**** BEGIN LOGGING AT Mon May 27 02:59:58 2013 May 27 06:27:41 A wonderful splendid Boomtime to You all. May 27 06:39:06 Hi guys, how can i tell HAM, not to check for updates, everytime i connect any GSM/Wi-fi network ? May 27 06:39:22 I need it to check it manually , only when i press "check for updates" May 27 09:02:21 there's a wiki page "maemo customization" OWTTE May 27 09:08:45 hello #maemo May 27 09:09:10 I just flashed a "new" n900, installed cssu-t and thumb, and I don't see the extras repo anywhere in HAM, does that sound familiar? May 27 09:15:59 DocScrutinizer05: gconftool -s --type int /apps/hildon/update-notifier/check_interval NEW_VALUE ? May 27 09:16:40 oh, right, that's important May 27 09:25:58 XATRIX: yes May 27 09:28:57 DocScrutinizer05: is it ok ? May 27 09:29:01 yes May 27 09:29:15 don't go to extremes with NEW_VALUE May 27 09:29:18 May 27 09:29:18 May 27 09:30:45 something equivalent to one year should be more than sufficient May 27 09:31:59 what about 'blink-after' ? May 27 09:32:06 Should i set it equal to ? May 27 09:35:28 nah, that's completely irrelevant May 27 09:35:58 afaik it says when blinking shall stop and go to steady signal May 27 09:36:25 you want blinking never to start May 27 09:37:54 probably setting blink-after set to 20 would save about 20mAh per year, if update checks weren't diabled anyway May 27 09:38:12 make an icon flash/blink costs some CPU cycles May 27 09:38:44 I guess that's why Nokia decided to switch to steady after 1440 seconds (minutes?) May 27 09:38:56 for sure seconds May 27 09:42:25 it also consumes brain cycles seeing it blink May 27 09:58:02 hehe right May 27 09:58:56 in the tmo thread somebody claims there was a way to stop update checks completely in HAM red pill mode settings menu - I can't find such option May 27 09:59:20 maybe a good feature request for CSSU HAM? May 27 09:59:26 uhh. Where did "" May 27 09:59:32 .."Mac OS" directory come from? May 27 09:59:42 been always there May 27 09:59:58 really? I don't recall this on my previous device May 27 10:00:20 and before you ask: yes you can delete it but it's not really recommended since it's tiny and might eventually be useful May 27 10:00:27 sure May 27 10:03:12 """"'"'"'' " '"' ''"''' ' ' ' " ' May 27 10:03:19 I think when i press Reload button on system menu, it simply resets the device, like if i press ALT+SysRq May 27 10:03:24 err sorry May 27 10:03:31 can i configure it to do "#reboot" ? May 27 10:03:56 reload button? May 27 10:04:10 Yea May 27 10:04:39 I enabled it in systemui May 27 10:04:43 * DocScrutinizer05 tries to recall any reload button anywhere May 27 10:07:07 Arrgh... it's not enabled by default May 27 10:07:21 Press your "Power button" you'll see Shutdown there May 27 10:07:46 If you enable Restart/Reload button inside the systemui.xml May 27 10:07:53 You'll be able to use it May 27 10:28:43 (re update) useful: alarmclient -L May 27 11:14:00 hm. Is changing the ordering of icons in app menu supposed to Just Work? May 27 11:14:22 here I was able to reorder things not the way I wanted, and now it's stuck that way. Maybe it's because of catorise :/ May 27 11:16:00 DocScrutinizer05: I now got AutoDisconnect properly uninstalled... or rather May 27 11:16:06 I see no issues even after reboot May 27 11:16:18 I did make sure to first set AD to be "never" active, and reconnected WLAN, before uninstalling it May 27 11:16:51 also, I used dpkg --purge to deinstall, not dpkg -r, if that matters May 27 11:17:35 (the steps before uninstalling served to make as many as AD's potential changes undone) May 27 11:18:02 as for people whose system got screwed up by AD, wouldn't reinstalling AD have helped? May 27 11:18:15 tadzik: moving icons in app menu with apmefo installed just breaks about everything, I guess for catorise it's the same. apmefo at least has a generic way to move icons in (sub)menus May 27 11:19:34 divVerent: unclear May 27 11:19:52 I never touched it (for obvious reasons) so never investigated May 27 11:20:19 divVerent: many thanks for the update on how to cleanly uninstall May 27 11:20:34 well, it was merely common sense to uninstall this way... worked for me, but no guarantee for others May 27 11:20:44 I did it with having a backup of / made before, though May 27 11:20:45 :nod: May 27 11:21:11 divVerent: you checked your extras access? May 27 11:21:13 now designing my autodisconnect logic I actually want... basically "no GPRS when locked" will do for me May 27 11:21:20 no, still have no SDK running anywhere May 27 11:21:31 if only I could create a package any other way... May 27 11:21:46 actually... is it very evil to try to upload packages "blindly"? May 27 11:21:56 i.e. take original debian sources, modify without test building, then uploading that? May 27 11:22:07 and then "seeing if the build server succeeds" May 27 11:23:21 I think that's exactly what this pypackager (or what's the name) is meant for May 27 11:23:55 DocScrutinizer05: apmefo? May 27 11:26:14 the better catorize May 27 11:26:49 UI a tad clumsy, but 'just works' May 27 11:29:01 I tested catorize, thought "SCREW DAT! WTF is calculator under system?" or sth like that, cursed it for not allowing to sort stuff your own way, and thankfully dumped it for apmefo when that came out May 27 11:29:11 I see. Catorise plus just worked for me usually, but now I tried to reorder stuff and shit hit the fan May 27 11:29:51 I tried looking at xmls in /etc/xdg-something, but they either look fine or have no effect on the ordering whatsoever May 27 11:29:58 dragging icons in appstarter is a HD thing that doesn't know of the tweaks done by any such stuff like catorize/apmefo May 27 11:30:35 so dragging icons is forbidden when you have submenus May 27 11:31:19 either your catorize/apmefo/whatever offers a generic method to do icon sorting, or you're screwed May 27 11:31:21 DocScrutinizer05: pypackager is somewhat... odd to use for binaries though :) May 27 11:31:33 apparently. Let's see if reinstalling catorise changes anything May 27 11:33:42 for apmefo it helps to edit and safe the config May 27 11:34:14 dunno if cacorise even has such thing like "edit" May 27 11:35:01 catoriseplus doea May 27 11:35:02 apmefo rewrites all the xml HD files according to own internal config, when you edit&save it May 27 11:35:46 I'll try it May 27 11:36:48 http://maemo.org/packages/view/apmefo/ May 27 11:40:22 it seems that I have to arrange everything manually with that, no? May 27 11:42:31 DocScrutinizer05: is there a way to find out which packages the build server has? e.g. if "po4a" exists May 27 11:42:41 it is a build dependency of deborphan, but I could patch around if it needed May 27 11:43:06 errr May 27 11:43:25 that's what SDK is meant for May 27 11:44:19 well, SDK is meant to provide that (among other purposes of SDK) May 27 11:44:36 yes, but I would have to wait one or two weeks till I can use the SDK again May 27 11:44:44 it sucks to be constrained by a Celeron 630 and 4G HDD May 27 11:45:11 sure it does. scratchbox alone is 5GB iirc May 27 11:45:33 that's why I wanted to ask how much damage I do if I upload a package of which I don't know if it will build properly or not May 27 11:45:46 