**** BEGIN LOGGING AT Sat Jan 26 03:00:00 2013 Jan 26 08:57:39 romaxa: is there any documentation of embedLite API? or header files are supposed to be enough? Jan 26 10:15:10 calibrated the battery, 1420mAh for 10$ Jan 26 10:23:29 I'm thinking about writin small set of scripta (tm) allowing people to automagically migrate to ext4 (and back to ext3) for /home, if desired Jan 26 10:24:59 it would consist of small backupmenu fork (for would be standalone, as it would be useful for anyone on kernel-power, allowing better compression - lzma instead of gzip - bq4... charging, and so goes on), and scripts modifying mount-opts to utilize ext4 benefits during home mount Jan 26 10:25:41 ext4-migrant would depend on this backupmenu fork and some feature of kernel-power Jan 26 10:26:24 now, I don't want to depend on kernel power, as what I really need is ext4 support, by this or any other kernel. So, I think that depending on certain *feature* provided by kernel power, would be better suited Jan 26 10:26:47 in case of hypothetical other kernel providing ext4 support, it could just provide same feature Jan 26 10:27:08 now, my question is - does anyone see caveats in this approach, in this certain case? Jan 26 10:27:28 Estel_: base it on rescueos instead Jan 26 10:27:42 of course stuff will be highly experimental and at own risk, yet, I would like to minimise risk of bricking... Jan 26 10:27:52 kerio, resuceos won't be automagical ;) Jan 26 10:28:02 make it automagical Jan 26 10:28:04 and I planned to fork backupmenu for kernel-power users only, too Jan 26 10:28:24 i.e. utilizing mkfs.ext4 with custom options suited for our flash Jan 26 10:28:41 and few other improvements that, sadly or not, require kp Jan 26 10:28:58 i think that the kernel used by rescueos supports ext4 Jan 26 10:29:03 not sure, though Jan 26 10:29:08 BTW, if someone use automagic to migrate to ext4, i *want* him to have backupmenu backups :P Jan 26 10:29:27 and i'm not sure that the kp modules will fit in the initrd Jan 26 10:29:37 but, one need to have kp at normal runtime, why migrating to ext4 otherwise Jan 26 10:29:50 have you seen anyone running hildon day to day on rescueos kernel? :P Jan 26 10:30:07 Estel_: because bootmenu is a piece of crap Jan 26 10:30:22 my whole idea is about nullifing need to use external PC... Jan 26 10:30:29 or custom mounts Jan 26 10:30:34 indeed Jan 26 10:30:34 or this whole magic... Jan 26 10:30:52 that's why i have rescueos in my uboot bootmenu :) Jan 26 10:31:00 it's surprisingly useful Jan 26 10:31:23 lately I was migrating my device to another motherboard, and Exploiting backupmenu for that (vanilla device -> my custom partition layout and ext4 etc) was very conveinent way Jan 26 10:31:50 BTW, I even repartitioned new device via old device and hostmode Jan 26 10:31:57 yeah, it's pretty neat Jan 26 10:32:01 and easy debian and gparted Jan 26 10:32:05 anyway Jan 26 10:32:16 i deleted the swap partition on emmc and just repartitioned to make more room for optfs Jan 26 10:32:18 and then just restored Jan 26 10:32:32 whole concept is that user install ext4-migrant Jan 26 10:32:41 hehe, I did whole new partition table Jan 26 10:33:12 ...and ext4-migrant depends on backupmenu fork and feature of kernel-power that prvides ext4 Jan 26 10:34:02 ext4-migrant updates mount-opts, genfstab.awk, and whatsnot, to utilize ext4 and special capabilities that our mkfs will do later... Jan 26 10:34:23 Estel_: why do you still use the fstab generation? Jan 26 10:34:26 then, it (ext4-migrant) instruct user to reboot device with keyboard slided out Jan 26 10:34:36 because device doesn't boot without it? :P Jan 26 10:34:42 works fine for me Jan 26 10:34:57 i just put the fstab generation between "if false; then"/"fi" Jan 26 10:34:59 hm, if I leave genfstab with ext3 hardcoded there, my device wont boot Jan 26 10:35:06 so change it and then reboot Jan 26 10:35:22 well, no idea what it could break in long run Jan 26 10:35:23 and you also need to change the "home was mounted as readwrite" verification, i know Jan 26 10:35:38 ...and then cssu update comes and overwrite my changes Jan 26 10:35:38 but i reckon there should be a more agnostic way Jan 26 10:35:53 then, I spend 4564565 hours determining what got changed and where Jan 26 10:36:08 and redo my old changes, that I barely remember at this time Jan 26 10:36:08 Estel_: it's your fault if you modify system files without diverting them Jan 26 10:36:24 but apt-get will prompt when it's about to modify one of your modified configuration files Jan 26 10:36:35 so I won't modify them and won't have a problem - i don't need to get rid of genfstab for anything Jan 26 10:36:41 it's art for the sake of art Jan 26 10:37:17 ...that's why I don't modify them if it doesn't gain me anything, and just update genfstab with ext4 instead of ext3 :P Jan 26 10:37:37 easier to remember (and code into automagic scripts) than whole magic behind getting rid of fstab generation Jan 26 10:37:57 if it would give me any benefits, then, sure, but currently, there are none Jan 26 10:38:05 Estel_: so what happens when cssu changes genfstab.awk? Jan 26 10:38:33 I change it back - not ideal, but easier than changing bunch of other long scripts in many places Jan 26 10:38:46 i just put the fstab generation between "if false; then"/"fi" Jan 26 10:38:52 and you also need to change the "home was mounted as readwrite" verification, i know Jan 26 10:38:58 you have to do the last thing anyway Jan 26 10:39:10 and the first thing is in the same file Jan 26 10:39:14 replacing "ext3" with "ext4" sounds easier to remember about ;) Jan 26 10:39:19 in one short file Jan 26 10:39:57 well, unless it result in any benefits, one file to edit with one value < two files to edit with many values Jan 26 10:40:55 Estel_: it's *the same file* Jan 26 10:41:04 anyway, back to ext4-migrant idea, after it asks user to reboot with keyboard slided out, he have to... no problem if he forget, as he will be in reboot loop then, so he have unlimited number of chances... :P Jan 26 10:41:08 it's literally 10 lines below Jan 26 10:41:16 kerio, first thing, but second isn't Jan 26 10:41:31 the second will be unneeded Jan 26 10:41:43 after booting into backupmenu, simply creating backup of optfs and restoring it will do all magic Jan 26 10:42:05 as mkfs.ext4 with all parameteres tailored to serve our flash best are coded already Jan 26 10:42:35 after restoring it, user bootsup normally into ext4-home powered device, without messing with desktop or whatsnot Jan 26 10:43:07 benefits - performance gains due to useless (on flash) barriers enabled, and usage of experimental stripe-stride config Jan 26 10:43:43 Estel_: http://i.imgur.com/qa0oGpG.png Jan 26 10:43:50 it's **the same file** Jan 26 10:43:55 and cssu will never change /etc/fstab Jan 26 10:43:59 but it will change genfstab.awk Jan 26 10:44:19 and mount-opts Jan 26 10:47:53 gg no re Jan 26 10:54:15 so I've changed my apt sources to maemo-archive but extras doesn't sync and others report hash mismatches when installing Jan 26 10:54:34 weird Jan 26 10:55:02 ~mirrors Jan 26 10:55:03 extra, extra, read all about it, mirror is http://maemo-archive.wedrop.it/ http://maemo.merlin1991.at/apt-mirror/ http://talk.maemo.org/showthread.php?p=1315143#post1315143 Jan 26 10:55:18 iirc the skeiron.org mirrors work Jan 26 10:55:23 http://skeiron.org/repo/repository.maemo.org/ Jan 26 11:05:04 where's the images and flashers hanging about now? Jan 26 11:05:53 yes, skeiron works, others have hashsum error Jan 26 11:06:09 at least that's the rumor Jan 26 11:06:52 i had errors on skeiron with the fremantle/tools repo Jan 26 11:06:54 well, forget about low battery while bme disabled :P Jan 26 11:07:12 Estel_: you need the whole bme replacement pack Jan 26 11:07:24 will wait till applet is fixed Jan 26 11:07:26 Estel_: anyway, yours was clearly a ragequit, i'm sorry Jan 26 11:07:28 i win Jan 26 11:07:32 :P Jan 26 11:07:50 better repeat what you've said, as my last message was last I've seen Jan 26 11:08:00 as i was saying, http://i.imgur.com/qa0oGpG.png Jan 26 11:08:00 and wiki for irclog locations doesnt want to work Jan 26 11:08:03 it's the same file Jan 26 11:08:10 and cssu won't change /etc/fstab Jan 26 11:08:17 however, it will change genfstab.awk Jan 26 11:08:21 o.O Jan 26 11:08:25 /topic Jan 26 11:08:59 true, true, I just wanted to read logs before connecting Jan 26 11:09:11 hold on now, what's preventing me from downloading sygic for free? Jan 26 11:09:32 fact that modrana is way better? Jan 26 11:10:05 it reminds me to reach MartinK, as google changes something re overlay Jan 26 11:10:13 and overlay fail to work Jan 26 11:10:24 Estel_: monav has offline everything Jan 26 11:10:27 yahoo overlay is quite outdated, too Jan 26 11:10:42 