**** BEGIN LOGGING AT Wed Jul 27 02:59:57 2011 Jul 27 03:06:21 I want people aware of good pygame and port to N900... Jul 27 03:53:08 Hey guys, what's new in terms of the next n900 and its openness? Jul 27 03:53:38 I'm tired of compromising so much for openness, without actually having openness Jul 27 03:53:40 next n900 not for sale Jul 27 03:53:50 yeah, the dev version Jul 27 03:53:52 right? Jul 27 03:54:29 so nvm, it end up with nobody own it - they all placed in Nokia history museum Jul 27 03:55:04 updateme: the current user will have to send them back to Nokia if I read the right Article Jul 27 03:55:47 yeah, but the device they're releasing to everyone won't be... But I hear it doesn't have a keyboard Jul 27 03:57:15 no keyboard on N9 but bluetooth keyboard still can be used. Jul 27 03:57:37 that sucks Jul 27 03:58:03 It uses MeeGo, right? Jul 27 03:58:09 Which means it isn't very open, once again? Jul 27 04:00:49 How much OpenNess can you expect from an already existing large international corporation ? Jul 27 04:01:29 I don't know. I'm only trying to compare the openness of meego to maemo, I guess Jul 27 04:01:54 harmattan and openness? Jul 27 04:01:55 lol Jul 27 04:01:57 Since there were so many things with dampered functionality, like the address book and calendar in maemo Jul 27 04:02:43 ug Jul 27 04:02:55 jiero, you can keep the device if you want to Jul 27 04:02:58 well when will there be a powerful phone that runs shit as open as a real OS Jul 27 04:03:14 updateme, aava phone Jul 27 04:03:47 again, we need the community to start our own "GTA04" Jul 27 04:04:00 newer board for the RX-51 Jul 27 04:04:05 let's call it the RX-52 Jul 27 04:04:28 hardware is hard (no pun), which helped openmoko to fail Jul 27 04:04:31 the only problem would be design, manufacture and distribution Jul 27 04:04:47 but if we can get a mass produced powerful device, which allowed for a non-crippled OS to run, then we can win Jul 27 04:04:57 oh, and it would cost $$$ to produce boards Jul 27 04:04:58 lol @ "the only problem" Jul 27 04:05:30 well I don't know anything about the Aava Jul 27 04:05:43 updateme, we can probably pay someone to make the PCB design files, and have some chinese factory churn the boards out Jul 27 04:06:03 do note that we'll have a hard time acquiring OMAP3630 parts Jul 27 04:06:19 likely need to buy them in lots of >1000 Jul 27 04:06:30 I know some people who have done startups with their own touch screen devices succesffuly Jul 27 04:06:42 so unless we get ALL of TMO together, we're pretty stuck in the mud Jul 27 04:06:48 But it seems there is a bottleneck other than hardware Jul 27 04:07:01 Like support, maybe Jul 27 04:07:17 "yeah, Tmo, I've a nokia phone" "okay no problem, here's a sim" ....etc Jul 27 04:07:17 good news though, as omap is very opensource friendly Jul 27 04:07:19 xloader etc are freely available Jul 27 04:07:33 oh right, the problem would be in a baseband Jul 27 04:07:51 transferring a nokia BB5 chip would be illegal wrt what IMEI really is Jul 27 04:08:18 so Jul 27 04:08:25 updateme, LOL SUPPORT Jul 27 04:08:32 yeah Jul 27 04:08:42 tmo's suddenly supporting devices not made by nokia Jul 27 04:08:52 anyways, that's beside the point Jul 27 04:09:02 What is the point? Jul 27 04:09:08 making the board Jul 27 04:09:19 That doesn't seem like the bottleneck Jul 27 04:09:26 then what is? Jul 27 04:09:50 It seems like it would be easier to support and develop an OSS solution for the most powerful established phone/hardware out there already Jul 27 04:09:54 instead of trying to build one from scratch Jul 27 04:10:14 since they put their R&D into it, if it is open enough, then can't it be "hi-jacked"? Jul 27 04:10:31 you mean linux? Jul 27 04:10:37 see meego Jul 27 04:10:40 +cordia Jul 27 04:10:51 Spell it out please Jul 27 04:10:59 MeeGo CE + Cordia