if it's ok to do this, I'd just try it and see what happens May 27 11:45:50 generally buold should just fail May 27 11:46:03 and would I get an error message or just a generic "build failed" May 27 11:46:05 throw error and send a mail to you May 27 11:46:23 po4a is used for i18n stuff in deborphan, I could patch i18n out if needed May 27 11:46:23 on web you can access the logs May 27 11:47:01 I never really used autobuilder, so no expert here May 27 11:48:07 hehe, patching out the po4a use is trivial :) May 27 11:48:12 removing doc from SUBDIRS = May 27 11:48:17 as doc is removed ANYWAY... May 27 11:48:26 :nod: May 27 11:48:39 though i *hate* that detail May 27 11:48:52 so if I patch this out right, the .deb won't contain the doc in the first placre May 27 11:49:35 never got the reasoning behind docpurge on a system with 100 times the storage of my first linux system that had proper full manpages May 27 11:50:37 it makes the flash images illegal May 27 11:50:39 that is the point :) May 27 11:50:55 (as it gets rid of possibly required license documents) May 27 11:51:05 SEE ALSO May 27 11:51:07 alarmd(8) May 27 11:51:09 IroN900:~# man alarmd May 27 11:51:10 No manual entry for alarmd May 27 11:51:11 right May 27 11:51:19 at least the manpages could have been kept May 27 11:52:38 divVerent: (illegal) HAHAHARRHARRRHRRR gooooood point! why didn't you come up with that in 2009? May 27 11:53:47 FSCK docpurge! May 27 11:54:48 ^^^ alarmclient --help; if you're wondering May 27 11:55:13 DocScrutinizer05: I had no N900 in 2009 :) May 27 11:55:24 and I do somewhat see the point of docpurge... some packages are bigger in doc than in content May 27 11:55:26 great pun! google for "man alarmd" to see how good May 27 11:56:25 well, for a distro maintainer who prefers messybox over proper unix shells&tools, this might actually be an issue May 27 11:57:13 I recently had namcap complain on an arch linux package I made May 27 11:57:18 that the doc part is bigger than the rest May 27 11:57:20 cause: LOTS of pngs May 27 11:57:33 for a distro maintainer maintaining a distro meant exclusively for a target platform that has 32GB(!) storage, it's so utterly silly though May 27 11:57:50 no, I want to use the 32GB May 27 11:57:57 I actually like the N900's partitioning May 27 11:58:01 and would not want to waste an extra G for the OS May 27 11:58:11 having said that - / and /home just never gets full anyway May 27 11:58:12 so what? you're free to purge the manpages any time May 27 11:58:33 better idea - linking manpages to MyDocs :) May 27 11:58:36 then anyone can delete them May 27 11:58:43 indeed May 27 11:59:02 well, back in PR1.0 times, they had not even any /opt May 27 11:59:27 so they thought they should fit everything into the 240MB of NAND May 27 11:59:49 and thus docpurge might have made sense, for histerical raisins May 27 12:01:19 however what *really* sucks is that it seems mandb-n900 pkg also lacking those fremantle-specific manpages like alarmd(8) May 27 12:01:59 I really wonder where to find that alarmd(8) page, when even aunt google fails May 27 12:03:43 ok, deborphan uploaded now... let's see what happens May 27 12:03:58 BTW, by what I did, I only removed the manpages, but not the other docs... weird May 27 12:04:06 but deborphan needs no manpages anyway, --help is sufficient May 27 12:05:04 ~#maemo autostart is /etc/X11/Xsession.d/* May 27 12:05:05 DocScrutinizer05: okay May 27 12:23:33 (http://maemo.org/packages/view/nelisquare/) hmmm, seems the 0.5.0 version somehow forgot it consists of a i386 as well as an armel May 27 12:30:02 DocScrutinizer05: ah, interesting :) as for autostart May 27 12:37:28 hello May 27 12:38:17 I'm wondering... according to the flashing guide there may be issues with win 64bit editions May 27 12:38:38 does it mean linux 64bit may have issues as well? May 27 12:39:03 I have plenty of linux distros to use for this operation but most of them are 64bit May 27 12:39:57 and only windows 7 64bit otherwise.. May 27 12:40:30 looks like my deborphan build worked May 27 12:41:04 GogoPogo: yes, there can be issues on linux 64bit, from how I read it May 27 12:41:07 I did it anyway and it worked May 27 12:41:12 I may use virtualbox as well, but I'm not sure I can get the usb connection to work inside the virtualmachine as well.. I always get some error.. May 27 12:41:15 just be prepared - if your N900 gets "bricked" and you see no way out May 27 12:41:21 try booting Ubuntu 32bit live CD and flash from there May 27 12:41:24 as a last resort May 27 12:41:28 no, do not use a VM for that May 27 12:41:35 ok May 27 12:41:51 also, you better have a backup of the device anyway May 27 12:41:58 flashing eMMC erases MyDocs, AFAIK May 27 12:42:11 old linux distros I still have around come with 32bit architecture... too old maybe?? May 27 12:42:22 ubuntu live cd should work May 27 12:42:25 but try your 64bit distro first May 27 12:42:49 I still have to understand if I have to flash combined + vanilla right after, or the contrary May 27 12:43:01 both work, but have different caveats from how I read May 27 12:43:18 follow the current docs exactly as written May 27 12:43:32 there is both kinds of instructions, both work, if you make sure you do nothing weird May 27 12:43:39 the biggest issue is if you flash emmc only, the device won't boot any more May 27 12:44:00 but that is then fixable by flashing the other part, AND emmc again May 27 12:44:52 so I suggest to just take whichever docs you find next, and follow them to the letter. That will work. May 27 12:46:08 mmm... ok! I really hope I'm not going to brick my phone, even if Doc said to me N900 is almost unbrickable May 27 12:46:52 GogoPogo: N900 (with working USB) is unbrickable May 27 12:47:13 good to know.. May 27 12:54:14 a question I have is.. will I be automatically re-signed up to MyNokia once rebooting? May 27 12:54:22 I really hope not May 27 12:55:32 freemangordon: finally register with nickserv please! May 27 12:55:42 I 'm not willing to share my SIM contract data with Nokia at all, and they surely store some even after unsubscription. I had to e-mail Nokia the first time this happened. May 27 12:56:09 DocScrutinizer05: will do when have some spare tme May 27 12:56:13 *time May 27 12:56:16 it takes 30s May 27 12:56:31 /msg nickserv help register May 27 12:56:34 sure, but I won't do it over RD over gprs ;) May 27 12:56:51 /msg NickServ REGISTER bar foo@bar.com May 27 12:57:19 Syntax: REGISTER May 27 12:58:56 thanks! May 27 12:59:13 you now have to answer verification mail May 27 12:59:16 DocScrutinizer05: where do you see that? May 27 12:59:22 DocScrutinizer05: i did it May 27 12:59:36 /ns info freemangordon May 27 13:00:04 DocScrutinizer05: now I am registered, anything else I need to do? May 27 13:00:24 /ns info freemangordon May 27 13:00:28 oops May 27 13:00:52 /ns id May 27 13:01:28 looks fine to me May 27 13:01:46 [2013-05-27 15:01:10] [Whois] freemangordon is logged in as freemangordon. May 27 13:02:56 DocScrutinizer05: "##maemo-admin Cannot join channel (+i) - you must be invited" May 27 13:03:05 retry May 27 13:03:10 I just fixed it May 27 13:10:24 this: /etc/X11/Xsession.d/04devlock-blocker *might* be interesting for some May 27 13:11:08 to break devlock, it should suffice to remove that file via rescueOS May 27 13:11:47 to lock file you might start that executable called in that file (though there are other - dbus-based - methods for that iirc) May 27 13:12:01 s/file/device/ May 27 13:12:02 DocScrutinizer05 meant: to lock device you might start that executable called in that file (though there are other - dbus-based - methods for that iirc) May 27 13:13:04 similar: /etc/X11/Xsession.d/05connui-conndlgs-cellular -> /usr/bin/startup-pin-query May 27 13:15:45 very educative: /etc/X11/Xsession.d/10hildon_welcome May 27 13:19:13 hi May 27 13:24:49 hi jonwil, long time no see May 27 13:25:03 :-) May 27 13:25:20 yeawh May 27 13:25:30 * jonwil wonders whats new in the community these days May 27 13:25:42 well, a lot of mess May 27 13:25:59 and worries, with infra May 27 13:26:22 and with elections, and whatnot May 27 13:36:07 hello,got a strange problem,my n900's imei is blank and can't connect to any network May 27 13:42:55 mmm... I can't complete my backup with PC Suite... at some stage of the copy it loses the connection with the phone. Ever happened to you? May 27 14:00:50 MohammadAG: http://mywiki.wooledge.org/ParsingLs May 27 14:01:06 mike__: that's bad May 27 14:02:23 mike__: might be modem chip hw error. Try removing battery, removing SIM, insert battery again, and test again May 27 14:03:18 mike__: try cmd "pnatd" and see if modem answers to "AT" May 27 14:04:10 (to stop pnatd you either close the shell or killall pnatd from a second shell) May 27 14:11:32 thank you! May 27 14:12:57 if it is the radio,is it repairable? May 27 14:21:31 there's that hack with springs inside the cans, which is the only thing that *might* help (if you don't consider reflow of the chips) May 27 14:22:09 is reflowing a temporary or a final solution? May 27 14:23:11 mike__: it's most likely permanent, considering how 90% of N900 failures can be attributed to cold solder May 27 14:23:44 great :) May 27 14:24:38 is it possible to do with a soldering station,or a rework station is essential? May 27 14:24:48 good question May 27 14:24:52 mike__: hot air evenly around the board may be enough May 27 14:25:35 DocScrutinizer05: it'd be a bad idea to simply aim a hot air gun on the board, right? May 27 14:25:42 I'd suggest preheating the whole board in oven to ~80°C, then hotair gun to the chips in question May 27 14:26:27 where "hot air" is a proper reflow station, nit a monster looking like a hairdryer May 27 14:26:33 not* May 27 14:26:34 what about to apply some flux with soldering tip? May 27 14:26:47 chops are BGA May 27 14:26:55 chips, dang May 27 14:27:07 which is part of the problem May 27 14:29:22 * DocScrutinizer05 frowns at http://talk.maemo.org/showpost.php?p=1346993&postcount=1225 May 27 14:29:38 * DocScrutinizer05 wonders why he bothers at all to answer May 27 14:30:18 >>I think freemangordon should not try to pester that lady via skype without prior email giving better info about what "we" are and what's our "project" and waiting for the (unlikely) invitation form Anne COUVIDAT to give her a call. To her freemangordon calling is similar to a video-club giving you a phonecall trying to convince you to become a member.<< May 27 14:30:21 tl;dr? May 27 14:33:04 mike__: you want to get the schematics (with component placement pics), and my highres scans of PCBA and PCB on maemo.cloud-7.de/N900-gallery May 27 14:33:28 ~schematics May 27 14:33:28 i guess schematics is http://wiki.maemo.org/N900_Hardware_Schematic May 27 14:33:37 DocScrutinizer05: HiFo has enough cash to buy IP? May 27 14:33:45 pfff May 27 14:33:51 sorry May 27 14:34:05 that's outright ridiculous May 27 14:34:25 yeah, there's no way they're just going to hand out the video decoder. May 27 14:34:37 even IF they buy the IP, they may not share it to any community member with sufficient knowledge to do sth with it May 27 14:34:49 what video decoder is this people are talking about? May 27 14:34:57 silly 720p May 27 14:35:03 jonwil: 720p codecs, episode 9 May 27 14:35:10 oh ok May 27 14:36:50 look, those ittiami (or whatever) guys *might* hand out the IP even without charge, but only under strict NDA and for sure not to a group of "3 stooges" like HiFo May 27 14:37:46 heck, until now HiFo not even been able to negotiate and sign contracts with Nokia May 27 14:38:19 and I'm 99.9% sure that's not Nokia's fault May 27 14:39:32 if somebody wants that 720p codec so badly, why not ask those guys to sell it to *him*? May 27 14:40:05 I for one don't need such 720p May 27 14:41:58 and I'd be really puzzled when HiFo would decide they want to waste a few kilobucks on such peripheral feature, while more pressing stuff around maemo.org infra at large gets neglected May 27 14:42:55 heck, we probably could get a fulltime sysop paid for one year instead that 720p stuff May 27 14:44:09 * DocScrutinizer05 anticipates whining "Now we have that awesome leete pretty 720p, and masterhacker can't build the gstreamer plugin since autobuilder is fsckdup" May 27 14:45:23 DocScrutinizer05: that'll just entice someone to make HDPatch.sh, and distribute it via Rapidshare, through Adf.ly links. May 27 14:45:40 great May 27 14:48:37 while I pull my hair about stuff like http://maemo.org/packages/view/pierogi/ May 27 14:57:39 I for one can think of far more useful things for Maemo than 720p video decode May 27 15:00:55 indeed May 27 15:01:37 particularly given the fact that not even AV-out can actually do 720 May 27 15:01:51 afaik May 27 15:02:12 so what for we need it anyway? May 27 15:02:17 like a better conversations app without the annoying scrolling issue I keep bumping into May 27 15:02:28 iirc the screen of N900 is 800*480 May 27 15:02:42 where it will randomly scroll to the start of the conversation for no reason May 27 15:02:54 DocScrutinizer05: simple: to avoid re-encoding videos May 27 15:02:56 hi, apart from iptables, is there a simple way to forbid any data transfer when the n900 is connected to my provider MMS APN ? May 27 15:03:07 DocScrutinizer05: but, there is more to be done for that... May 27 15:03:40 peetah: ? May 27 15:03:43 by any, I mean other than data targeted to the MMS access point May 27 15:03:54 ooh May 27 15:04:13 ask frals May 27 15:04:28 hm... gstreamer0.10-ffmpeg is installed, yet I cannot play .ape music files May 27 15:04:36 where to fix? May 27 15:04:50 nfc May 27 15:05:03 DocScrutinizer05: ok thuttu77 May 27 15:05:06 it would also be nice to have some new games to play on the N900 that are actually a good fit for mobile devices and not just a port of some desktop game that doesn't fit well in a touchscreen environment May 27 15:05:13 DocScrutinizer05: ok thanks (sorry) May 27 15:05:32 DocScrutinizer05: renaming file from .ape to .ogg fixes it May 27 15:05:39 so, where are the file type assocations? :) May 27 15:06:01 (this works because ffmpeg doesn't care for the extension at all - it always probes) May 27 15:06:02 peetah: I don't know if MMS APN IF is supposed to get used by anything except fmms anyway May 27 15:06:33 divVerent: haha May 27 15:06:46 so probably an issue of minetypes? May 27 15:07:04 check