sure, I'm not talking about routing Jan 26 11:11:13 I mean visible map, I use google satelite view + overlay as overlay, which result in streets transparently mapped on satelite view Jan 26 11:11:18 pester him to add vector rendering instead Jan 26 11:11:24 he is working on it Jan 26 11:11:48 personally, I preffer satelite view, but of course vector map have nice things to offer, too Jan 26 11:12:07 now I got gpg expired errors and gzip crashing :S Jan 26 11:12:17 ccxmo: which repo? Jan 26 11:13:04 and really, why does ovi have a store when you can just apt-get install the apps for free? :o Jan 26 11:13:38 not payware ones, methink? Jan 26 11:13:46 sygic is definetely payware Jan 26 11:13:49 no idea, never used or want to use ovi store Jan 26 11:13:53 unless it has its own internal method of payment Jan 26 11:14:12 strange, maybe someone who used ovi at least once could explain Jan 26 11:17:59 kerio: you got CSSU, eh? Jan 26 11:18:05 kerio: it was because of skeiron missing -testing apparently Jan 26 11:18:23 yes, we left testing out of skeiron Jan 26 11:19:03 since somebody convinced me it's useless since only a subclass of devel Jan 26 11:19:08 got backupmenu installed at last, thanks Jan 26 11:19:34 on hindsight we probably should have kept it Jan 26 11:19:37 * ccxmo goes to reboot Jan 26 11:26:08 oh right, the current testing has a fairly big problem Jan 26 11:26:23 it needs a package from extras-testing or extras-devel Jan 26 11:40:49 ~repos Jan 26 11:41:51 I got them Jan 26 11:42:26 Hi Jan 26 11:43:02 Can anybody tell me if the fremantle images will be available on the new server Jan 26 11:46:45 which images? Jan 26 11:48:17 the firmware images for n900 Pr1.3 Jan 26 11:48:35 DocScrutinizer05: do we have any "official" mirrors for tablets-dev? Jan 26 11:49:26 i hadn't downloaded a backup image before the servers went down Jan 26 11:49:34 :-( Jan 26 11:50:11 ~mirrors Jan 26 11:50:12 well, mirror is http://maemo-archive.wedrop.it/ http://maemo.merlin1991.at/apt-mirror/ http://talk.maemo.org/showthread.php?p=1315143#post1315143 Jan 26 11:50:23 no tablets-dev there :( Jan 26 11:50:27 Milhouse_: http://maemo.jacekowski.org/binary/firmware/ <-- N900 firmwares there Jan 26 11:50:42 kerio: the flasher and fiasco stuff? Jan 26 11:50:48 yep Jan 26 11:51:36 http://maemo.cloud-7.de/maemo5/fiasco+co/ but you have to "figure" the names of files in there, and if you dare to post the full URL anywhere, they will be gone same moment Jan 26 11:54:32 wow, thanks a lot, you're great Jan 26 11:54:58 LICENSE Jan 26 11:54:59 MD5SUMS Jan 26 11:55:00 RX-51_2009SE_10.2010.13-2.VANILLA_PR_EMMC_MR0_ARM.bin Jan 26 11:55:02 RX-51_2009SE_20.2010.36-2.002_PR_COMBINED_002_ARM.bin Jan 26 11:55:03 RX-51_2009SE_20.2010.36-2.003_PR_COMBINED_003_ARM.bin Jan 26 11:55:05 RX-51_2009SE_20.2010.36-2.004_PR_COMBINED_004_ARM.bin Jan 26 11:55:06 RX-51_2009SE_20.2010.36-2.203.1_PR_COMBINED_203_ARM.bin Jan 26 11:55:08 RX-51_2009SE_20.2010.36-2_PR_COMBINED_MR0_ARM.bin Jan 26 11:55:42 DocScrutinizer05: are those only the fiasco images? Jan 26 11:55:51 i wonder where i can get the os x flasher, at this point Jan 26 11:55:53 yoh Jan 26 11:56:00 also, for fuck's sake, why did nokia just drop that server? Jan 26 11:56:13 NFC, really Jan 26 11:56:33 is it dropped or just usual weekend fail? Jan 26 11:56:49 I'd hope the latter Jan 26 11:56:56 a DNS query might tell Jan 26 11:57:28 if the DNs vanished, we're fsckd Jan 26 11:57:36 mhm .. when will it back to normal? .. i dont need urgend upgrades so i didnt changed any config yet Jan 26 11:57:42 points to akamai Jan 26 11:58:05 and those akamai are still existent? Jan 26 11:58:27 I get two IPs Jan 26 11:58:31 I mean, hard to believe a whole serverfarm goes down by hw defect Jan 26 11:58:54 do they ping Jan 26 11:58:57 ? Jan 26 11:59:07 didn't check Jan 26 11:59:07 it was a small asteroid impact, nothing serious :) Jan 26 11:59:15 hehe Jan 26 11:59:37 was the name of the asteroid "elop"? (SCNR) Jan 26 12:00:02 yep, pings Jan 26 12:00:17 then probably nothing really bad happened yet Jan 26 12:00:29 they messed up, nothing unusual Jan 26 12:01:31 questuin is if anyone is around to fix it :) Jan 26 12:01:53 * DocScrutinizer05 idly wonders what's the "stage.maemo.org" for tablets-dev Jan 26 12:02:10 ShadowJK: TODAY? you're kidding Jan 26 12:03:24 no access to 'fiasco+co' Jan 26 12:05:06 forbidden Jan 26 12:05:17 toldya you need to "figure" the filenames, and DO NOT post the full URL *anywhere* Jan 26 12:06:19 Milhouse_: look, for your convenience I even disclosed what file I have on my private backup, some lines above Jan 26 12:06:33 now append those filenames, and you're done Jan 26 12:08:07 got it? Jan 26 12:09:17 ok thanks Jan 26 12:14:33 ups, sorry Jan 26 12:15:28 I think I'll wait for a couple if there'll be any official server. Nokia sucks! Jan 26 12:15:44 thank a lot for your support Jan 26 12:15:50 Elop sucks! Jan 26 12:16:41 worst service ever! Jan 26 12:20:21 maemo_flasher-3.5_2.4.5.3_beta.tar.gz Jan 26 12:20:33 I just added to my private archive Jan 26 12:25:23 how do I select gnu coreutils instead of busybox? Jan 26 12:25:35 hardly Jan 26 12:26:05 install them and you get them in bin/gnu/ (which you can add to your path, just before bin/) Jan 26 12:26:23 or invoke with a leading "g" Jan 26 12:26:31 so ls becomes gls Jan 26 12:26:38 find becomes gfind Jan 26 12:27:02 those are symlinks from bin/ to bin/gnu/ Jan 26 12:27:59 you have to pick your tools, find e.g. is in findutils pkg iirc Jan 26 12:28:14 maybe even findutils-gnu Jan 26 12:28:24 kthnx Jan 26 12:28:26 there are other *utils-gnu Jan 26 12:29:24 DocScrutinizer05, do you remeber problem with sudo gainroot? Jan 26 12:29:37 faintly Jan 26 12:29:49 gainroot is part of sudo package Jan 26 12:29:49 tracker-processes -r nukes tracker database, but who remember command to force it to start rebuilding? Jan 26 12:29:54 and I want to fix it in cssu Jan 26 12:30:13 is calling "su -" enought (instead /bin/sh)? Jan 26 12:30:16 after cssu update I *again* find myself with tracker ignoring flac and ogg files (not sure yet if it's really cssu fault) Jan 26 12:30:30 Pali: gainroot is part of sudo?? Jan 26 12:30:34 yes Jan 26 12:30:42 I'd think it's part of rootsh Jan 26 12:30:55 rootsh patching that gainroot Jan 26 12:31:04 M4rtinK, ping Jan 26 12:31:05 but script comes from sudo package Jan 26 12:31:08 :-S Jan 26 12:31:23 nasty shite Jan 26 12:31:23 I can fix both packages Jan 26 12:31:31 but we need working extras Jan 26 12:31:42 obviously Jan 26 12:31:46 I think that calling "sudo su -i" is ok Jan 26 12:31:50 Estel_: tracker-status ought to do it Jan 26 12:31:54 or something that calls tracker Jan 26 12:31:56 Estel_: ? Jan 26 12:31:57 better then "sudo sh" Jan 26 12:32:03 M4rtinK, I've been testing cause of some modrana problems I was facing lately, and it seems, that google changed something re their satelite and overlay Jan 26 12:32:15 as a result, it doesnt work in modrana anymore Jan 26 12:32:23 can we obliterate that piece of shit rootsh from extras please Jan 26 12:32:32 tiles show error with map service - change layout, try again later, etc Jan 26 12:32:32 sudser is also horrendously broken Jan 26 12:32:47 Pali: pretty please triple-check, since if that stuff breaks on vanilla systems, you fscked up epically Jan 26 12:32:51 kerio, are you goint it to fix? Jan 26 12:33:04 DocScrutinizer05, I know Jan 26 12:33:08 OTOH, with google overlay, it's strange, as it does work for very *far* view - aka countries - but refuses to work on close look Jan 26 12:33:21 and I'm pretty sure messybox su is fubar Jan 26 12:33:27 Pali: gainroot is to be used as the "official" way to get root Jan 26 12:33:31 DocScrutinizer05, when deb packages will be created, are you going to test if all working? Jan 26 12:33:35 aka, when rdmode is enabled Jan 26 12:33:37 (before pushing to cssu and extras?) Jan 26 12:33:44 sadlyp with google overlay, I'm not sure WTF, as modrana just stuck on "loading" instead of tile image, *not* downloaing and map service error Jan 26 12:33:54 yahoo overlay works ok, though Jan 26 12:34:07 so I would say it's about google, not modrana staight bug Jan 26 12:34:08 the whole concept of rootsh is broken, because gainroot is broken, and gainroot has to be like it is for histerical raisins Jan 26 12:34:14 Pali: might do, if I find the time Jan 26 12:34:34 Pali: my systems are pretty non-vanilla thhough Jan 26 12:34:51 DocScrutinizer05, after merlin1991 create sudo repo on gitorious you can look at patch Jan 26 12:35:26 looking at patch doesn't help much here, since that's related to messybox broken su Jan 26 12:35:32 M4rtinK, if I can help by providing some more detailed info, just ask Jan 26 12:35:44 it might look perfect, but fails with messybox Jan 26 12:35:50 Estel_: Google