UX Jul 27 04:11:06 looks like Fremantly Jul 27 04:11:09 *Fremantle Jul 27 04:11:23 the only thing that makes the N900 shine is the multitasking Jul 27 04:11:29 so why not take just that? Jul 27 04:11:55 Take the n900? Jul 27 04:11:58 I am/have one Jul 27 04:12:08 take the user interface of the N900 Jul 27 04:12:19 also known as fremantle's Hildon Jul 27 04:12:36 put it on a linux distro with glibc and gtk+ Jul 27 04:12:52 there, no more nokia dependency, plus we can use it on any phone Jul 27 04:13:05 the problem is, what phones properly support meego Jul 27 04:13:08 And it would have enough support to be a full smartphone OS on the n900 before the entire userbase starts using/supporting a new phone? Jul 27 04:13:32 I have a feeling I'd rather use the n900 with meego than aara Jul 27 04:14:14 updateme, would meego CE have enough support -- a widely accepted Fremantle replacement before only 2 N900s on Earth still work? Jul 27 04:14:31 face it, a replacement OS is still just a replacement OS Jul 27 04:14:39 the HTC HD2 is the ONLY exception to that rule Jul 27 04:15:00 So what are you saying? Jul 27 04:15:05 You support my argument and not yours? Jul 27 04:15:06 lol Jul 27 04:15:19 uh wut? Jul 27 04:15:21 That's what I was hinting at - just a replacement OS Jul 27 04:15:25 So you're in agreement Jul 27 04:15:47 the problem is, what phones properly support meego Jul 27 04:16:04 i think we were agreeing before a slight language slip ;) Jul 27 04:16:20 i mean, who's going to make a meego phone? Jul 27 04:16:21 hmm Jul 27 04:16:29 the world is on Android Jul 27 04:16:40 there's a massive ecosystem there Jul 27 04:16:50 indeed Jul 27 04:16:59 N900 has advantages still Jul 27 04:17:11 N900 has it's advantages for US people Jul 27 04:17:15 some of them due to its openness, but at the same time is hurt significantly by not being open enough Jul 27 04:17:19 not the other plebians Jul 27 04:17:30 at first I read "united states" people Jul 27 04:17:31 lol Jul 27 04:17:38 a lot of people still treat their shiny new droid devices as toys Jul 27 04:17:51 what do we do with our devices? Jul 27 04:17:54 I treat my n900 as an expensive toy Jul 27 04:17:55 massive disconnect there Jul 27 04:18:00 haha Jul 27 04:18:02 yeah, I agree Jul 27 04:18:10 again, Mobile Computer vs Smartphone Jul 27 04:18:24 there is a difference Jul 27 04:18:32 but the gap seems small enough to close without much $$$ Jul 27 04:18:51 like how much it is worth closing vs how much it costs to close Jul 27 04:18:52 updateme, we'll never know Jul 27 04:19:01 I'm an optimist I guess Jul 27 04:19:06 the time that N900 was being sold, the N97 was being promoted out the wazoo Jul 27 04:20:00 I just want an n900v2 Jul 27 04:20:08 with a fully open OS Jul 27 04:20:17 So I can go in and change senseless shit Jul 27 04:20:44 Why is fully open so important? Jul 27 04:20:49 It isn't Jul 27 04:21:00 just open enough to have options Jul 27 04:21:07 the fact that the N95 didnt ship with Maemo proves that Nokia's an old man, afraid to take a leap of faith, to become young once more Jul 27 04:21:11 Unfortunatly, carriers and content partners HATE openness on phones Jul 27 04:21:17 and now, they're filled with regret Jul 27 04:21:17 I agree that it is nice for software to be open so that you can fix those stupid annoyances. Jul 27 04:21:57 maemo would keep up better for us if it were open Jul 27 04:21:59 Like TomTom car navigators that have shitloads of UI stupidity (mostly in the form of completely useless confirmation prompts) that would be trivial to remove if you had the source code. Jul 27 04:22:04 fixing all the things that each of us care about Jul 27 04:22:39 from a technical standpoint, the N95(NAM) would've easily handled OS2007 Jul 27 04:22:53 Or if the n900 were more open we could sort out more issues with the GPS