mediaplayer .desktop! May 27 15:07:25 when the only connection available is MMS APN, and modest or telepathy starts looking for mail or xmmp account, won't they try to use this connection ? May 27 15:07:49 depends on route May 27 15:07:55 and whatnot else May 27 15:08:14 DocScrutinizer05: ok, know where to look then May 27 15:08:20 e.g no maemo app is using USB network connection per default May 27 15:08:24 also, it probably is a bug of the gstreamer0.10-ffmpeg package... May 27 15:08:52 yeah, might not ship a comprehensive set of mimetypes May 27 15:09:01 no, no bug in that package May 27 15:09:07 it has no mimetypes at all shipped... these go elsewhere May 27 15:09:20 o.O May 27 15:09:36 it's the original debian gstreamer ffmpeg package, why should it ship mimetypes? May 27 15:10:03 well, the original pkg probably doesn't work too well on maemo anyway May 27 15:10:07 BTW, .ape decoding is... somewhat broken May 27 15:10:15 it barely works on non overclocked CPU May 27 15:10:20 do ANYTHING, and it stutters May 27 15:10:33 didn't know .ape was that CPU intense May 27 15:10:57 so... I take back anything I said, and will NOT suggest adding the mime type May 27 15:11:07 ;-P May 27 15:11:15 renaming hack is good enough for me May 27 15:11:48 also, this CD sucks anyway :) May 27 15:12:22 hah May 27 15:13:55 http://www.youtube.com/watch?v=zDzUG-bno6Q May 27 15:24:38 DocScrutinizer05: haha, this is SO pointless May 27 15:24:55 as if ANY N900 could reach the end of a DVD on battery May 27 15:25:12 hostmode after all means the USB port is taken, and can't be used for the charger May 27 15:25:33 ~hostmode-powered May 27 15:25:33 from memory, hostmode-powered is http://talk.maemo.org/showthread.php?p=921203#post921203 May 27 15:26:28 haha, insane May 27 15:29:21 insane??? May 27 15:29:51 yes, but I can see that it would work May 27 15:30:05 N900 USB is insane, H-E-N hacks been my best contribution to maemo ever May 27 15:30:36 yes, doing hostmode on the N900 is already insane enough :) May 27 15:30:36 a lot of people like USB hostmode in N900 May 27 15:30:48 is is an interesting feature, but the N900 clearly was not made for it May 27 15:31:05 making it even more interesting that it works at all May 27 15:31:53 of course, this USB cable is dangerous... but that's a different story :) May 27 15:32:04 well, took me only ~6 months of investigations and developing to make it work May 27 15:32:11 basically, an A-A cable is something that should never have been made... but it has May 27 15:32:42 yes... it is a cool feature to have, but it's the kind of thing I'd never do May 27 15:32:49 the cables alone are enough for me to object to May 27 15:33:08 even though the N900 does not power the port (so in the speciifc case of the N900 the "danger" part does not exist) May 27 15:33:24 err wut? May 27 15:33:33 or is it able to also power the USB port? May 27 15:33:38 sure May 27 15:33:44 the "danger" is when you accidentally use hostmode, AND are connected to a PC May 27 15:33:50 two USB hosts is normally no big deal May 27 15:33:58 but if both try to power the bus, that can get dangerous May 27 15:34:38 I'd more think it gets dangerous when one of them is *not* powering the VBUS lines May 27 15:34:41 so in a "proper" world, hostmode should require special host side connectors, and not abusing a USB device socket for host May 27 15:35:17 that's also how USB was designed to avoid this issue May 27 15:35:43 that's why N900 officially can't support hostmode or OTG, it has a USB-micro-b receptacle May 27 15:36:03 and I have seen the same hack on Android devices... and hate it equally there May 27 15:36:14 to me it looks as bad as an AC plug-AC plug adaptor May 27 15:36:26 (that one admittedly is a lot MORE dangerous) May 27 15:36:57 and no, please don't find a google images link where someone has actually made this adaptor I DON'T WANT TO SEE IT :) May 27 15:38:30 meh May 27 15:38:51 your argumentation isn't based on good facts May 27 15:39:25 it is, you just can't have two power sources on the same bus... unless the bus/the devices are designed for it May 27 15:39:30 in case of USB that is not the case May 27 15:39:36 connecting two USB-hosts together first instance does nothing at all May 27 15:39:49 apart from the power issue May 27 15:40:08 what power issue? May 27 15:40:22 the devices may disagree on their definition of GND :) May 27 15:40:27 no May 27 15:40:46 not more than any other two sevices May 27 15:40:50 also, they may simply not be build to expect another device already powering the bus, possibly causing harm May 27 15:41:36 basically, the circuits are "allowed" by the standard to take damage if there is already a higher voltage on the bus than what they would provide... and be it 5.1 V vs 4.9 V May 27 15:41:50 and I really don't trust PC motherboards more than I can throw them May 27 15:42:28 e.g. assume the voltage regulator tries to short any excess voltage (like a zener diode does) May 27 15:43:05 meh May 27 15:43:29 don't argue with an EE ;-) May 27 15:43:56 e.g. assume the voltage regulator tries to short any excess voltage (like a zener diode does) May 27 15:43:58 e.g. assume the voltage regulator tries to short any excess voltage (like a zener diode does) May 27 15:44:00 sorry May 27 15:44:00 the only "dangerous" case I could think of is plugging a powered host to a PC that's off May 27 15:44:04 thing is, ARE normal conforming USB host sockets designed to deal with that kind of abuse, or not? May 27 15:44:11 yes, that can be even worse May 27 15:44:16 if it then shorts the pins (which it may May 27 15:44:18 ) May 27 15:44:36 short is no problem at all May 27 15:44:50 USB is short-resilient May 27 15:44:53 haha May 27 15:44:56 I've seen the opposite May 27 15:45:09 N900 *is* May 27 15:45:10 someone forcing a USB plug in the wrong way round, even causing a spark and reboot of the system May 27 15:45:13 (but the USB device survived) May 27 15:45:19 the N900 side I trust a lot more here :) May 27 15:45:27 I mean, there is only so much current the battery can yield May 27 15:45:44 the battery? May 27 15:46:01 where else should the N900 get the current from when it powers the port? May 27 15:46:13 from mars? May 27 15:46:16 *sigh* May 27 15:46:16 that could work May 27 15:46:38 battery is 3.7V, USB is 5.0V May 27 15:46:45 yes May 27 15:46:50 voltage convertors exist May 27 15:46:59 thing is, if the N900 is in host mode AND powers the port on its own, shorting the pins there should still be pretty safe, mainly because the current is at the very least limited by what the battery can do May 27 15:47:03 afk May 27 15:47:23 too silly discussion May 27 15:47:46 well, you said the N900 can power the port, so I assumed a voltage up-convertor somewhere inside May 27 15:51:23 and you think this step-up converter was designed in a way so it could blow the overcurrent protection in battery? May 27 15:51:44 sounds like a pretty poor design May 27 15:52:26 bq24150 in boost mode can provide 200mA@5V, then overload protection kicks in and it throws error on status May 27 15:52:56 frals: several questions for you when you'll have time: 1/ when the MMS APN IF is activated, does modest or telepathy attempt to send data through this connection ? 