Overlay seems to work for me & Satellite seems broken Jan 26 12:35:51 DocScrutinizer05: i'm not entirely sure that messybox su is *that* wrong Jan 26 12:35:57 it's just not suid root, so it can't work Jan 26 12:35:58 it *is* Jan 26 12:36:05 yes Jan 26 12:36:10 so run it as root Jan 26 12:36:10 M4rtinK, hm, for me, overlay works only from local storage Jan 26 12:36:22 and it MUST NOT get suid *ever* Jan 26 12:36:27 "sudo su" doesn't execute root's .profile but at least it sets the correct envvars Jan 26 12:36:28 Estel_: oh, taht can be it Jan 26 12:36:33 where I dont have tiles, it's stuck on "loading" instead of downloading Jan 26 12:36:45 "sudo su -l" gives me a login shell Jan 26 12:36:49 much like "sudo -i" Jan 26 12:36:53 kerio, gainroot script script checkig if you have R&D enabled and if yes, then it call /bin/sh Jan 26 12:36:59 is there something like factory-wipe so I can remove previous user's setup? Jan 26 12:36:59 yea, but I'm curious why it doesn't act like satelite, i.e. trying to download and failing... satelite does it, overlay seems stuck on loading? Jan 26 12:37:11 if you not call gainroot with sudo you will ged R&D is disabled Jan 26 12:37:12 Pali: change it to run su -l Jan 26 12:37:17 CcxCZ_, ~reflash Jan 26 12:37:26 what is difference between su - and su -l? Jan 26 12:37:29 Estel_: nope, overly works even for an unknown area Jan 26 12:37:33 *overlay Jan 26 12:37:40 Pali: oh, the same Jan 26 12:37:58 flashing guide anywhere? Jan 26 12:37:58 M4rtinK, strange, maybe it's me about overlay then. But I'm surprised, as everything else works Jan 26 12:38:25 anyway, we can confirm that satelite got broken, at least, yep? Jan 26 12:38:27 Pali: which problem are you fixing with this patch? Jan 26 12:38:31 Estel_: they periodically chnage the URLS - there is a numbered component in each of them Jan 26 12:38:52 yea, I'm aware of that, very irritating Jan 26 12:39:07 Estel_: it is probably to track which service is using which and when it started using it Jan 26 12:39:10 btw, before next version, is there any file user can edit by hand as pathwork fix? Jan 26 12:39:20 DocScrutinizer05, correct $HOME correct $USER Jan 26 12:39:29 Estel_: so they might be phasing out some of the older URLs Jan 26 12:39:36 is gainroot supposed to give you a login shell, though? Jan 26 12:39:37 ~reflash Jan 26 12:39:37 extra, extra, read all about it, reflash is zImage and/or initrd.bin on CF and press C+D+Reset (collie) OR updater.sh, zimage.bin and/or initrd.bin on CF/SD and press OK while rebooting, then option 4, then CD or SF, then HAI (yes), then wait and cross fingers (all other models) Jan 26 12:39:40 DocScrutinizer05, try to run: sudo gainroot and sudo -i and sudo su Jan 26 12:39:46 wut? Jan 26 12:39:48 Estel_: .modrana/map_config.conf Jan 26 12:39:48 you will see different envs Jan 26 12:39:54 infobot, that wasnt helpful Jan 26 12:39:57 M4rtinK, lol Jan 26 12:40:07 I get correct $HOME and $USER with command `root`. Command sudo gainroot is NOT meant to have root envs Jan 26 12:40:19 haven't expected it's there :) btw, what is best way to determine correct address now? Jan 26 12:40:26 and I absolutely deprecate changing semantics here Jan 26 12:40:48 Estel_: just note that the file will be replaced once a new "official" one is available & the old one should get renamed Jan 26 12:40:57 * Estel_ nods Jan 26 12:41:09 you even MUST NOT change semantics of sudo gainroot Jan 26 12:41:12 ~2119 Jan 26 12:41:13 Estel_: search -> Address online ? Jan 26 12:41:13 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119. Jan 26 12:41:17 Pali, I agree with DocScrutinizer05 here, unless there is an issue unknown to me, that you're fixing by it Jan 26 12:41:19 DocScrutinizer05: rootsh already changes them Jan 26 12:41:36 M4rtinK, no, I havent thought about "that" address :P Jan 26 12:41:37 kerio: BS Jan 26 12:41:45 I mean what to put into config Jan 26 12:41:50 to make satelite working again Jan 26 12:42:01 i.e. how to best determine correct new address they're using Jan 26 12:42:07 sudo gainroot == su; root == su - Jan 26 12:42:10 period Jan 26 12:42:18 sorry, I'm kinda green when it comes to google ever-changing services Jan 26 12:42:26 Estel_: just open google maps, switch to satellite view & check what urls are the tiles using Jan 26 12:42:32 gainroot, ran as root, is supposed to run sh as root if rd-mode is enabled, and it's supposed to be runnable via sudo without being asked a password Jan 26 12:42:35 DocScrutinizer05, no Jan 26 12:42:42 sudo gainroot doing sudo sh Jan 26 12:42:44 M4rtinK, ah, it's easy as that Jan 26 12:42:47 OK, thanks Jan 26 12:42:50 Estel_: yep Jan 26 12:42:57 and calling sudo sh is *bad* idea Jan 26 12:42:59 source: the "sudo" package, as it appears in stock fremantle Jan 26 12:43:08 I'll play a little more with overlay, and report back if problem persist, with some debug output Jan 26 12:43:19 Pali: I done't give a ***** what it *executes*, i talk about what it *does* Jan 26 12:43:22 Estel_: there is also another possibility - they might not like the modRana user agent Jan 26 12:43:29 Estel_: thanks ! Jan 26 12:43:45 DocScrutinizer05, so then sudo gainroot does *not* do su Jan 26 12:43:54 yes it does Jan 26 12:43:55 M4rtinK, user agent, huh? Jan 26 12:43:59 DocScrutinizer05, no Jan 26 12:44:03 in such case it's also configurable? Jan 26 12:44:03 yes Jan 26 12:44:06 sudo su != sudo sh Jan 26 12:44:15 dafaq Jan 26 12:44:17 DocScrutinizer05: "su" will still set $HOME and $USER Jan 26 12:44:20 sudo gainroot won't Jan 26 12:44:27 NO IT DOESNT Jan 26 12:44:40 sudo will not set envs correcty Jan 26 12:44:42 kerio: get your shit sorted Jan 26 12:44:43 su will do it Jan 26 12:44:55 Estel_: hmm, I think not yet Jan 26 12:45:15 DocScrutinizer05: my shit is properly sorted and packed Jan 26 12:45:26 open any proper unix system, type su and then echo $HOME Jan 26 12:45:34 M4rtinK, ok, we will worry about it if they block (unless you want to be prepared against that offence to freeedom :P if it's not much work to add it into new version) Jan 26 12:45:35 Estel_: but I think it should be added as an option for enterprising users :) Jan 26 12:45:46 :) Jan 26 12:46:00 Estel_: yeah, fixing the URLs comes first Jan 26 12:46:06 DocScrutinizer05, then your version of "su" is incorrect Jan 26 12:46:26 damn, my version of su on PC is incorrect Jan 26 12:46:48 call su and then call echo $HOME Jan 26 12:46:59 so are my statements above, but that doesn't change anything regarding NOt TO CHANGE SEMANTICS of sudo gainroot Jan 26 12:47:05 do it on n900, call sudo gainroot and call echo $HOME Jan 26 12:47:08 why not? Jan 26 12:47:43 so why we should have broken su in our devices?? Jan 26 12:47:58 because we always had Jan 26 12:48:02 indeed Jan 26 12:48:03 because nokia done it? Jan 26 12:48:08 kerio, Pali, quick question. Will fixing it broke many packages or scripts? Jan 26 12:48:15 if not, lets fix it Jan 26 12:48:20 and screw what nokia did Jan 26 12:48:24 we have NO version of su on our N900 Jan 26 12:48:30 if it is riskyp then think again Jan 26 12:48:31 Estel_, sudo gainroot is for R&D mode Jan 26 12:48:37 we have sudo gainroot and root Jan 26 12:48:43 I do not know how if any app depends on R&D mode Jan 26 12:48:52 Estel_: any package that relies on rootsh doesn't deserve its space in bits Jan 26 12:48:58 Pali, sure, I'm just asking if program or scripts relying on our current implementation of root or gainroot will get broken Jan 26 12:49:00 and that has a well known semantics that many packages rely on Jan 26 12:49:13 Estel_: qtmobilehotspot depends on gainroot in a bad way Jan 26 12:49:18 you don't want to 2fix" dat shite for shits'n'giggles Jan 26 12:49:22 indeed Jan 26 12:49:29 Estel_, I did not see any app which depends on gainroot Jan 26 12:49:36 Pali: qtmobilehotspot Jan 26 12:49:39 pffff Jan 26 12:49:41 and any package rdepending on rootsh Jan 26 12:49:44 so until you can get in touch of maintainers of such widely used things as qtmobilehotspot (and others), and ask them to fix... Jan 26 12:49:49 or fix them yourself Jan 26 12:49:51 kerio, really? and it checking $HOME?? Jan 26 12:49:54 please don't touch it :) Jan 26 12:49:59 Pali: i don't know Jan 26 12:50:03 i'm staying the hell away from it Jan 26 12:50:06 sorry but this is stupid Jan 26 12:50:23 why some application running as root needs HOME to be set to user? Jan 26 12:50:25 Pali: just configure sudo properly Jan 26 12:50:34 ok, while I agree with You Pali that bad tings should be fixed, not kept due to habit, this is special situation Jan 26 12:50:34 no, messing with such silly "fixes" is stupid Jan 26 12:50:35 Pali: maybe PATH Jan 26 12:50:40 maybe it relies on the current working directory Jan 26 12:50:54 maybe it relies on some kind of freaky job control that