functionality and replce our TomToms Jul 27 04:23:17 updateme, GPS functionality is mostly because of the crappy antenna Jul 27 04:23:17 Hurrian: although, it is pretty slow to boot and such on the n900 Jul 27 04:23:27 Has anyone modded it? Jul 27 04:23:34 I didn't know that *interested* Jul 27 04:23:44 updateme, only did external GPS mod on N95 Jul 27 04:23:52 Also, I think, because it doesn't have a compass Jul 27 04:23:58 so directionality doesn't work so well Jul 27 04:24:10 Btw, isn't N950 supposed to have a compass? Jul 27 04:24:25 Doesn't seem to be working in Maps yet :( Jul 27 04:24:27 I'd care if I could get one Jul 27 04:24:45 crikey, going from maemo42 to maemo46 just ate 10+mb of my rootfs Jul 27 04:25:18 fusi, if rootfs isn't full, no need to worry Jul 27 04:26:35 why don't you have 47, fusi? Jul 27 04:26:59 yeah, turn by turn navigation would be so nice on the n900 Jul 27 04:27:19 I run with tunes sometimes, and keeping that on during a long run would be fscking sick Jul 27 04:27:39 "turn right" *hot surreal british chick voice* Jul 27 04:28:05 updateme, arduino + compass shield + USB Hostmode + USB Serial Console + software Jul 27 04:28:16 not the best solution, but it's a solution Jul 27 04:28:47 it is a solution at home, but if someone wants to be able to have it in pocket or walking without breaking it, then not really Jul 27 04:29:58 Sorry, I'm more of an elitist user than a dev Jul 27 04:30:05 hope that can be appreciated Jul 27 04:30:35 agreed Jul 27 04:30:55 well, there isnt much space inside the n900's case for yet another PCB+wires Jul 27 04:31:13 well, I'm unclear on what is wrong with the GPS antenna Jul 27 04:31:15 it's rather tight from all the very thick plastic and metal cage Jul 27 04:31:46 Some of the issues I've dealt with lead me to believe it is partially software... Since some programs say they get signals more quickly than others Jul 27 04:32:12 Surely, which is why n900v2 could pack it all in tighter and still have more features (moore's law etc) Jul 27 04:33:22 5:20 < updateme> why don't you have 47, fusi? Jul 27 04:33:34 not as stable i find Jul 27 04:33:47 Ah, what part of it? Jul 27 04:34:03 So you update manually instead of with 'the thing pops up'? Jul 27 04:34:04 heh Jul 27 04:34:18 :) Jul 27 04:34:34 dsp overclocking seems unstable Jul 27 04:34:51 is fine on 46 o_9 Jul 27 04:35:08 what has dsp overclocking allowed you to do? Jul 27 04:35:32 well, i dont :) Jul 27 04:35:56 only the cpu Jul 27 04:36:04 but i dont like the feeling Jul 27 04:36:07 :p Jul 27 04:36:08 Is it night and day? Jul 27 04:36:39 I already change out my batteries every few hours when I'm carrying, so I don't need to be using anymore power anyway Jul 27 04:36:39 ~80% @250mhz Jul 27 04:37:38 get bout 28hrs avg use, just bought 2 of those scud batts :) Jul 27 04:38:12 im off to bed Jul 27 04:38:18 5:30am :s Jul 27 04:38:24 nn o/ Jul 27 04:38:45 same, I carry 3 batteries with me where ever I go Jul 27 04:45:35 DocScrutinizer: have you tried any turn by turn GPS using the n900? Jul 27 04:46:06 Also, do you know the the n900s GPS issues are limited by just one thing, or several? Jul 27 04:46:19 Like, is it just a hardware antenna issue, or softare too, or what? Jul 27 04:46:20 mappero and marble Jul 27 04:46:55 can three people vote for this please? Jul 27 04:46:56 http://maemo.org/packages/package_instance/view/fremantle_extras-testing_free_armel/pyradio/3.7/ Jul 27 04:47:02 and for me there are no GPS issues. AIUI those are limited to devices without SIM Jul 27 04:47:04 need to move it to extras already Jul 27 04:47:40 mappero? Jul 27 04:47:48 seriously, can I even get street names with that app? Jul 27 04:48:13 I feel like I'm in the 90s using that app Jul 27 04:48:38 I mean 'directions' rather Jul 27 04:49:03 updateme: modrana does turn by turn Jul 27 04:51:24 