2/ if yes is there a simple way to forbid data transmission other than from fmms to the MMS access point ? 3/ I activate a VPN with a /etc/network/if-up.d script when the Wifi AP I'm connected to is not one that I control: I guess that the VPN is activated when the MMS APN IF May 27 15:54:31 your post got trunkated? May 27 15:55:02 it may be: I tried to ask everything on a single line, but it was quite long May 27 15:55:14 lets try again May 27 15:55:18 DocScrutinizer05: no, I did assume there is ALSO an overcurrent protection there May 27 15:55:24 there's a max-line-length enforced by freenode May 27 15:55:27 just was saying that even in worst case, not too much damage can happen May 27 15:55:37 frals: 1/ when the MMS APN IF is activated, does modest or telepathy attempt to send data through this connection ? May 27 15:55:49 we got that, peetah May 27 15:55:56 because the battery also has an overcurrent protection (that then breaks the battery for good) May 27 15:55:57 frals: 2/ if yes is there a simple way to forbid data transmission other than from fmms to the MMS access point ? May 27 15:55:58 ...ted when the MMS APN IF May 27 15:56:05 was last that arrived May 27 15:56:34 frals: 3/ I activate a VPN with a /etc/network/if-up.d script when the Wifi AP I'm connected to is not one that I control: I guess that the VPN is activated when the MMS APN IF is activated; how can I detect in my script that the MMS APN IF is activated ? May 27 15:56:36 while I've seen PCs just connect the +5V line of the PSU directly to USB May 27 15:56:55 thus quite large currents can happen on the USB port, way overloading the mainboard or cable May 27 15:57:22 DocScrutinizer05: thanks for noticing, and sorry to every one for the noise May 27 15:58:16 divVerent: I just don't like to defend H-E-N against that poorly based accusations. H-E-N been properly evaluated in every possible aspect and considered safe May 27 15:58:29 I do believe H-E-N's side of things is safe May 27 15:58:40 I accuse the mere existance of the cables because they can be abused elsewhere May 27 15:59:49 I mean, it is quite easy to engineer the N900 side of things so that no matter what happens (and which direction), the current flowing is bounded somewhere sanely May 27 16:00:02 I mainly fear PC-PC connections May 27 16:00:34 which are safe except when one PC is in off state May 27 16:00:45 especially if the other PC does not limit current on USB May 27 16:00:52 like I have seen before (thus the spark and reboot) May 27 16:01:21 when both PCs try to power the bus, USUALLY the current is still limited by the inner resistances and the small voltage difference May 27 16:01:33 yes May 27 16:01:41 it can still be like 1 or 2 A, but nothing that melts cables May 27 16:02:17 particularly since a power supply usually isn't any kind of power sink, even when you feed a few volts more than it's set to provide May 27 16:02:20 thing is, I just don't like the existence of such cables with which you have to be VERY careful to not misuse them May 27 16:02:31 thus, a hostmode cable with USB micro on one side and USB-A on the other is still quite safe May 27 16:02:48 as USB micro is typically used for small devices which won't take large amperages anyway May 27 16:03:08 but the USB A-USB A adaptor idea offends me May 27 16:03:53 * DocScrutinizer05 shrugs and idly wonders if freenode has a #usb channel May 27 16:04:10 also, I mentioned this because none of these cables are in any USB spec, for a good reason May 27 16:04:29 heck, do you think I didn't ever heard of that? May 27 16:04:33 and I wouldn't oppose if the USB consortium used all lawyer forces available to get rid of all nonconforming cables May 27 16:04:40 no, I don't May 27 16:04:52 so what are you telling us? May 27 16:05:05 I consider these cables/adaptors evil and risky. That is all. May 27 16:05:20 we already heard and got that May 27 16:05:30 that's all I wanted to say May 27 16:05:44 and it's why I'll never use hostmode on such a device, no matter how convenient May 27 16:06:00 DDTT May 27 16:06:16 up to you May 27 16:06:22 having said that, I did read about that hack where they used a AC plug-AC plug cable to insert a UPS into a running circuit without breaking it at any time May 27 16:06:35 I'd never do such a thing, but yes, it was neat that they could do that May 27 16:06:55 insert an UPS into the circuit without interrupting it by clever use of power strips and that dangerous plug-plug AC cable May 27 16:08:30 I think this is going from bashing H-E-N as silly useless dangerous to sth completely OT now May 27 16:09:18 yes, sorry May 27 16:09:42 to be a bit more ontopic - I'd like to see a photo of all the cabling for that external DVD drive setup if we also power the N900 using an external power supply :) May 27 16:09:58 it probably would remind me of the mac mini... it's just too small to hide all the cabling mess behind it May 27 16:13:03 as for the hostmode issues, maybe the USB consortium should sort out that issue for USB 3.1... basically some kind of safety measures to fix hostmode May 27 16:13:25 actually enforcing negotiated current limits would probably suffice :) May 27 16:18:33 there *is* a safety limit of 500mA/port May 27 16:18:46 always been May 27 16:19:54 if your crappy mainboards don't even provide a blow-fuse for the USB port, that's not USB-cert's fault May 27 16:20:45 I told you USB is supposed to be short-resilient May 27 16:21:11 per definitionem May 27 16:22:28 it's not mandatory to enforce the possibly lower supply current negotiated durin ENUM May 27 16:23:13 and while some borads do it under windows OS, it's usually not done at all under linux May 27 16:23:15 DocScrutinizer05: oh, you say these mainboards are non conforming? THAT is good to know May 27 16:23:35 the mainboards where USB V+ is connected to +5V from the PSU directly May 27 16:23:52 as evil as it gets May 27 16:24:26 since which USB standard is this the case, BTW? May 27 16:24:27 they're abusing the PCB copper traces as blow-fuse May 27 16:24:33 hehe May 27 16:24:40 in the specific case, the PSU turned off first :) May 27 16:24:41 thus the reboot May 27 16:25:00 so at least the PSU was good May 27 16:26:18 well, the specs afaik say: short-resilient, must support 500mA May 27 16:26:39 for standard PC USB hosts May 27 16:27:05 and the plug specs are rated at 1...2A max May 27 16:27:34 so it's pretty clear how the short protection has to work May 27 16:27:44 and it's been like this since 1953 May 27 16:27:57 peetah: depends on the connection mode and if the carrier supports multiple apn contexts open at once - if its the "ugly hack handler" (or whatever i called it), it brings up a parallell connection and sets up a route to the mmsc so in that case it *should* only route the mms packages over that connection May 27 16:28:51 peetah: you could probably check what the gprs0 apn context is - but i can provide any example from the top of my head May 27 16:29:29 peetah: or you could hack around in the fmms scripts to add a tmp file or whatever when it brings up the iface, its just a shell script anyway May 27 16:29:40 