fails with a proper login Jan 26 12:50:54 since always we had it, and many things may get broken if you change it just for the sake Jan 26 12:51:07 ship a new rootsh-pali-finally-correct package, and do in it whatever you want Jan 26 12:51:10 are you willing to inspect every package in extras for any mention of gainroot? Jan 26 12:51:19 DO NOT change semantics of rootsh Jan 26 12:51:28 so unless we can fix all things that rely on it, by hand, or perform some "compatibility layer" that will allow them to work... Jan 26 12:51:36 changing this piece of bit is little risky Jan 26 12:51:41 kerio, sudo gainroot changing directory to /root/ Jan 26 12:51:48 su chaning it to /root/ too Jan 26 12:51:53 btw, rootsh is a package that violates about every kind of packaging guideline Jan 26 12:52:07 keri, so go and fix it as soon as repos are online Jan 26 12:52:18 it's still single most popular package in maemo Jan 26 12:52:18 Estel_: the only fix is to delete it and pretend it never existed Jan 26 12:52:35 kerio: we simply don't care since it ever did and we can't change the past Jan 26 12:52:37 just take into account that I'm asking seriously. Jan 26 12:52:48 either we fix it on big scale Jan 26 12:53:02 or make sad face and accept that fixing it now would cause fubar mess Jan 26 12:53:18 I have to agree with Estel_ Jan 26 12:53:23 I have no idea how to fix it on big scale, if someone have - by some mad in the middle script or whatsnot - lets talk about it Jan 26 12:53:39 hrmpf, i suppose that it could at least avoid the packaging issue Jan 26 12:53:55 I'm all for fixing nonsense made by nokia or early maemo adopters, but only if we wont break half or 1/4 of repo content Jan 26 12:54:32 particularly since I don't see a single valid bug ticket Jan 26 12:54:34 and no matter if we like it or not, rootsh is still more popular than anything else, including cssu ;) Jan 26 12:55:07 can you do "alias su=gainroot" and "alias 'su -'='root'" somewhere? ;-) Jan 26 12:55:13 well, here we a little disagree, as I would have nothing against fixing it just for sake of future less-mess, but only id it wouldn't hurt (too bad) Jan 26 12:55:15 ShadowJK: different syntax Jan 26 12:55:38 * ShadowJK doesn't even see a problem keeping it the way it is Jan 26 12:55:45 but now it seems to be hurting badly Jan 26 12:55:55 it is future-maximum-mess Jan 26 12:55:56 so there are 3 of us already :P Jan 26 12:55:57 ShadowJK: you have packages doing stuff like echo 'command' | root Jan 26 12:55:57 it's not like we have working pam either Jan 26 12:56:03 su != gainroot --> which should be fixed Jan 26 12:56:04 DocScrutinizer05, right, currently, yes Jan 26 12:56:06 that shit is fucked up, yo Jan 26 12:56:13 thats why i said "if it wouldn't hurt (too bad)" Jan 26 12:56:25 Pali: says who? Jan 26 12:56:43 replace should with "could" Jan 26 12:57:02 it could be fixed and I would say yes, eagerly, if it wouldn't bring risk of fubar'ing half of repo Jan 26 12:57:34 Let's do it the Sun way, create a /usr/gnu/ prefix for gnulike replacements, then you can choose maemo standard or gnu standard at will Jan 26 12:57:34 it definitely WILL fubar stuff Jan 26 12:57:52 qtmobilehotspot wasn't written by idiot, and if they did it like they did - ok, I agree that they should use other way, but I can only imagine that many others followed same scheme Jan 26 12:57:56 Estel_: fwiw, a package that breaks because you change something in rootsh's gainroot means that it was doing something seriously wrong, and it should be fixed Jan 26 12:58:05 and the developer should be fixed too Jan 26 12:58:11 so we reduce the damage to the genetic pool Jan 26 12:58:28 kerio, sure, but you should tell them when they released Jan 26 12:58:36 not a year after they're not contactable Jan 26 12:58:55 and the one breaking rootsh on purpose to force others to fix their stupid progs will get shot Jan 26 12:59:00 I know rootsh is doing it the wrong way, but, it's special case due to it being legacy of early maemo adopters Jan 26 12:59:01 I promise that Jan 26 12:59:19 meh, at least we should fix rootsh's packaging Jan 26 12:59:21 note that rootsh is 3rd application from extras which provide root binary, so each package which depends on it is in maemo package interface Jan 26 12:59:38 we can look at it and find if something could break Jan 26 12:59:41 kerio, we may not like it, but it's like trashing half of the city, because socialist's architects did it wrong way Jan 26 12:59:48 and leaving people homeless, for greater good Jan 26 13:00:06 Pali: go ahead Jan 26 13:00:13 http://maemo.org/packages/view/rootsh Jan 26 13:00:20 in Poland we have many examples of sad architecture, but nuking it all at once isn't good fix Jan 26 13:01:08 when you're done with finding all pkgs using rootsh in *-install or anyewhere in program, maybe even without proper DEPENDS in pks, when you got all fixed and tested and claimed maintainership, than come back here and report Jan 26 13:01:57 Pali: honestly, it's a silly idea Jan 26 13:03:01 Pali: ship a better-rootsh pkg, deprecate rootsh pkg, in 2 or 3 years you can start to fade it out. That's the way such stuff gets handled Jan 26 13:05:21 why not ship a proper su-package? Jan 26 13:06:28 su is borked just because silly messybox. Ship a proper stand-alone su binary and fine you are Jan 26 13:06:57 how does one apply for maintainership of a package? Jan 26 13:07:02 will conflict with messybox, but so what? Jan 26 13:07:08 kerio, you get logged in Jan 26 13:07:16 and from package page, you apply Jan 26 13:07:27 you can send new version to devel without being maintainer, though Jan 26 13:07:31 one does not "conflict" with messybox Jan 26 13:07:44 replace or divert or change messybox and update both at the same time Jan 26 13:08:00 Pali, in fact after shipping proper rootsh package you could start fading it away in 6 months :P Jan 26 13:08:01 hm, would it be a stretch for the busybox package to include a proper su binary? Jan 26 13:08:07 as things are going faster in Maemo Jan 26 13:08:20 kerio, busybox power is the answer Jan 26 13:08:26 nope Jan 26 13:08:33 busybox-power's su is also broken Jan 26 13:08:37 but then it would need to get into cssu without idiotic requiments to split it into two packages Jan 26 13:08:42 but it can get fixed ,) Jan 26 13:08:53 because you'd have to make busybox suid Jan 26 13:08:55 no, seriously Jan 26 13:09:00 which is a Bad Idea™ Jan 26 13:09:01 Pali's idea is good Jan 26 13:09:04 but not suddenly Jan 26 13:09:10 no, seriously, su needs to be a separate binary Jan 26 13:09:18 first, lets fix broken packages, starting with rootsh Jan 26 13:09:34 M4rtinK, sorry for pestering, but whats difference between: Jan 26 13:09:44 kh/v=107 Jan 26 13:09:48 and question again: why we should install another binary to n900, and have also old bad binary? Jan 26 13:09:57 and vt/lyrs=s Jan 26 13:10:03 suffix of address? Jan 26 13:10:11 I see old value commented out in .map-config Jan 26 13:10:11 why should I have installed both good and bad binary in my n900? Jan 26 13:10:29 new calue, if satelite map is open from browser, have vt/lyrs= suffix Jan 26 13:10:49 but, it should be updated too, or kept kh/v=107? Jan 26 13:11:19 M4rtinK, ^. Again, sorry for pestering, I just don't have idea what it uses and if mimicing browser view entirely is good idea Jan 26 13:11:46 I for the life of mine don't see the root problem (pun intended), use `root` to get proper root env and perms, use `sudo gainroot` to get root without changing environment Jan 26 13:11:54 Pali: the issue is that /bin/su MUST be suid root Jan 26 13:11:58 to work properly Jan 26 13:12:04 busybox works by having one binary and a bunch of symlinks Jan 26 13:12:14 symlinks can't be made suid root Jan 26 13:12:23 and you MUST NOT make /bin/busybox suid root Jan 26 13:12:34 kerio, I know and this is reason why I want to call /bin/su from gainroot called via sudo Jan 26 13:12:50 this will work because su will be called as root Jan 26 13:12:54 (I tested it) Jan 26 13:12:56 WTF we *must* change semantics of `sudo gainroot`??? Jan 26 13:12:58 Pali: sudo gainroot --use-su Jan 26 13:13:01 isn't it achievable by not breaking "root" and "gainroot" as used by rootsh? Jan 26 13:13:13 kerio, that working only if you have installed rootsh package Jan 26 13:13:16 which is what is ran by `root` Jan 26 13:13:23 ok then Jan 26 13:13:25 sudo gainroot Jan 26 13:13:29 and then type su - Jan 26 13:13:32 gainroot comes from sudo package from SSU Jan 26 13:13:43 indeed Jan 26 13:13:48 root and patched gainroot comes from extras package rootsh Jan 26 13:13:50 and that one doesn't have --use-su Jan 26 13:14:26 calling su that doesn't do any su is a meaningless execise Jan 26 13:14:36 but if i had to guess, i'd say that that separate option was