joga: the question was if I have tested any. I only tested mappero and marble for turn by turn (voice) guidance Jul 27 04:52:30 monav backend even does offline routing for marble Jul 27 04:52:37 I thought only sygic did turn by turn - at least that's how it was marketed..... as the only one that could Jul 27 04:55:12 so there is a way to do turn-by-turn on the N900 for zero cost? Jul 27 04:56:15 torrents Jul 27 04:56:22 That's what I thought Jul 27 04:56:57 jonwil: marble Jul 27 04:58:02 voice of marble is ueber-new, as in "will arrive tomorrow" Jul 27 04:58:43 only English voice so far, contest for other lang still running Jul 27 04:59:43 or you use tomtom voicefiles, they are compatible Jul 27 04:59:56 where does Marble get maps from? Jul 27 05:00:00 OpenStreetMap? Jul 27 05:00:05 OSM? Jul 27 05:00:21 I think you can config several map sources Jul 27 05:01:40 DocScrutinizer: I heard a story about the cost of capacitive touch screens being based upon the transparent metal they're built with - do you know anything about this? Jul 27 05:01:54 no Jul 27 05:03:59 man, there is so much crap and misinformation having to do with the pros and cons and functionalities of resistive and capactive touch screens Jul 27 05:21:48 * jonwil is surprised that the prediction of smartphones + free-as-in-no-cost nav software killing standalone GPS hasn't come to pass Jul 27 05:23:02 jonwil: such easily solutions are possible too Jul 27 05:23:08 early days Jul 27 05:23:48 I think the problem is that the free-as-in-no-cost nav software isn't as good as TomTom/Garmin/etc Jul 27 05:24:20 and people question the accuracy of free vounteer-edited map data Jul 27 05:25:30 and rightly so Jul 27 05:25:32 haha Jul 27 05:30:25 man, these test and extra apps never really work Jul 27 05:30:35 just tried GeePS Jul 27 05:50:21 wow this is the best gps app I've tried hands down Jul 27 05:50:23 on the n900 Jul 27 05:50:53 can't scroll around quickly like you can on android, but I know know what causes that bottleneck... probably not the network though (since it would be the same datarate for map downloads) Jul 27 05:52:48 updateme: what is? Jul 27 05:53:19 try monav Jul 27 05:53:34 almost perfect routing, vector maps Jul 27 05:53:53 does it speak? Jul 27 05:53:58 soon Jul 27 05:54:15 christoph has already a private branch with speech support Jul 27 05:54:41 for now I think I'll use modrana, I like its interface Jul 27 05:54:43 http://monav.openstreetmap.de/ Jul 27 05:55:53 cloudgps is nicely smooth, but has no much functionality besides eye-candy Jul 27 07:04:19 argh, why on earth will scratchbox not work no more? it worked fine yesterday, now i get a "/usr/bin/apt-get: Accessing a corrupted shared library" when trying to install new software. fakeroot? the same... Jul 27 07:05:35 did you install something odd? Jul 27 07:06:58 i used the sb-menu to "select" the arm target Jul 27 07:07:11 never used that iirc Jul 27 07:07:13 i probably should have just selected it from the command line Jul 27 07:08:01 can still do it ;) Jul 27 07:10:49 ruskie: yah, tried it, but doesnt work, i think i must start from scratch Jul 27 07:10:54 haha, no pun intended :( Jul 27 07:12:58 Anybody have a guide to hide X cursor ...? Jul 27 07:21:09 ruskie: ok, got it fixed, needed to go through the whole setup process (of sb-menu) again Jul 27 07:21:24 anybody have an idea on how to get svn 1.6.x on scratchbox? Jul 27 07:22:16 last I checked you could use an older version with latest servers easily Jul 27 07:22:20 for checkouts iirc Jul 27 07:24:41 its giving me a warning Jul 27 07:25:13 svn: This client is too old to work with working copy 'myRepo'; please get a newer Subversion client Jul 27 07:28:51 erm Jul 27 07:28:55 working copy? Jul 27 07:29:05 you copied in an already checked out working copy? Jul 27 07:31:52 yes, was i supposed to check out the repo again? Jul 27 07:33:06 yup Jul 27 07:33:55 ok, will try that after the ./configure run Jul 27 10:51:22 I herald you, my supreme master! Lead us into the light of your wisdom and power Jul 27 10:52:30 is there a way to extract WI-FI passwords from a stock N810? Jul 27 10:52:39 infobot: botsnack Jul 27 10:52:39 aw, gee, RichiH Jul 27 10:53:21 at least on n900 its stored in gconf /system/osso/connectivity/, could be same on N810 with any luck Jul 27 11:05:00 hello. 