ifconfig and route should give some data May 27 16:31:06 bb5 supports usual 4 apn contexts? May 27 18:45:15 hello, any reason why the maemo mail client on my n900 cannot authenticate with the SMTP server (LOGIN) after issuing STARTTLS? May 27 18:45:35 According to the mail server logs, it advertises "AUTH LOGIN PLAIN" to the client, but the client then doesn't do anything. May 27 19:06:00 hmmm May 27 19:06:14 ipv6? (far shot) May 27 19:06:53 check cssu changelogs if there's sth mentioned about TLS May 27 19:07:39 somehow TLS rings a bell here, but I can't recall what it was May 27 19:07:57 or I might be mistaken on it May 27 19:17:13 frals: ok thanks May 27 19:17:36 gah, catorise is insane May 27 19:49:53 get apmefo, I seen "catorize compatibility" in a changenote May 27 19:57:20 whats catorise May 27 19:57:40 oh ok May 27 20:01:19 &g 25 May 27 20:01:27 *grr* May 27 20:06:24 is http://maemo.merlin1991.at/light/extras-devel/ still updated May 27 20:06:27 ? May 27 20:06:47 or there's just nothing new in extras-devel since mar 27? May 27 20:06:58 check http://maemo.merlin1991.at/light/extras-devel/dists/fremantle/free/binary-armel/Packages.diff/ May 27 20:07:04 and you'll see last update happened today May 27 20:07:30 oh, awesome May 27 20:07:39 Last modified doesn't refer to files inside, that confused me May 27 20:07:47 http://maemo.merlin1991.at/light/extras-devel/dists/fremantle/ shows may 27 alright May 27 20:07:56 thanks merlin1991 May 27 20:08:24 it's because it's a nested folder structure May 27 20:08:35 right below dists nothing gets touched ever May 27 20:08:46 right May 27 20:31:24 SDK et al: http://privatepaste.com/e4a5b13fad May 27 20:41:25 ~#maemo SDK is http://privatepaste.com/e4a5b13fad or "wget http://repository.maemo.org/tabletsdev/explicit/maemo-dev-env-downloads/downloads/Readme_Ubuntu_Lucid_Desktop_SDK_Virtual_Image_Final.txt" May 27 20:41:25 okay, DocScrutinizer05 May 27 20:55:17 hi... quick question.. alternative to ifconfig for maemo x-terminal? May 27 20:55:33 or is ifconfig installable maybe? May 27 20:55:59 GogoPogo: what? May 27 20:56:21 GogoPogo: did you try to type ifconfig in the terminal? May 27 20:56:28 yes it doesn't find it May 27 20:57:02 go root? May 27 20:57:05 BusyBox v1.10.2 (Debian 3:1.10.2.legal-1osso31+0cssu0) multi-call binary May 27 20:57:12 Usage: ifconfig [-a] interface [address] May 27 20:57:37 ok, it is supposed to work then May 27 20:57:38 thank you May 27 20:57:51 I haven't checked /bin May 27 20:58:04 GogoPogo: it is part of busybox May 27 20:58:26 and BB is surely pre-installed May 27 20:59:23 yes it is May 27 20:59:30 of course :) May 27 21:00:15 ok ok, it was a user permission related issue May 27 21:00:35 it still behaves strangely May 27 21:00:41 I want to reflash my phone May 27 21:02:17 the file /etc/network/interfaces lists 3 network cards: lo, eth0, usb0 May 27 21:02:30 but ifconfig finds way more interfaces May 27 21:02:32 normal? May 27 21:02:39 GogoPogo: and what do you expect? May 27 21:03:06 phonet0 is GPRS May 27 21:03:22 I'm trying to learn May 27 21:03:29 ooh, ok :) May 27 21:04:32 upnlink0 is wifi? May 27 21:04:44 hmm, no idea May 27 21:05:19 if they are listed inside ifconfig results I should be able to edit their parameters somewhere inside the root shouldn't I? May 27 21:05:43 make believe I want to edit mac addresses? May 27 21:06:59 why would you want to do that? May 27 21:07:31 I'm used to changing my MAC addresses on and off on PC May 27 21:07:55 .oO(???) May 27 21:08:26 Doc! I can't complete my Backup.. is there another way to copy all the /home folder at least? May 27 21:08:39 I'm used to shooting my foot with a frozen cherry every now and then May 27 21:08:51 PC Suite always fails May 27 21:09:14 whf pc suite? sounds like windows. sorry no idea what's on windows May 27 21:09:28 Doc.. why ? I change my ethernet netcard mac... May 27 21:10:12 GogoPogo: you don;t need to use PC Cuite to make backups May 27 21:10:13 I thought you were referring to backing it all up with Nokia PC Suite... how do you suggest I back it up? May 27 21:10:24 BUT... May 27 21:10:45 backup does not mean "copying the whole /home" May 27 21:11:01 yeah I know May 27 21:11:12 if you want to copy that on yoiyr PC, install WinSCP and use that May 27 21:11:12 at least that would be good to me May 27 21:11:30 *your May 27 21:11:39 ~bm May 27 21:11:40 it has been said that backupmenu is http://talk.maemo.org/showthread.php?t=63975 May 27 21:11:50 and there is "backup" application May 27 21:12:02 and what DocScrutinizer05 said, BM May 27 21:12:54 backup application works smoothly May 27 21:13:07 I'll go for backupmenu May 27 21:13:41 they serve complementary purposes May 27 21:13:58 yeah I got it May 27 21:15:16 Doc, is there a list of network interfaces names correspondences? like phone0 = GPRS? May 27 21:15:26 no May 27 21:15:51 I think most of them are pretty self-explaining May 27 21:16:08 not to me :\ May 27 21:16:29 well, you don't usually bother about them anyway May 27 21:16:39 but I do! May 27 21:17:06 last time I had to *do* sth about that been ~3 years ago when I traced phonet IF with wireshark May 27 21:18:13 well the mac address I'll be messing with is the wlan0 anyway May 27 21:18:36 so I recognize that name corresponds to wifi card May 27 21:18:53 but there are plenty more and I'm so curious :) May 27 21:18:56 anyway there are only 4: lo phonet0 wlan0 wmaster0 May 27 21:18:57 GogoPogo: iirc wlan and BT MAC addresses are kept in CAL ;) May 27 21:19:15 and probably gprs when you're logged in to GSM May 27 21:19:34 CAL stands for? May 27 21:19:37 DocScrutinizer05: usb0 may appear too as some other if connected to pc suite May 27 21:19:47 sure May 27 21:19:48 yeah indeed it appears May 27 21:20:02 ~CAL May 27 21:20:02 extra, extra, read all about it, cal is a calendar. try $(cal 1752) May 27 21:20:04 all pretty self-explanatory May 27 21:20:08 hmm :( May 27 21:20:28 I have others though May 27 21:20:34 ~liskeys cal May 27 21:20:40 won't fly May 27 21:20:46 yep :) May 27 21:20:48 ~listkeys cal May 27 21:20:50 Factoid search of 'cal' by key (16 of 335): /usr/local ;; bcalc ;; calendars for the web ;; calvarez ;; classical ;; cpu scaling ;; debian things > 2mb of locale files ;; don't recall which ;; dos call ;; fcal ;; iirc, the xcalibrate ;; jargon automagically ;; jargon canonical ;; libical ;; mscal2html ;; mythical man month. May 27 21:21:15 ~listvalue talk.maemo.org May 27 21:21:21 ~listvalues talk.maemo.org May 27 21:21:22 Factoid search of 'talk.maemo.org' by value (18 of 35): #maemo bnf ;; #maemo donate ;; #maemo kp ;; #maemo mirror ;; #maemo truecrypt ;; #maemo u-boot ;; #maemo vkbd-portrait ;; aegis-no-thanks ;; alopex ;; backupmenu ;; bq-calibrate ;; cssu-thumb ;; ereswap ;; flashplayer ;; hen ;; hostmode ;; hostmode-powered ;; hotswap. May 27 21:21:40 I have: lo, e:0, ier:0, phonet0, er:0, upnlink0, :0, me:0, rier:0, wmaster0, :0 again May 27 21:22:05 nfc, never seen May 27 21:22:10 DocScrutinizer05: "Configuration Access Library" May 27 21:22:37 DocScrutinizer05: https://gitorious.org/community-ssu/libcal/blobs/27df167a08103f060f297cbdfec2a360e5d1eec4/src/cal.c#line4 May 27 21:22:52 anyway, I dunno if changing wlan MAC is that easily done May 27 21:23:01 freemangordon: eh? May 27 21:23:05 so what? May 27 21:23:18 DocScrutinizer05: in case you want CAL in the factoid May 27 21:23:18 I know that May 27 21:23:25 nope May 27 21:23:58 ~#maemo CAL is http://talk.maemo.org/showthread.php?t=20465 May 27 21:23:59 DocScrutinizer05: okay May 27 21:25:28 GogoPogo: AFAIK wlan MAC is read from CAL end send to kernel via a netlink socket or whatever was it called May 27 21:25:33 *and May 27 21:25:59 GogoPogo: not sure how easy is to be changed May 27 21:26:16 I usually set a line inside /etc/network/interfaces May 27 21:26:40 below the card namehwaddress
May 27 21:26:40 well, try it, it might work May 27 21:27:01 chances are low, but still May 27 21:27:07 hwaddress