added *because* they didn't want to break the existing semantics Jan 26 13:14:48 DocScrutinizer05: su - makes a login shell of some sorts Jan 26 13:14:57 DocScrutinizer05, no, it create *proper* shell Jan 26 13:14:58 it's probably the same as typing login, or something akin to that Jan 26 13:15:06 Estel_: I mostly just try to discard all components from the URL as long I can get a working link to the tile image Jan 26 13:15:22 so cut-and-check, until it dies? Jan 26 13:15:27 it should be the same as running login -f root Jan 26 13:15:31 then revert to shortest working link? Jan 26 13:15:55 Estel_: hmm, now when I think about it, Google might then think about the URL as less trustworthy :) Jan 26 13:16:05 Estel_: yeah, basically Jan 26 13:17:53 M4rtinK, holy shit, we need open soource satelite swarm Jan 26 13:18:01 this google thing is way to messy Jan 26 13:18:14 Pali: you ever thought that maybe those who use `sudo gainroot` do *not* *wnat* a proper shell? Jan 26 13:18:28 IIRC there are even some licence compatible aerial data sets Jan 26 13:18:47 no, because sudo gainroot shoudl do proper shell Jan 26 13:18:51 but no one is currently hosting them in an usable form Jan 26 13:18:51 also it write "Root shell enabled" Jan 26 13:18:58 Pali: says who? Jan 26 13:19:10 DocScrutinizer05, sudo gainroot say it Jan 26 13:19:24 DocScrutinizer05, then remove bad message from gainroot Jan 26 13:19:24 Estel_: see: http://openaerialmap.org/Main_Page Jan 26 13:19:33 it not doing root shell Jan 26 13:19:40 Pali: just because _you_ _think_ it 2should do a proper shell" is no argument to patch mess into mess Jan 26 13:19:42 Estel_: currently offline Jan 26 13:20:00 DocScrutinizer05, look at what sudo gainroot writing and what doing Jan 26 13:20:00 Pali: no, I want that messy message Jan 26 13:20:10 M4rtinK, I see... Jan 26 13:20:14 it is not root shell what gainroot doing Jan 26 13:20:31 remove that message or fix sudo gainroot to do what it prints Jan 26 13:20:36 I don't give a shit what it's writing, I care what it did so far, and I expect it continues to do exactly that Jan 26 13:20:48 well, Pali have a point here Jan 26 13:21:05 and I expect that some application will do what writing Jan 26 13:21:09 I don't see reason why we can't fix that simple message in rootsh, and it's packaging problem at the same time Jan 26 13:21:11 Estel_: then he should remove that message Jan 26 13:21:17 * Estel_ nods Jan 26 13:21:35 that message is part of the semantics Jan 26 13:21:43 while i can live with this idiotic message, if cosmetic is what he want to spend time on, why not. all for the better Jan 26 13:21:43 or instead that message it should write: this is not proper root shell Jan 26 13:22:01 this is bug: application not doing what printing to user Jan 26 13:22:04 Estel_:hmm, changing v=107 to v=124 might be enough Jan 26 13:22:11 as long as it doesnt break half of repos I'm all for fixes :) Jan 26 13:22:29 how can you be sure that there's no program that, say, expects that line to send commands? Jan 26 13:22:33 M4rtinK, hm, thanks, I already did vt/lyrs=s, will see what it ends up like Jan 26 13:22:34 BWAHAHAHAAAAHAA "i'm lieing, *always*" Jan 26 13:22:50 Estel_: but it is quite hard to test if google already autobanned you for downloading to many tiles (happens all the time) Jan 26 13:23:02 M4rtinK, lol, true Jan 26 13:23:07 and really I do not have any idea how can changing sudo sh --> to sudo su break application Jan 26 13:23:15 sudo su --> sudo sh is possible Jan 26 13:23:26 if yahoo sat wouldnt be too outdated, I wouldn't gie a rat's shit about google sat view Jan 26 13:23:27 but our direction not Jan 26 13:23:29 Estel_: an interesting experiemtn might be to leve also some of the other options in the url, just to see it makes it more trustworthy :) Jan 26 13:23:35 or if we would have open aerial map ;) Jan 26 13:23:48 why not, I can testdrive it Jan 26 13:23:54 Estel_: yeah, Google is unfortunately the only globally usable Jan 26 13:24:01 I can see couple of reason why some applications could be broken with current situation Jan 26 13:24:36 Estel_: virtual earth & yahoo are mostly unusable outside outside from the most major areas Jan 26 13:24:38 Pali: don't start talking silly! You said your patch is meant to change $HOME now, while formerly `sudo gainroot` doesn't. That sure as hell will break things Jan 26 13:24:41 Pali, maybe cssu-devel is good place to test it? Jan 26 13:24:58 cssu-devel users could then check if, for example, qtmobilehotspot get hit or not Jan 26 13:24:59 if you call sudo gainroot for gettin "root shell" and then you will call some application which needs to be started as root and that application check "$USER" it will fail Jan 26 13:25:09 M4rtinK, yea Jan 26 13:25:30 Pali: $USER is not to be relied upon Jan 26 13:25:42 Pali: that's exactly what it's meant to be Jan 26 13:25:54 and you MUST NOT change that Jan 26 13:26:08 can I use debian repos to install software missing from the maemo ones? Jan 26 13:26:20 DocScrutinizer05, without my patch some application started from sudo gainroot will not work Jan 26 13:26:30 so what? Jan 26 13:26:32 CcxCZ_, sure, but most of it wont work ;) Jan 26 13:26:49 it never did Jan 26 13:27:13 in root shell you should have correct $HOME and $USER and if not message "root shell" should not be shown to user Jan 26 13:27:24 but with your patch an app that formerly did a wget to current directory will now wget to root's $home Jan 26 13:27:38 and gainroot priting message that this *is* root shell - which is not truth Jan 26 13:27:40 M4rtinK, strange, now my modrana stopped working alltogether Jan 26 13:27:43 i.e. it loads Jan 26 13:27:50 but it doesn't see any layers, at all Jan 26 13:27:56 as available Jan 26 13:27:57 DocScrutinizer05, sudo gainroot chaning PWD to /root Jan 26 13:28:00 only "empty layer" Jan 26 13:28:05 su doing it too Jan 26 13:28:07 Pali: again, and last time, if that message bugs you so much, remove it! Jan 26 13:28:16 config file looks ok... well, maybe I deleted some " unnecessary Jan 26 13:28:58 yea, silly me, it was just typo Jan 26 13:29:05 DocScrutinizer05, you did not show me any possible situation when some application could be broken after change Jan 26 13:29:12 sorry, I'm quite new to mess with modrana internals Jan 26 13:29:54 and I see couple of situation when root applications will not work in shell spawned by gainroot Jan 26 13:30:56 ~ $ id Jan 26 13:30:57 uid=29999(user) gid=29999(users) Jan 26 13:30:58 ~ $ echo $HOME Jan 26 13:31:00 /home/user Jan 26 13:31:01 ~ $ pwd Jan 26 13:31:03 /home/user Jan 26 13:31:04 ~ $ sudo gainroot Jan 26 13:31:06 Password: Jan 26 13:31:08 Root shell enabled Jan 26 13:31:09 BusyBox v1.10.2 (Debian 3:1.10.2.legal-1osso31+0cssu0) built-in shell (ash) Jan 26 13:31:10 Enter 'help' for a list of built-in commands. Jan 26 13:31:12 /home/user # echo $HOME Jan 26 13:31:13 /home/user Jan 26 13:31:15 /home/user # pwd Jan 26 13:31:16 /home/user Jan 26 13:31:20 dat paste Jan 26 13:31:56 since I'm fed up with it Jan 26 13:32:18 [2013-01-26 14:27:57] DocScrutinizer05, sudo gainroot chaning PWD to /root Jan 26 13:32:36 Pali: NO it's NOT Jan 26 13:32:56 and you'll BREAK THINGS if you change that Jan 26 13:33:14 M4rtinK, I can confirm, that google overlay sa Jan 26 13:33:23 started working for me only *after* changing it Jan 26 13:33:34 from "mt1" to "mts1" Jan 26 13:33:40 in map_config file Jan 26 13:34:04 mts1.google.com seems to be required one, currently. No idea how it could work for you with old address Jan 26 13:34:18 Estel_: ok, noted :) Jan 26 13:35:25 Estel_: maybe also try some geographically distinct areas ? Jan 26 13:35:33 tried that Jan 26 13:35:42 I'm sure i was never in africa from modrana Jan 26 13:35:50 at zoom 16 :P Jan 26 13:35:53 Estel_: hh Jan 26 13:37:46 M4rtinK, depending on zoom, address change from mts1.google.com, to mts0.google.com (for closer looks) Jan 26 13:37:51 isnt it problem? Jan 26 13:40:33 (([2013-01-26 14:29:55] and I see couple of situation when root applications will not work in shell spawned by gainroot)) then don't use sudo gainroot for those apps. It's not meant to get used this way! And evidently nobody did so far, since... it doen't work. So WTF *is* the bug you're fixing? Jan 26 13:40:34 Estel_: mts0 & mts1 are IIRC are just mirrors, that enable the javascript in the web map to download multiple times at once Jan 26 13:41:11 ah, i though its related to new "angled view" at close zoom Jan 26 13:41:25 sudo gainroot is only way how to get root without any 3rd application Jan 26 13:41:32 i.e. fact, that, for web page, it starts showing photos took from angle, not high above, for close zooms Jan 26 13:41:49 and it not spawn proper root shell, so some root applications will not work here Jan 26 13:41:55 