2 questions about marble routing. 1) does "auto zoom" work for you? for me it doenst 2) is there a way to (automatically) rotate map while driving...cant find any option like that Jul 27 11:56:04 which maemo version does fremantle 1.3 correlate with? Jul 27 11:56:14 or the other way: how do i verify i have fremantle 1.3? Jul 27 11:56:36 osso-product-info Jul 27 11:57:12 about product, i assume? Jul 27 11:57:27 that's the other way Jul 27 11:57:30 but there, the version states 20.2010.36-2 Jul 27 11:57:37 ah ok, ima install your package Jul 27 11:57:39 and you compare the output versus tablets-dev.nokia.com Jul 27 11:58:01 good hint, thanks Jul 27 13:10:40 how is xmpp messaging for telepathy called on n950? Jul 27 14:03:55 is there some way to stop the app manager looking for updates when I'm on roaming data? Jul 27 14:05:13 you can't stop it, you can extend the time when HAM goes to check for updates Jul 27 14:05:19 while :; do killall .. ; sleep 5 ; done :D Jul 27 14:05:37 isn't there some "check for updates" checkbox anywhere ? :P Jul 27 14:05:40 HAM suck, that's it Jul 27 14:05:42 yeah an ugly script would work Jul 27 14:06:13 you need to use this gconftool thing Jul 27 14:06:18 there is check for updates box option if one has red pill enabled, however in my case unchecking that has no effect Jul 27 14:06:25 it reminds me of windows registry... Jul 27 14:06:38 there is check for updates box option if one has red pill enabled, however in my case unchecking that has no effect Jul 27 14:06:49 red pill was disabled post-PR1.0? Jul 27 14:07:16 at least i can't access it anymore Jul 27 14:07:32 FIQ|n900, no its not disabled, its made slightly harder to enable Jul 27 14:07:46 oh Jul 27 14:07:52 i liked the previous way :D Jul 27 14:08:17 reminds me of cheatcodes in games lol Jul 27 14:08:28 don't think it's enabled, know it's enabled Jul 27 14:08:48 IDKFA Jul 27 14:09:11 Robot101, http://wiki.maemo.org/Customizing_Maemo#Disabling_Auto_Updates_Check <-- here's one way Jul 27 14:09:55 FIQ|n900, http://wiki.maemo.org/Red_Pill_mode#Activating <--- Re: red pill mode, manual method of activating Jul 27 14:27:31 mbarcode somehow does not process MECARDs Jul 27 14:36:09 edheldil: Process? Jul 27 14:37:26 I'm not sure code was ever written to parse out the DoCoMo formats. Jul 27 14:38:20 well, it displays some helper in addition to the web search one, but it does not do anything visible when clicked Jul 27 14:38:59 Ah, I didn't even realize it did that much. Jul 27 14:39:54 "Open iMode MECARD data" Jul 27 14:55:13 are any of you N900 users have bi-directional calendar sync via CalDav working in Maemo? Jul 27 14:55:25 s/are/do Jul 27 15:05:57 May i flash EMMC Image "RX-51_2009SE_10.2010.13-2.VANILLA_PR_EMMC_MR0_ARM.bin" using combined "RX-51_2009SE_20.2010.36-2_PR_COMBINED_MR0_ARM.bin" ? Jul 27 15:06:33 no Jul 27 15:06:37 these are two different images Jul 27 15:06:41 you have to flash both Jul 27 15:06:46 or just combined Jul 27 15:06:53 never flash emmc alone Jul 27 15:11:56 LarsN: I need mere WebDAV + iCal. Jul 27 15:12:08 but no luck as well Jul 27 15:13:26 edheldil: I'm pretty sure I also have funambol available on our SOGo server. Jul 27 15:13:39 I guess I could grab the syncevolution client and give that a shot. Jul 27 15:14:03 I should likely backup my device, then flash / wipe it clean