hear :-) Mar 24 13:46:39 DocScrutinizer05, here is 3.8 kernel with n900 patches: https://gitorious.org/linux-n900/linux-n900 Mar 24 13:46:49 he also denied that there were any insurmountable obstacles with it Mar 24 13:46:50 some of them will be included in 3.9/3.10 Mar 24 13:47:23 so probably I told BS, sorry Mar 24 13:47:32 Pali: btw I rebased your patches to 3.9-rc1, boots :) Mar 24 13:47:54 Skry, I started cleaning my patches Mar 24 13:48:05 nice Mar 24 13:48:15 I will push new version of cleaned patches when I finish it Mar 24 13:48:30 ok, could you ping me when that happens? Mar 24 13:49:29 Pali: did you have working audio? I get insta-crash with both 3.8 and 3.9. I was thinking if our configs are different and would that be the reason-- Mar 24 13:49:56 audio not working... Mar 24 13:50:15 maemo not crashing, but no sound Mar 24 13:50:36 yeah, too bad. that's currently the only thing keeping me from updating arch packaging to 3.9 :\ Mar 24 16:24:55 hm Mar 24 16:25:06 i'm a bit confused as to how phones advertise their freqs nowadays Mar 24 16:25:31 quad band GSM (850MHz, 900MHz, 1800MHz, 1900MHz) Mar 24 16:26:16 is t-mobile US now 1800? Mar 24 17:01:07 1800 is used in europe Mar 24 17:01:39 basically, 900 became full in cities, so 1800 was added Mar 24 17:03:16 tmobile is 850/1900 or something, simultaneously Mar 24 17:03:26 ie one is used for transmit and the other for receive Mar 24 17:04:10 Macer: https://en.wikipedia.org/wiki/GSM_frequency_bands Mar 24 17:04:23 i looked at that Mar 24 17:04:30 and it makes 0 sense as well ;) Mar 24 17:04:38 freqs are listed twice Mar 24 17:05:12 what do you mean? Mar 24 17:05:16 like take 1900MHz.. according toe wikipedia it is an edge freq OR a UMTS/HSPA/HSPA+ freq Mar 24 17:05:20 so does that depend on the device? Mar 24 17:05:30 plus doesn't it use 1700 for uplink and 1900 for downlink? Mar 24 17:05:38 that is where i get confused Mar 24 17:06:00 http://www.amazon.com/G13-Black-Quad-band-Watch-UNLOCKED/dp/B0064U47HC/ref=pd_sim_cps_2 Mar 24 17:06:04 i was looking at that thing Mar 24 17:06:07 looks kind of neat Mar 24 17:07:12 Unlocked, quad-band GSM 850/900/1800/1900 MHz Mar 24 17:07:24 and what uses that 1800MHz freq? is that japan or something? Mar 24 17:08:28 "the quad-band phone, also known as a world phone, supporting all four major GSM bands, allowing for global use" except non-gsm like japan Mar 24 17:08:34 In europe UMTS and GSM/edge are both on 900MHz, though obviously the operators themselves have allocated part of their 900MHz holding for 3g and some for 2g Mar 24 17:09:06 1800 remains 2g only, but may be opened up to have both 2g and lte Mar 24 17:09:23 and 2100 is umts/lte shared, I think Mar 24 17:10:18 sixwheeledbeast: so that means regardless of carrier it should support GSM 3G? Mar 24 17:10:36 or does it just mean you can talk on it but 3G or faster maybe not? Mar 24 17:11:43 that device has no 3g. Mar 24 17:13:13 Macer: depends on the country Mar 24 17:13:58 No, I'm pretty sure as the device doesn't have 3g/umts radio, 3g will not work anywhere, regardless of country :) Mar 24 17:14:14 ShadowJK: but the same goes for actual phones which do have the 3G modems Mar 24 17:14:32 the listing for quad bands is usually the same with the oddball 1700MHz being here and there Mar 24 17:14:53 sorry no not that device as ShadowJK i mean 3G devices Mar 24 17:15:15 1700 is japan AFAIK Mar 24 17:15:59 Unlocked quad-band GSM cell phone compatible with 850/900/1800/1900 frequencies and US/International 3G compatibility via 850/900/1700/1900/2100 UMTS/HSDPA plus