interesingly, some other programs still show "straight-down" view for close zooms Jan 26 13:42:12 Estel_: "Generally several subdomains (server names) are provided to get around browser limitations on the number of simultaneous HTTP connections to each host. Browser-based applications can thus request multiple tiles from multiple subdomains faster than from one subdomain." Jan 26 13:42:29 hm, can modrana exploit it too? Jan 26 13:42:31 Estel_: from http://wiki.openstreetmap.org/wiki/Slippy_map_tilenames#Tile_servers Jan 26 13:42:47 i.e. use mirrors, to go around risk of autoban? Jan 26 13:43:08 instead of pulling dozens from one mirror, pull some from mts0, 1, 2, 3... Jan 26 13:43:09 modRana already downloads multiple tiles at once Jan 26 13:43:26 the question is if it makes a difference Jan 26 13:43:30 but from same mirror? or not Jan 26 13:43:42 well, if my browser doesnt get banned so quickly using mirrors... Jan 26 13:43:47 if the mirrors share accounting Jan 26 13:43:48 even if I hover like mad Jan 26 13:44:03 yeah, good point Jan 26 13:44:05 but maybe google uses it to pinpoint "illegal" apps? Jan 26 13:44:12 might be also sue to user-agent though Jan 26 13:44:18 if something download many tiles at once from one mirror, it's suspicious Jan 26 13:44:27 and most likely, app Jan 26 13:44:34 well, we can use chrome user agent :P Jan 26 13:45:04 I always wondered ahy I can, very quickly, hover above whole city in close zoom and download hundreds of tiles in browser... Jan 26 13:45:10 but in modrana, I get banned quickly Jan 26 13:45:29 maybe user customizable user agents + using mirrors is a way to go to avoid limits Jan 26 13:45:46 after all, everyone is free to user any useragent she(e want, it isn't patented (yet) Jan 26 13:46:18 and mirrors are publicly available. +, it's user responsibility, is he uses google sat map outside eula... BTW, in Pland, google eula for maps isn't ever legal Jan 26 13:46:46 Estel_: most EULAs aren't :) Jan 26 13:46:58 Yes, because going around the technological limitations that people set up so you wouldn't abuse them is so "ethical". Jan 26 13:47:05 :) Jan 26 13:47:29 my ethic doesn hurt when i use google map from fast modrana, instead from clunky user interface Jan 26 13:47:34 of browser Jan 26 13:47:34 you can experiment with the user-agent here: https://github.com/M4rtinK/modrana/blob/master/modules/mod_mapTiles.py#L194 Jan 26 13:47:44 thanks :) Jan 26 13:48:05 This is pretty much like the idiots who thought they could use our API for free. We had setup a reasonable limit of x API calls per day, per IP. Jan 26 13:48:08 teotwaki_, it isn't about abusing, it's about forcing to use browser over app, to get same results Jan 26 13:48:30 well, it's not api, that we're talking about Jan 26 13:48:48 Some company tried to use our API for profit, but didn't want to pay, so they started setting up multiple servers they could use our API from. Jan 26 13:49:02 anyway, call me evil, but I won't shed a single tear about google being not able to force browser usage on me Jan 26 13:49:09 for getting same results Jan 26 13:49:18 (to get around the API calls limit) Jan 26 13:49:41 well, you could write more security... or go FOSS way :) Jan 26 13:49:53 Estel_: or better here: https://github.com/M4rtinK/modrana/blob/master/core/configs.py#L220 Jan 26 13:50:00 ok... Jan 26 13:50:07 They were very quick to subscribe and use a single IP when I added a data randomizer for all of their requests :) Jan 26 13:50:58 sounds fun. Problem with google is that they don't give a shit about ammount of data used Jan 26 13:51:09 modRana is not a commercial application & users are free to use it how they like :) Jan 26 13:51:09 they just want to enforce to know everything about user using it Jan 26 13:51:27 Estel_: we are FOSS. Our whole API software is dual licensed GPL, with the libraries under LGPL :) Jan 26 13:51:53 M4rtinK, excuse my noobines, but this is drop-in replacement for python file, or I have to do something else before using it? Jan 26 13:52:04 teotwaki_, then why on earth they tried to abuse? Jan 26 13:52:11 BTW, that is why OSM based routing is so important Jan 26 13:52:19 yea Jan 26 13:52:27 as Google can log your routing requests Jan 26 13:52:27 Estel_: because they're idiots Jan 26 13:52:40 Estel_: because the API is FOSS, our infrastructure that gives you access to placing millions of phone calls isn't :) Jan 26 13:52:41 still it is much harder from modRana than from a browser Jan 26 13:53:09 teotwaki_, well, but api allowed to do some free requests to everyone, yep? Jan 26 13:53:20 Estel_: just edit /opt/modrana/core/config.py :) Jan 26 13:53:23 Everyone with an account, yes. Jan 26 13:53:35 M4rtinK, thats what I thought, just wanted to be sure Jan 26 13:54:09 teotwaki_, then hunting abusers is part of fun Jan 26 13:54:23 Oh it definitely is Jan 26 13:54:52 We operate call centre solutions, recently, I've been tasked to down call centre agents who place personal calls on their work time :) Jan 26 13:55:09 but i hope you see distinct difference between modrana accessing data exactly the same way any browser is able too (you can even call modrana browser with strange features, like voice nav) and your case Jan 26 13:55:17 Estel_: I do, no worries :) Jan 26 13:55:31 it's me ...or talk.m.o is progressively slowing down? Jan 26 13:55:38 track down** Jan 26 13:55:46 I don't "down" agents, whatever that may mean. Jan 26 13:55:49 xes, everything related to maemo infra is, one way or another Jan 26 13:57:00 teotwaki_, also teach them than "reboot your system" isn't always best first part of conversation with client Jan 26 13:57:05 or reboot your router Jan 26 13:57:16 I don't talk to the agents :) Jan 26 13:57:22 just joking ;) Jan 26 13:57:27 I really hope the Fremantle OBS Pali usually mentions gets resurrected Jan 26 13:57:44 its dead? Jan 26 13:57:46 it works really nice for Harmattan & Mer/Nemo Jan 26 13:57:57 I thought it is what maemo repos was heading into Jan 26 13:58:06 I don't meen the MeeGo COBS Jan 26 13:58:11 but yeah, I found one guy, working for the (french) government, who spent 300 hours, over the last year, calling friends in the US. Jan 26 13:58:18 yeah, there was some progress in the summer Jan 26 13:58:47 but then everything switched to emergency mode due to the Nokia announcements Jan 26 13:58:55 I even logged into it once Jan 26 13:59:45 well, last time when i checked on late autumn, it was still goal Jan 26 14:00:22 well, I've heard that TMO got funny new administrator - lets hope that he will be able to deal with technical issue Jan 26 14:00:23 I hope it still is once everything is stabilized Jan 26 14:00:25 issues* Jan 26 14:00:45 hopefully to the point that he will be too sick with TMO to even think about being trigger-happy again Jan 26 14:00:56 well, whatever Jan 26 14:01:03 same here, M4rtinK Jan 26 14:01:11 autobuilder is enough pita as it is Jan 26 14:04:00 OBS makes collaboration easy Jan 26 14:04:13 xes: tmo also slowing down now? LOL, I guess the guys at Nemein will have a very tough weekend Jan 26 14:05:16 also for Mer, you can pull packages directly from the SUSE OBS instance (wont directly work for Fremantle though) Jan 26 14:05:50 DocScrutinizer05: yes... slowing down is not a good. You know that usually is just the beginning Jan 26 14:06:25 hmm, >>Cpu(s): 5.0%us, 0.8%sy, 0.0%ni, 93.8%id, 0.3%wa, 0.0%hi, 0.0%si, 0.2%st << Jan 26 14:06:54 and tmo is _not_ behind that friggin firewall/NAT Jan 26 14:07:52 it's indeed a tad slow, but that might be some backup via network, or sth Jan 26 14:08:11 wha sort of FW is it? Jan 26 14:13:33 M4rtinK, fremantle cobs is really dead Jan 26 14:14:24 jacekowski: xen-grid vm Jan 26 14:39:20 jacekowski: Endian Jan 26 14:40:02 hm, did we forget to backup the source of the packages in the repos? Jan 26 14:40:31 I hope not Jan 26 14:40:32 ~mirror Jan 26 14:40:33 i guess mirror is http://maemo-archive.wedrop.it/ http://maemo.merlin1991.at/apt-mirror/ http://talk.maemo.org/showthread.php?p=1315143#post1315143 Jan 26 14:40:49 oh, it's in source/ Jan 26 14:40:54 phew Jan 26 14:43:09 * DocScrutinizer05 starts to feel too sick with post sin this channel to NOT feel trigger-happy Jan 26 14:43:20 posts in* Jan 26 14:43:44 ? Jan 26 14:43:46 why so, rootsh discussion was better than popcorn Jan 26 14:43:51 kerio: i'm not entirely insane, i did save the sources aswell :D Jan 26 14:44:20 i remember RMO having the sources in the same place as the .debs but maybe i'm confused Jan 26 14:45:33 kerio: rmo always had the sources in sources/ cssu-thumb has them in place Jan 26 14:45:54 no, i was thinking of the debian repos Jan 26 14:47:06 M4rtinK, I've nuked overlay database which seems like fixed problem I was mentioning earlier. Also, changed useragent to chrome, will