before starting Jul 27 15:14:11 it's full of failed attempts at this point. Jul 27 15:23:35 angelox|laptop: isn't the queston a bit paradox? Jul 27 15:24:32 angelox|laptop: lease report any ambiguities or missing bits in Jul 27 15:24:35 ~flash Jul 27 15:24:36 well, flashing is http://wiki.maemo.org/Updating_the_tablet_firmware Jul 27 15:24:45 P*lease Jul 27 15:27:41 because i saw the numbers inside image name are different Jul 27 15:28:03 notice: ~flashing is now -maemo-flashing Jul 27 15:28:22 angelox|laptop: ooh, that doesn't matter Jul 27 15:28:24 * ShadowJK wonders if ham would behave better with bindmounts instead of symlinks Jul 27 15:28:37 because it definitely wasnt this bad in pr1.0 Jul 27 15:28:52 angelox|laptop: there's only been iirc 2 VANILLA files, but some 4 or 5 COMBINED files Jul 27 15:29:01 yes,and old and that one Jul 27 15:29:13 s/and/a/ Jul 27 15:29:14 angelox|laptop meant: yes,a old and that one Jul 27 15:29:20 always use freshest of both Jul 27 15:29:54 Understood,thank you DocScrutinizer Jul 27 15:30:00 yw Jul 27 15:30:54 Man, anyone ever used the projects.developer.nokia.com thingy? I got a "want to join project" request from some guy but the site has no option to write messages to anyone?!? Jul 27 15:31:06 ShadowJK: ??? Jul 27 15:31:59 gri: projects.developer.nokia.com feels like it's in beta state Jul 27 15:32:13 That's more than beta state Jul 27 15:32:30 or less, depends on how you see it :) Jul 27 15:32:47 s/projects.// Jul 27 15:32:49 DocScrutinizer meant: gri: developer.nokia.com feels like it's in beta state Jul 27 15:33:41 I'm deleting my source there and switch to gitorious, there people atleast have the chance to contact each other Jul 27 15:34:11 DocScrutinizer, I'm wondering if they accidentally stat() the symlink instead of the intended file, and thus pop up "Checking for updates" constantly.. Also notable is that I never noticed the behaviour on PR1.0 brought to 1.3 through SSU, only on flashed-1.3.. Jul 27 15:37:07 ShadowJK: that might be a reason why I'm missing what you're talking about Jul 27 15:38:35 what popup "Checking for updtaes"? A hildon notifier? how often does that show? Jul 27 15:39:21 Has meego been progressing on the N900 or is it still primarily for playing around with? Jul 27 15:40:10 ShadowJK: seems to me I never seen such notifier - maybe coffee not yet kicked in Jul 27 15:41:06 LarsN: best place: /join #meego-arm Jul 27 15:42:20 anyone want to sell their nokia n8? Jul 27 15:42:32 i readed the warning "Never boot up the device between flashing the FIASCO image and the eMMC image",but i'm curious,why we can't do that? remembering,i wan't not to do,but i want know :) Jul 27 15:43:12 ShadowJK: though I "can not duplicate the issue", might it be related with updater process is closer checking the connectivity, so maybe the looking-for-update is triggered each time some connectivity state change happens? Jul 27 15:46:07 angelox|laptop: I guess because there is not /usr or something like that Jul 27 15:46:08 angelox|laptop: first-boot optifies a lot of binaries and files, means they get copied to /opt which will get erased on flashing VANILLA Jul 27 15:46:19 ah Jul 27 15:46:30 s/copied/moved/ Jul 27 15:46:30 DocScrutinizer meant: angelox|laptop: first-boot optifies a lot of binaries and files, means they get moved to /opt which will get erased on flashing VANILLA Jul 27 15:47:05 ah ok,understood,thank you :) Jul 27 15:47:46 angelox|laptop: that's why the foolproove way is to flash vanilla first, then combined, as then a intermediate accidental boot won't hurt Jul 27 15:48:57 and an accidental boot is happening *really* easily, and esp "noobs" don't even notice it Jul 27 15:49:25 plugging in USB to the device already makes it boot Jul 27 15:49:59 DocScrutinizer, a "modal" progress bar in ham. Pops up when tapping