GPRS/EDGE capabilities Mar 24 17:16:21 i'm pretty sure tmobile us uses 1700 for hspa+ Mar 24 17:16:27 that's an e7 Mar 24 17:16:58 https://en.wikipedia.org/wiki/UMTS_frequency_bands Mar 24 17:17:35 ooh pentaband umts Mar 24 17:17:57 quad band will normally be 850/900/1800/1900 to cover most of the world. Mar 24 17:18:50 Band II (W-CDMA 1900) in North America and South America (ITU Region 2). Mar 24 17:18:55 Band IV (W-CDMA 1700 or Advanced Wireless Services) in the United States (T-Mobile USA), Canada (WIND Mobile, Mobilicity, Vidéotron) and Chile (VTR Movil, Nextel Chile) Mar 24 17:19:08 * Macer points at 1700 and t-mobile usa Mar 24 17:19:24 i believe tmob calls that the AWS band Mar 24 17:20:37 they seriously need to make a simpler way of doing this :) Mar 24 17:20:57 ShadowJK: yeah. e7 supports everything heh Mar 24 17:21:03 i honestly just got it for the qwerty tho Mar 24 17:22:21 http://maemo.org/community/brainstorm/view/850-1900_mhz_frequency_support/ Mar 24 17:22:24 oldie but goodie Mar 24 17:26:34 my point being that the n900 doesn't support the 1900MHz freq? Mar 24 17:27:23 http://www.nokia.com/gb-en/products/phone/n900/specifications/ Mar 24 17:27:39 yet in its listed freq bands.... it has it listed.. it just doesn't support it for data :) Mar 24 17:28:12 am i the only one not grasping how overly complicated these bands are ? heh Mar 24 17:29:40 and... marina harisson is kind of cute Mar 24 17:30:56 N900 = quad band for GSM and three band for UMTS. No 1900 for data, from what I gather. Mar 24 17:34:36 Well data on gsm on 1900 Mar 24 17:35:26 yes, sorry 3G data. Mar 24 17:38:25 And of course, support for 2g on 1900 is useless for talking to a 3g tower on 1900 :) Mar 24 17:49:13 there's no specific "data" capability for any band Mar 24 17:50:11 either the modem can use a band, or it can't. If the data transferred is GSM-codecvoice data or http:// browser data is relatively irrelevant for the modem Mar 24 17:51:36 and yes, the problem with N900 in USA been that it doesn't support 3G/UMTS on a band that's originally meant to be 2G/GSM Mar 24 18:24:31 what is the quickest/easiest way to copy-paste a long filepath in Maemo? Mar 24 18:25:57 copy and paste in what? Terminal? Mar 24 18:26:27 yes, into terminal Mar 24 18:26:49 e.g. I have a .deb that I want to dpkg Mar 24 18:27:02 and its path+filename are really long Mar 24 18:27:29 how would you copy-paste that path into terminal? Mar 24 18:27:57 copy it onto the clipboard then paste using the top menu in XTerm Mar 24 18:28:32 for instance, in uBuntu you can copy the file in file manager, and then if you paste in terminal, the filepath is pasted Mar 24 18:28:38 same procedure here? Mar 24 18:28:55 I tried that before, didn't work Mar 24 18:29:00 nothing happened Mar 24 18:29:19 text or whole file? Mar 24 18:29:41 do you copy it from Maemo stock file manger, or another one? Mar 24 18:29:49 whole file Mar 24 18:30:16 I mean if you copy just the file name, you still have to type the path first Mar 24 18:31:38 and a different issue, apt-get -f install suggests to remove kernel-power-flasher, I wonder if I should do that? won't this touch the kernel itself? Mar 24 18:32:38 I use hamsterfiler to copy full locations, if that's what you mean. Mar 24 18:34:35 ok. how exactly in hamsterfiler? in Maemo stock fm, if you try to copy sth., you immediately have to specify where to copy it to. If you don't, I think it's not kept on the clipboard Mar 24 18:35:36 I'd have to install Hamsterfiler. need to resolve this apt-get issue though bfore I can install anything Mar 24 18:43:00