May 27 21:27:22 where do I exactly find the CAL? May 27 21:27:32 ~cal May 27 21:27:32 i guess cal is http://talk.maemo.org/showthread.php?t=20465 May 27 21:27:40 ^^^ :) May 27 21:28:08 GogoPogo: you need libcal May 27 21:28:27 but you don't want to do that, honestly May 27 21:28:29 and some code to use it May 27 21:28:39 DocScrutinizer05: :nod: May 27 21:28:55 it's actually the only way to thoroughly brick your N900 May 27 21:29:09 HAM hasn't got it May 27 21:29:19 GogoPogo: which one, libcal? May 27 21:29:28 yeah May 27 21:29:32 GogoPogo: you have that preinstalled May 27 21:29:41 but need some code to use it May 27 21:29:57 libcal is closed source and no documentation is available ;) May 27 21:30:17 I'm honestly failing to get the approach GogoPogo is taking on a new unknown system May 27 21:30:18 sh*t :) May 27 21:30:31 GogoPogo: honestly, you'd better follow DocScrutinizer05's advice ant not play with CAL May 27 21:30:35 Doc, is there a good approach? May 27 21:30:40 *and May 27 21:31:16 well, maybe one: don't use root or sudo for first 6 months May 27 21:31:22 :) May 27 21:31:28 non sense! :) May 27 21:32:16 unless you're adventurous ;) May 27 21:32:19 maemo is linux and linux is fun... I should sell it back otherwise :D May 27 21:32:20 you know what su tells you? with great power comes great responsibility May 27 21:32:33 you should know your shit before you mess with it May 27 21:32:34 so ? It's my phone.... May 27 21:32:54 it's our nerves May 27 21:33:08 GogoPogo: really, the only way to make your n900 unusable is to mess with CAL, everything else can be fixed May 27 21:33:41 and I honestly fail to see why you would need to change MAC, to start with May 27 21:34:08 I'll try not to touch it /angel face/, I haven't found it yet anyway /devil face/ May 27 21:34:16 not even experienced users/devels ever asked for it May 27 21:34:43 doc... mac address can be retrieved by packets deep analysis May 27 21:34:56 GogoPogo: on the other hand you're right, it is your device, here is libcal replacement source code https://gitorious.org/community-ssu/libcal May 27 21:34:59 errr .....k May 27 21:35:09 it usually can be changed.. May 27 21:35:23 probably even by not-so-deep "analysis" May 27 21:35:49 GogoPogo: You have to figure out what needs to be done to change tha MAC address in CAL May 27 21:35:52 your statements make no sense May 27 21:35:55 so what? You agree with me it can be found, and it can be used to identify your machine May 27 21:36:20 GogoPogo: use iptables May 27 21:36:32 if you think your MAC needs to be changed May 27 21:36:35 sure, an APN would be able to associate your MAC to your device May 27 21:36:42 that's what a MAC is used for May 27 21:36:56 AP May 27 21:36:59 not APN May 27 21:37:38 freemangordon: iptables is an app? May 27 21:37:38 but so what again? the AP/hotspot will need to authenticate you anyway, unless you're using public open hotspots May 27 21:38:23 and - just in case you suffer a misconception here - the traffic from AP to internet at alrge does _not_ carry your MAC May 27 21:38:42 :nod: May 27 21:39:14 GogoPogo: ask google for iptables May 27 21:39:20 I completely fail to see the purpose of this stuff you plan to do May 27 21:39:47 unless you're an evil blackhat May 27 21:40:04 Doc, you're not supposed to. I don't understand why you want me to justify.. May 27 21:40:35 that plans to penetrate some company's WLAN and still you don't want to throw away your N900 after having committed the crime May 27 21:40:56 ahahaha May 27 21:41:25 GogoPogo: simply because you're asking nonsensical questions that nobody else would ever benefit from if we waste hours to explain all that stuff to you May 27 21:42:11 if that latter is what you plan: forget it, IT forensics can identify your device even when you changed the MAC May 27 21:42:28 :D May 27 21:43:04 that's paranoia Doc, something I suffer from and can't help doing May 27 21:43:11 there are other ways than simple MAC addr to identify a platform and even a particular wlan chipset May 27 21:43:53 I honestly suggest to color your hair twice a day and waer another sunglasses every 10 min May 27 21:44:09 you got it :D May 27 21:44:10 changing MAC is mere nonsense May 27 21:44:33 well, many would disagree May 27 21:44:35 completely useless for anything I could think of May 27 21:45:32 there's exactly one usecase for MAC change: when you want to fake another WLAN client, on an AP with MAC filter May 27 21:46:30 GogoPogo: it is useless, MAC lives only in LAN, the first router you're connected to changes your MAC with its own (to put it simplified) May 27 21:46:57 well, I said that May 27 21:47:33 yep, http://en.wikipedia.org/wiki/MAC_spoofing May 27 21:48:35 when you hope to stay anonymous when you use same public hotspot several times, then forget it as well. Your traffic patterns will expose your identity May 27 21:49:16 so changing your MAC onkly shows you're rogue but silly and hope to hide something May 27 21:49:53 unlike MAC, your traffic patterns _will_ propagate to larger internet May 27 21:50:01 sometimes you need to change MAC if ipv6 priv address not working in open network (and you do not want to use calculated public ipv6 address from mac) May 27 21:50:38 Pali: n900 does not have ipv6 (by default) :) May 27 21:51:15 Doc, you're indirectly saying that I'm silly, like I offended you.. really.. chill out. That's getting kinda tiresome. May 27 21:51:58 no, i say what you plan to do is silly, unless there's some very obscure sound rationale that we all fail to see May 27 21:53:08 so far your reasoning why you want to change MAC wasn't convincing May 27 21:53:33 I don't need to convince you in order to talk on here May 27 21:53:39 it's plain useless for all purposes you gave hint towards and we tried to figure May 27 21:54:14 freemangordon: mine has :-) May 27 21:54:28 no, you don't. But you will lose my support (and probably everybody else's) when you ask for stuff that nobody sees a good reason it should get done at all May 27 21:54:29 it's not plain useless, you can find many good and bad reason, and Pali wrote down one May 27 21:55:09 and now I'm done with that, and afk May 27 21:55:51 just one last advice: don't touch CAL, you lack the mandatory skills for that May 27 21:57:19 lol, all boils down to skills ownership.. like it wasn't obvious I lack them.. it's good to stress it isn't it? May 27 21:57:36 btw, if you want to change