download, pops up on returning to main menu, pops up on going to download again, etc Jul 27 15:50:27 aaaaah that one Jul 27 15:51:16 yeah, it feels a bit broken Jul 27 15:51:41 i dont think it was this broken before ham/apt data got optified Jul 27 15:52:22 I hardly can recall that time Jul 27 15:53:07 but probably you *could* flash PR1.0 and first eMMC image, and see what it's like there Jul 27 15:53:41 stat64("/home/user/.hildon-application-manager-last-update", 0xbeb5d0a8) = -1 ENOENT (No such file or directory) Jul 27 15:53:44 hm Jul 27 15:53:59 keep in mind though that extras wasn't enabled by default in PR1.0 Jul 27 15:54:44 i think enabling it was first thing I did after unboxing and setting clock Jul 27 15:55:15 aaah yes, that's more like it - as HAM ran into "not enough space on /" issues for all the cache and whatnot, it eventually changed to use eMMC to store those things Jul 27 15:55:43 AIUI Jul 27 15:55:57 lol, i created that file, but ham wrote something to it and then deleted it Jul 27 15:56:07 haha Jul 27 15:58:00 while true; do chmod a-w /home/user/.hildon-application-manager-last-update; ln /home/user/.hildon-application-manager-last-update /home/user/.hildon-application-manager-last-update-STAY; done Jul 27 15:58:37 or cp -l /home/user/.hildon-application-manager-last-update /home/user/.hildon-application-manager-last-update-STAY Jul 27 15:58:51 DocScrutinizer, actually it writes another file Jul 27 15:58:51 instead ln Jul 27 15:59:10 /home/user/.hildon-application-manager/last-update Jul 27 15:59:35 HAHAHAHA Jul 27 16:00:17 the benefits of hardcoded strings in code, rather than using decent #define Jul 27 16:00:23 or const Jul 27 16:01:16 isn't "/" == Shift+"-", on qwerty? Jul 27 16:02:22 on 104-key qwerty normal kbd Jul 27 16:03:06 DocScrutinizer: nope, on qwerty / is next to the right shift key Jul 27 16:03:16 ShadowJK: I guess you squished an age-old buf in HAM that made it suck so hard for all of us Jul 27 16:03:46 hiemanshu: that doesn't help to clarify my question Jul 27 16:03:49 DocScrutinizer: no Jul 27 16:03:54 DocScrutinizer: it's just / Jul 27 16:04:05 DocScrutinizer: on UK layout Jul 27 16:04:19 DocScrutinizer: I just answered isn't "/" == Shift+"-", on qwerty? Jul 27 16:04:37 hiemanshu: you did? Jul 27 16:04:50 DocScrutinizer: DocScrutinizer: nope, on qwerty / is next to the right shift key Jul 27 16:04:59 DocScrutinizer: and shift+- is _ Jul 27 16:05:12 yes, jacekowski answered it Jul 27 16:05:20 DocScrutinizer, it didn't solve anything? Jul 27 16:06:10 ShadowJK: HAM is open, no? Jul 27 16:07:26 All I found was that it checks in /home/user first, then .hildon-application-manager/last-update, and deletes the one in user/ if it exists, or something Jul 27 16:07:39 I'd bet somebody used inline string constants in several locations, and occasionally used "/home/user/.hildon-application-manager-last-update" instead of "/home/user/.hildon-application-manager/last-update" Jul 27 16:08:29 ahoy Jul 27 16:08:31 Note that /home/user/.hildon-application-manager-last-update looks definitely fishy Jul 27 16:09:09 only need kernel-power-headers if devving on device right? Jul 27 16:09:28 DocScrutinizer: i agree that path looks wrong Jul 27 16:09:31 it stats it twice, deletes it, then opens the "correct" file Jul 27 16:09:39 it stinks like typo, so terribly intense that I can sense it thru internet ;-) Jul 27 16:09:48 hehe Jul 27 16:09:54 typo sense is tingling Jul 27 16:11:19 look, kids! That's why you define string (and other) constants in one location Jul 27 16:11:46 helps with translation too Jul 27 16:11:49 where's that friggin HAM source? Jul 27 16:11:56 mxr? Jul 27 16:13:21 does anyone know why mutt doesn't work properly in osso-xterm? Jul 27 16:17:17 hmmmm http://mxr.maemo.org/fremantle/search?find=/hildon-application-manager/&string=last-update