report after a week or so if it make me less bannable by google :P, Jan 26 14:47:19 can't test right now as I'm already temp banned due to tiles Jan 26 14:47:30 thanks for your help and interesting talk! Jan 26 14:48:51 is there key combo mapped to esc on N900? Jan 26 14:49:35 Estel_: thanks - yeah, semi-longterm testing ins the only way to work :) Jan 26 14:50:42 is the wiki still down? Jan 26 14:52:32 yep, definatly looks that way to me :) Jan 26 14:53:09 jonwil: same here Jan 26 14:57:28 CcxCZ_: nope Jan 26 14:59:38 how do I play angband then? D: Jan 26 15:00:19 I've found xmodmap, but can't find xev Jan 26 15:10:08 anyway, off to shopping, and cooking for 8. What a waste of fucking time. Jan 26 16:06:28 does phone functions work with an alt os on n900 (like debian / ubuntu with ofono) ? Jan 26 16:06:38 afaik no Jan 26 16:07:35 kerio: thanks Jan 26 16:13:59 Paul1210: there are some issues with audio quality making it unuseable afaik Jan 26 16:53:48 I suspect the audio quality issues are because the alternative operating systems do not have the relevant Nokia-special PulseAudio algorithms that the stock OS has Jan 26 17:33:38 is it posible to rum maemo on any of the android tablets Jan 26 17:33:42 run* Jan 26 17:36:09 is it posible to run maemo on any of the android tablets? Jan 26 17:42:53 greenbarel: depends what exactly you mean. Maemo as a whole will not work anywhere else that the NIT it belongs to. Jan 26 17:43:21 s/that/than/ Jan 26 17:43:22 Sicelo meant: greenbarel: depends what exactly you mean. Maemo as a whole will not work anywhere else than the NIT it belongs to. Jan 26 17:45:49 i think someone ported it to htc hd 2 phone Jan 26 17:47:38 here it is Jan 26 17:47:38 http://www.techknots.com/mobiles/htc-hd2-running-maemo/ Jan 26 17:48:50 i realy like maemo i had n900 but battery life is short, and it is shaped like brick Jan 26 17:49:24 mine is not shaped that way. Jan 26 17:54:26 ~NIT Jan 26 17:54:27 extra, extra, read all about it, nit is Nokia Internet Tablet Jan 26 17:55:09 my spare N900 has weeklong batterylife :P Jan 26 17:55:20 * ShadowJK only uses it for calls and alarm clock Jan 26 17:56:35 yes NIT but thats long forgoten project, last device was n9 Jan 26 17:56:58 and it has meego on it Jan 26 17:57:34 isn't it a little inconvenient to have several devices for daily usage? Jan 26 17:57:46 n900 with 1.5 cm i think is the thickest phone Jan 26 17:58:01 i just have pc and phone Jan 26 17:58:41 and i am looking for some tablets and i am interested in buying some if i can install maemo on it Jan 26 17:58:45 was thinking about ShadowJK's usage :) Jan 26 17:59:05 xD Jan 26 18:01:46 btw have anyone here experience with metasploit and armitage i am having some dificulties with it Jan 26 18:02:14 * Sicelo wouldn't be surprised is that HTC+Maemo video is bogus. I haven't viewed it btw Jan 26 18:03:35 i read it but idk about it, htc hd is capable of running lots of mobile os idk why could't run maemoo Jan 26 18:03:50 the colors are all weird Jan 26 18:03:53 i bet it's vnc Jan 26 18:04:25 "This video is a Dramatization of what it could be like." Jan 26 18:04:25 yep Jan 26 18:06:04 0:10 Jan 26 18:06:43 do you want to know why it couldn't run maemo? Jan 26 18:06:46 ~closed Jan 26 18:06:47 somebody said closed was http://wiki.maemo.org/Why_the_closed_packages or https://wiki.maemo.org/Fremantle_closed_packages Jan 26 18:10:26 ..vnc is painting the screen from the upside to downside, the screen locker is something not part of maemo.. Jan 26 18:22:53 kgu, just that the spare device is usually on desk and main device is in pocket, so I usually pick up the spare one when someone calls Jan 26 18:23:01 (all my phones have the same phone number associated) Jan 26 18:26:33 ShadowJK: So all of them ring, when someone calls? Jan 26 18:26:53 yes Jan 26 18:28:07 what is the second one? Jan 26 18:28:21 * ShadowJK has 2 N900 Jan 26 18:28:27 in use Jan 26 18:28:43 hot spare :s Jan 26 18:29:00 you mad Jan 26 18:30:49 no, u mad Jan 26 18:33:40 i mean in good way Jan 26 18:42:08 4 N900, 1 N800, 1 N810, 1 Nokia E75, 1 Nokia E70, 1 Asus Transformer tf700 Jan 26 18:42:53 8 N900 Jan 26 18:42:56 beat that Jan 26 18:43:05 although they're mobile in a specific way Jan 26 18:43:14 most of them sold when good buyer found Jan 26 18:43:19 and new ones bought Jan 26 18:43:25 2 of them are dead, part donors Jan 26 18:43:35 oh, sorry, 3 of them are dead Jan 26 18:44:00 (but with good screen module, 2 of them perfect, 1 quite scratched) Jan 26 18:44:08 one main + one backup Jan 26 18:44:18 + 3 rotating, i.e. sold and bought Jan 26 18:44:27 how much is a new screen? Jan 26 18:44:31 so their number varies Jan 26 18:44:51 it depends, most of the time you find chinese replacement with varying quality Jan 26 18:45:05 some really suxx. I keep original ones Jan 26 18:45:25 also define screen Jan 26 18:45:27 lcd module? Jan 26 18:45:30 digitizer? Jan 26 18:45:40 or whole screen module, including slider? Jan 26 18:45:41 the upper scratcy part :) Jan 26 18:45:48 digitizer Jan 26 18:46:14 well, you may fin original for quite cheap, but it's quite impossible to find ones without scratches Jan 26 18:46:20 most of them have one or two big ones Jan 26 18:46:32 perfectly new are, usually, replacements Jan 26 18:46:57 some work as well as original, reportedly, and some are worse than digitizers of pda's from pre 2000 Jan 26 18:47:09 yeah, mine has areas where the protection has gone Jan 26 18:47:17 (digitizer, in case someone is not aware, also work as input device for touch) Jan 26 18:47:30 hm, I always use protecting foils Jan 26 18:47:38 just replace them if got scratched too much Jan 26 18:48:02 yeah, I know. too late now.. Jan 26 18:48:03 interestingly enough, I replaced mine only once. found el-cheapo brand that makes really durable ones Jan 26 18:48:44 wanna buy new digitizer for 50 euro? :P Jan 26 18:49:12 dream on ;) Jan 26 18:49:19 lol Jan 26 18:50:30 seriously, I just thought about price that would convince me to spare my original replacements in perfect condition Jan 26 18:50:46 you never know when you may be in need of using them, so I keep it Jan 26 18:51:28 well, vi____ pays more, reportedly, as he is buying original parts from Nokia care including freplace, but *not* allowing them to replace it Jan 26 18:51:53 preffering to replace it on his own, instead of giving morons at nokia ocymoron care chance to touch and break something Jan 26 18:52:16 sounds like rip off, but he seems happy Jan 26 18:52:38 s/ocymoron/oxymoron Jan 26 18:52:54 anyway Jan 26 18:53:02 I got problem for maemo sound system gurus Jan 26 18:53:03 Estel_: there's that dude on tmo with like 5 n950s Jan 26 18:53:14 he pwns you pretty hard Jan 26 18:53:14 Can it really be a $7 buy on ebay, what's the scam? http://www.ebay.com/itm/New-Touch-Screen-Digitizer-for-NOKIA-N900-N-900-Tools-/330745729904 Jan 26 18:53:19 kerio, not likely Jan 26 18:53:31 as in additiion to N900's i have N950 too :P Jan 26 18:53:34 how many? Jan 26 18:53:50 one is enough, you don't want them to start reproducing Jan 26 18:54:01 btw, he have uncle in R&D, or what? Jan 26 18:54:22 kgu, it may be real, but crappy quality probably Jan 26 18:54:31 dr_frost_dk bought one of those once Jan 26 18:54:37 they work, but very, very poorly Jan 26 18:54:57 BTW, how much is N950 sold nowadays? Jan 26 18:55:22 wonder if you can buy digitizer directly from nokia Jan 26 18:55:26 I'm buying new house and need to check if selling one device is enough already :P Jan 26 18:55:47 Nokia could give last breath, at last Jan 26 18:55:59 it would then be quicker to become cult collector device Jan 26 18:56:04 last good product of nokia Jan 26 18:56:14 last good product of first mobile phone company Jan 26 18:56:39 in 30 years tech collectors will get orgasm just by seeing one Jan 26 18:56:53 ~450 available in world Jan 26 18:57:11 enough joking Jan 26 18:57:22 Me got dillema, as said, for audio gurus Jan 26 18:57:29 while using bluetooth adaptor... Jan 26 18:57:40 I can put normal headphones into jack too, and radio works Jan 26 18:57:58 but, fm radio works only on those wired headset or speakers Jan 26 18:58:15 I wan't it to forward music via bluetooth at the same time. Doable? Jan 26 18:58:21 via some fancy alsa mixer settings? Jan 26 18:58:41 or won't work, as when bluetooth is in fm mode, it won't work with bluetooth headset? Jan 26 18:58:59 fm radio needs an antenna which the headphone cable doubles as... Jan 26 18:59:04 sure Jan 26 18:59:14 thats why i said bluetooth headset & wired one connected Jan 26 18:59:16 at the same time Jan 26 18:59:18 ah Jan 26 18:59:21 dunno then Jan 26 18:59:42 actually, I'll check if it plays radio at all with bluetooth headset paired, via anything... Jan 26 18:59:50 if yes, it should be 100% doable