mac address, why not to use ifconfig/ip or udev script? May 27 21:57:43 thanks for your time, I won't bother with nonsense anymore May 27 21:58:15 Pali, I was trying in fact :) May 27 21:58:54 $ sudo ip link set dev "$interface" address "$mac_address" May 27 21:59:18 install package iproute2 from extras-devel which contains command ip May 27 21:59:34 this *should* work for changing mac address May 27 21:59:40 thanks a lot! I'll try right away May 27 22:00:24 will mosh (mobile shell) be accepted into the repo at some point? May 27 22:01:11 ecc3g, is there debian source package for that application? May 27 22:01:19 or is it just the usual 'needs a maintainer'... May 27 22:01:36 ecc3g: maybe, maybe not. Might have immanent power consumption issues May 27 22:01:49 no clue... haven't seen one, just raw src May 27 22:02:07 which should rule it out on testing. So it wouldn't go to extras, and stay in extras-devel May 27 22:04:27 curious how it does, say vs straight ssh, for power drain... May 27 22:05:20 iirc mosh protocol specs rely on very frequent keep-alive pings, that are by design a bug/flaw for mobile apps May 27 22:05:32 yow May 27 22:05:39 why to use mosh instead ssh?? May 27 22:05:45 indeed May 27 22:06:06 since it cuts thru your battery in no time, when on 3G May 27 22:06:21 mosh is broken by design May 27 22:06:51 yes that would suck... was hoping it would do the opposite - sendling chunks vs individual characters like ssh May 27 22:07:15 i'd wish for a ssh with a local readline option May 27 22:07:34 thats what mosh sounded like... May 27 22:07:46 false advertising? May 27 22:07:55 a big hoax May 27 22:08:02 only tab and return and certain control command send data and it watches when the terminal is switched into raw mode May 27 22:08:38 i tried mosh from a laptop to server, its useful if you have high latency links, but it is not power efficient May 27 22:08:38 well, afaik that's what ssh does, when configured correctly May 27 22:08:53 there are uses for it May 27 22:09:04 which results in ^C or ^D not working until you hit enter May 27 22:09:08 DocScrutinizer05: tell me how please May 27 22:09:31 iirc it can not do it May 27 22:09:52 ssh sends each and every keystroke May 27 22:10:06 (maybe it batches a few milliseconds together) May 27 22:10:21 it depends on your pts and terminal emulator what will be sent May 27 22:10:35 ask aunt google for "^C not working via ssh" OWTTE, and then see what they recommend to avoid. That's what you want to _do_ then ;-) May 27 22:10:37 and shell May 27 22:11:39 I think that there is no better protocol for remote shell May 27 22:11:52 you can define buffer fill threshold, buffer expiry time, and transmissions initiation chars, iirc May 27 22:12:47 today's standard config is buffer-fill-treshold=0 May 27 22:13:04 would be nice if there was something that would auto goto line mode in readline apps, and eat power in vim... but remote echo still works when in line mode (faked) May 27 22:13:29 or = 256 or sth and a timeout os 0.1s May 27 22:14:20 it is not always easy to detect... but always wishes out there. May 27 22:16:22 cehteh: yes, exactly. Mosh is for high-latency situations May 27 22:17:03 probably invented in times of GSM CSD May 27 22:17:25 or acoustic couplers ;-P May 27 22:17:49 shell over satellite May 27 22:17:49 the beloved "datenklo" May 27 22:17:59 or that, good point May 27 22:18:46 they still have datarates of 4800baud usually May 27 22:18:54 and a RTT of >1s May 27 22:19:16 (via sat phones and the like) May 27 22:19:40 barely sufficient to poll your email May 27 22:25:08 here you are: https://mg.pov.lt/maemo-irclog/%23maemo.2012-04-18.log.html#t2012-04-18T18:39:03 May 27 22:35:16 on hindsight, my final verdict on mosh back when should've been "yes, mosh acts almost like ssh - when you press backspace in ssh every 3s" May 27 22:36:08 wcdma/3G never will stop sending, the timeout is longer than 3s iirc May 27 22:48:34 hi, anyone has debian chroot with testing? it wants to upgrade libc6, which says it needs a more recent kernel May 27 22:49:31 and upgrading it seems a must for some essential packages May 27 22:57:57 cehteh: >>Mosh doesn't fill up network buffers, so Control-C always works to halt a runaway process<< -- and see `man stty` May 27 23:05:38 hahahahaha May 27 23:06:11 somebody does not know about pts and terminal emulator, so creating mosh? May 27 23:55:46 cehteh: man 3 cbreak May 27 23:56:03 cehteh: we're getting closer May 27 23:56:55 cehteh: my suggestion about aunt google was a tad... unspecific May 28 00:08:58 cehteh: info:/libc/Noncanonical Input May 28 00:09:40 > >The MIN slot is only meaningful in noncanonical input mode; it specifies the minimum number of bytes that must be available in the input queue in order for `read' to return.<< May 28 00:10:02 > >The TIME slot is only meaningful in noncanonical input mode; it specifies how long to wait for input before returning, in units of 0.1 seconds.<< May 28 00:11:27 but yes, it seems there's no option to make nifty stuff like "send on NL" happen May 28 00:38:30 cehteh: though there's those obscure tabN, crN etc "delay mode settings" which I have no clue what they do May 28 00:39:36 might be delays to allow slow tty to move carriage before next char to type gets sent, after cr, nl, tab, del, ... May 28 00:40:56 but then I fail to grok why most have only 0|1, just cr has 0|1|2|3 May 28 00:41:34 if those were delays, they should have a parameter in fractions of a second, not 2 or 4 modes May 28 00:43:37 so might be: cr0->don't send buffer on cr; cr1->send buffer on cr; cr2->send buffer on cr+nl, cr3->send buffer on nl+cr May 28 00:44:18 tab1->send buffer on tab, aso... May 28 01:12:33 I'm currently trying to chroot into a debootstrapped debian-sid, but looks like 2.6.28 (power kernel) is too old (FATAL: kernel too old). any ideas what else I could try? May 28 01:14:06 is there an easy way to get a newer kernel version running or would that be a little more tricky? May 28 01:15:44 (actually I just want to get opusenc running on maemo, but looks like that isn't in maemo, not even the community ssu. and thought a debootstrap would be the easiest way) May 28 01:16:29 T_X1: you're SOL, the glibc's targeted 2.6.32, IIRC May 28 01:24:15 rikanee: hm, okay. thx. do you happen to know what the minimum kernel would be for debian-wheezy (which has opus-tools, too, I think)? or how to figure that out? May 28 01:24:39 otherwise, I guess I'll just have to try May 28 01:26:11 Debian's minimum supported kernel is the last supported one on kernel.org at time of release, which is ~2.6.32 (Wheezy) May 28 01:26:40 that makes Squeeze the last version of Debian supported by Maemo kernel May 28 01:27:14 rikanee: hm, okay, thx. though squeeze doesn't have opus-tools, hm May 28 01:27:56 T_X1: you're just going to have to build it, I guess May 28 01:28:46 oki doki May 28 02:28:52 GogoPogo: you want to change your wlan mac get macchanger package. or cleven which gives you a gui to that package May 28 02:30:20 if you're pentesting and want to try throw the blue team off, be careful not to connect with your real mac first. ;) **** ENDING LOGGING AT Tue May 28 02:59:59 2013