Jan 26 18:59:53 btw, CFM Radio works even without earphones, of course, mostly a hiss Jan 26 19:00:13 well, we need to spank javispedro Jan 26 19:00:34 he got finished version, that allow to listen to radio without cpu used Jan 26 19:00:34 Hm, is fmradio connected analogically to sound mixer? Jan 26 19:00:37 analog passthrough Jan 26 19:00:45 yea Jan 26 19:01:01 but only unreleased version of javispedro's radio uses this feature Jan 26 19:01:06 other things use digital Jan 26 19:01:16 digital should be able to send via bluetooth Jan 26 19:01:25 as it goes through cpu Jan 26 19:02:00 yeah Jan 26 19:08:42 Meanwhile: Unemployed Chinese Graduates Say No Thanks To Factory Jobs Jan 26 19:13:38 Will ubuntu on the n900 be able to handle calls and sms? Jan 26 19:14:13 not yet Jan 26 19:14:44 sms should actually already work Jan 26 19:16:47 hope that ubuntu giving effort on the smartphone market will have good sideeffect on a working ubuntu on n900 :) Jan 26 19:17:12 hmm Jan 26 19:23:36 vi____: *poke* Jan 26 19:31:23 buntkuh on N900? i'm not sure whom I hate more, Canonical or Google. So probably I might rather switch to nitfroid than to buntkuh Jan 26 19:36:05 but wasnt harmattan all Kuh-Tee based? :) Jan 26 19:47:54 btw, sygic is kinda open to download from the mirrors Jan 26 19:49:32 oh, it's open to downloads from the officila mirror too Jan 26 19:49:58 official mirror? Jan 26 19:50:04 er, from the official repo Jan 26 19:50:09 that is, ovi Jan 26 19:50:41 sure Jan 26 19:51:05 it makes little sense, i thought it wasn't freeware Jan 26 19:51:53 Yeah you get unlock code when you buy it, changes it from demo to full Jan 26 19:51:55 iirc Jan 26 19:52:14 oic Jan 26 20:02:44 this sandisk mobile ultra card is shit Jan 26 20:03:07 6M seq write Jan 26 20:03:14 class 10 / uhs-i Jan 26 20:03:40 possibly not configured/recognized correctly Jan 26 20:04:07 sd is a pain with all its modes and configuration parameters Jan 26 20:05:27 sandisk sucks on usb sticks ... no surprise SD cards are the same :p Jan 26 20:05:43 except that sandisk sds are the only good ones Jan 26 20:05:55 kingstons work ok Jan 26 20:05:59 Well also its cpu owuld be running slower when not in a uhs host Jan 26 20:06:08 kingstons are the worst :-) Jan 26 20:07:40 i've got patriot EP pro in my dslr and so far it's faster than sandisk Jan 26 20:08:15 well any random crap is fast in cameras and similar workloads Jan 26 20:09:04 not really Jan 26 20:09:22 I wonder if this card needs a offset Jan 26 20:09:29 kingston does not produce sd-cards .. they just brand them Jan 26 20:09:34 alignment offset Jan 26 20:09:41 i've tested a lot of cards Jan 26 20:09:47 ShadowJK: skip the hyper-optimized space for vfat Jan 26 20:09:55 sandisk, lexar, kingston, patriot, samsung Jan 26 20:10:06 and sandisk has the best controlers, but linux is often a bit flacky, especially with older kernels and new cards Jan 26 20:10:06 and few others Jan 26 20:10:23 and i could measure a difference Jan 26 20:10:26 with kingston you almost never get kingston.. they're mostly a rebranding shop Jan 26 20:10:32 and patriot is the best so far Jan 26 20:11:52 there are only 3 or 4 producers .. sandisk, samsung, micron .. any else i cant rememnber? Jan 26 20:12:02 for flash and controlers Jan 26 20:12:26 hm, it seems to be doing pretty well my cardkiller test Jan 26 20:13:00 write block at pos x, write block at pos x+512b Jan 26 20:13:03 repeat Jan 26 20:13:10 4 threads doing that Jan 26 20:13:44 Samsung Essential I tested earlier gave about 100 kbytes/sec Jan 26 20:14:05 looks like about 4M/sec on this sandisk Jan 26 20:14:24 worst I've seen was a Kingston at about 20 kilobytes/sec Jan 26 20:14:57 vhttp://www.theregister.co.uk/2013/01/25/unlocking_change/ seen? Jan 26 20:15:05 sandisk is known to perform very well on concurrent/random access Jan 26 20:15:09 Hm... sneaky "v"... Jan 26 20:15:34 and class ratings dont help you for smartphone like usage patterns .. Jan 26 20:16:05 cehteh: intel, toshiba Jan 26 20:16:09 cehteh: SMC Jan 26 20:16:41 cehteh; more recent ones aren't nearly as good as 2009 sandisk class 4 :-( Jan 26 20:16:45 intel produces flash chips .. but iirc they dont sell to/for SD card producers Jan 26 20:16:57 TLC and 6-12 megabyte block sizes nowadays :( Jan 26 20:17:17 it's all about the controller anyway :) Jan 26 20:17:35 yes i have a old class 4 sandisk in my n900 and it performs very well Jan 26 20:17:37 bunny had decapped a samsung card and it looked like it had a massive arm cpu in it :) Jan 26 20:17:41 i still want a ram-based uSD :c Jan 26 20:17:48 so i was hoping for better perf from the samsungs Jan 26 20:17:56 but they've probably changed them Jan 26 20:19:54 Sandisk has a thing where they can treat part of flash as SLC and the rest as MLC, allegedly some of their cards have slc cache for small random writes Jan 26 20:24:40 I finally found a place selling Samsung "Pro" card that freemangordon said was good.. costs like 80euro though :s Jan 26 20:28:26 what package should I install to be able to compile stuff? Jan 26 20:28:39 CcxCZ_: on device? Jan 26 20:28:43 nope Jan 26 20:28:49 can't do that in any meaningful way Jan 26 20:28:58 the toolchain is too big Jan 26 20:29:14 sadface Jan 26 20:29:19 looks like a minimal write size of 64k or 128k.. Jan 26 20:30:00 tmo down for maintenance back at 9pm Jan 26 20:30:11 kerio; get a microsd card, format it ext3, copy / to it, chroot into microsd card, apt-get install build-essential... Jan 26 20:30:13 I used to compile dwm and urxvt on my G1... Jan 26 20:30:35 ShadowJK: meh Jan 26 20:31:06 the stuff in sdk and sdk/tools otherwise is installed on / which is just 256M Jan 26 20:31:22 you can also do ~m32g Jan 26 20:31:25 ~m32g Jan 26 20:31:30 /home also doesn't have all that much space for all the random -dev crap :-) Jan 26 20:31:32 hrmpf Jan 26 20:31:39 ~m32 Jan 26 20:31:45 oh Jan 26 20:31:53 DocScrutinizer05: infobot is AWOL Jan 26 20:36:47 there's one partitioned nand inside and a mmc or sd card and no easy way to plug in other storage, am I correct? Jan 26 20:38:40 onenand 256M, emmc 32g (vfat, 2g ext3 /home, 768M swap), empty MicroSDHC slot Jan 26 20:40:42 also, fstab is a lie, dont try edit it. :-) Jan 26 20:44:28 unless you're a serious person who disabled the fstab generation Jan 26 20:49:33 * CcxCZ_ wonders if he could run new enough kernel to have btrfs spanning both mmc and sd Jan 26 20:52:26 kerio: I'm quite serious at trimming nonsense, but I haven't had time to delve that deep in the two days I have this device :-) Jan 26 20:53:10 beware that mmc drivers get loaded pretty late during boot :) Jan 26 21:00:13 where I can see details on what gets mounted? I think the best approach for now is shrinking the vfat and putting gentoo chroot on separate partition there Jan 26 21:04:27 CcxCZ_: 'mount' should show you what mounts where Jan 26 21:05:00 it's done by a series of scripts in /etc/event.d and rc.d, calling other scripts Jan 26 21:05:58 just shrinking vfat and extending ext3 is safe? Jan 26 21:07:14 yes Jan 26 21:07:25 hmm, ext3? Jan 26 21:08:13 usually people extender /home, or create new partitions in the free space Jan 26 21:08:18 -r Jan 26 21:08:49 yes, by ext3 I meant /home Jan 26 21:10:26 * ShadowJK 'd just use a uSD :D Jan 26 21:12:11 I'd rather have system on mmc and data on sd, but alas I do lack a suitable sdcard anyway Jan 26 21:42:24 oh, the wiki is back Jan 26 21:42:27 and garage too Jan 26 21:42:37 main site still 503 Jan 26 22:12:12 what's 503 Jan 26 22:12:39 Service unavailable Jan 26 22:13:28 ah Jan 26 22:49:34 Sandisk extreme pro microsdh 16g.. Looks kickass so far. Jan 26 22:55:59 Seems to do 45-50 iops 4k random write steady state (300 ish on non-aged card). Mobile Ultra (2012): 4 iops, Mobile Ultra (2009): 8 iops Jan 26 22:56:16 Sandisk Class 4 (2012) 32g: 1 Jan 26 22:56:17 :) Jan 26 22:57:24 not bad Jan 26 22:57:57 A-data "super fast" (2012): 22 iops Jan 26 23:06:01 ... the hell.. it's slower at 16k random write than 4k and 8k :-) Jan 26 23:10:33 4k: 48, 8k: 28, 16 and up: 4 Jan 26 23:11:20 "ultra" Jan 27 01:14:55 dunno if this has been asked Jan 27 01:14:57 was there an announcement about TMO maintenance? Jan 27 01:48:09 ok, so TMO is down but the wiki is back up :) Jan 27 02:15:48 Can't we get to a point where all the Maemo infra is all up at the same time? :P Jan 27 02:18:37 Give HiFo $20 Jan 27 02:31:21 I am broke and between jobs, I cant afford to donate $ (which is why I use my skills to donate code and reverse engineering to the Maemo community instead :) Jan 27 02:31:30 well not broke, just have no spare cash Jan 27 02:33:47 Sorry, I was being facetious. Jan 27 02:59:17 jonwil: tmo was not down... at least not in terms of "unreachable" Jan 27 02:59:36 true **** ENDING LOGGING AT Sun Jan 27 02:59:59 2013