Jul 27 16:19:22 http://mxr.maemo.org/fremantle/ident?i=UFILE_LAST_UPDATE Jul 27 16:24:37 YAY http://mxr.maemo.org/fremantle/source/hildon-application-manager/src/user_files.c#83 Jul 27 16:24:50 ShadowJK: ^^^ Jul 27 16:27:25 old_path = g_strdup_printf ("%s-%s", full_state_dir, name); Jul 27 16:28:55 honestly, smells like cruft Jul 27 16:34:22 http://mxr.maemo.org/fremantle/source/hildon-application-manager/src/user_files.c#44 looks fishy - or I'm not smart enough to see how smart it is Jul 27 16:35:43 err I'm not smart enough (unless I look a little closer) Jul 27 16:36:08 what u lookin for Doc? Jul 27 16:36:10 still this is all so incredibly convoluted Jul 27 16:36:55 if mkdir fails for some other reason than EEXIST, return failure.. Jul 27 16:36:59 for a bug in HAM that makes it forget timestamp of last update as saved in either /home/user/.hildon-application-manager-last-update or /home/user/.hildon-application-manager/last-update Jul 27 16:37:06 that code is fugly :( Jul 27 16:37:07 xkr47: yup Jul 27 16:37:16 fusi: extremely Jul 27 16:38:07 migration handling mixed into normal functions for even *reading* the file Jul 27 16:38:42 all that old-fike/new-file cruft should go completely or at very least get called on HAM startup once Jul 27 16:39:32 AIUI it's only needed to migrate from old HAM version to "new" HAM version, so exactly one time, and even then it's not really of much help Jul 27 16:42:51 tbh all this cruft with renaming /home/user/.hildon-application-manager-last-update to /home/user/.hildon-application-manager/last-update, creating /home/user/.hildon-application-manager/ dir etc - this should be done in HAM.deb postinstall Jul 27 16:43:39 really insane to do it inside the program, *each time* you run HAM since PR1.1(?) Jul 27 16:45:02 so, is this busybox again or am I doing something wrong? ~ $ echo -n $(( 2>1 )); echo -n $(( 0>1 )); (( 0>1 )) && echo -n y; (( 2>1 )) && echo y Jul 27 16:45:05 10yy Jul 27 16:46:30 Shapeshifter: what's that?? Jul 27 16:46:53 looks terrible Jul 27 16:47:19 why do both 0 and 1 resolve as true when coming from (( ))? Jul 27 16:47:56 (( 0>1 )) && echo y should'nt print anything Jul 27 16:48:13 $(( 2>1)) evaluates to "1" in my book, $((0>1)) is completely weird I'd think, and wtf is >> echo -n y; (( 2>1 )) < huh? Jul 27 16:49:13 whats this in Jul 27 16:49:14 what are you talking about. Jul 27 16:49:15 c? Jul 27 16:49:53 DocScrutinizer: first I'm printing the two, hence the $(()), then I'm using them as tests, hence no $ Jul 27 16:49:58 afaik the compiler will only evaluatethe first conditional with && Jul 27 16:50:09 ooh bash Jul 27 16:50:11 nvm Jul 27 16:50:21 echo -n $(( 0>1 )) -->0 Jul 27 16:50:36 echo -n $(( 2>1 )); -->1 Jul 27 16:50:58 ja Jul 27 16:51:10 $ echo -n $?; (( 2>1 )); echo $? Jul 27 16:51:11 00 Jul 27 16:51:13 but what is >> echo -n y; (( 2>1 )) << meant to do? Jul 27 16:51:18 uh wait Jul 27 16:51:45 $ (( 0>1 )); echo -n $?; (( 2>1 )); echo $? Jul 27 16:51:45 00 Jul 27 16:51:49 here Jul 27 16:51:54 Shapeshifter: the whole thing looks completely meaningless to me Jul 27 16:52:08 hehe Jul 27 16:52:11 the return values of both are 0 Jul 27 16:52:53 RM680-22-6_PR_RM680:~# $ (( 0>1 )); Jul 27 16:52:55 /bin/ash: syntax error: unexpected "(" (expecting ")") Jul 27 16:53:08 switch the conditions around, do both outputs become 1 Jul 27 16:53:11 ]$ (( 0>1 )); echo -n $?; (( 2>1 )); echo $? Jul 27 16:53:12 10 Jul 27 16:53:21 here's the same in bash on my pc Jul 27 16:53:28 if they do ur not assigning the second one correctly? Jul 27 16:53:35 WTF is ((2>1)) ????? Jul 27 16:53:39 if you need to compare numbers, what's wrong with the good old POSIX [ 0 -gt 1 ] ? Jul 27 16:53:52 DocScrutinizer: a test if 2 is bigger than 1 Jul 27 16:54:09 with (( ? Jul 27 16:54:10 (( )) is like [[ ]] for numbers Jul 27 16:54:14 yes Jul 27 16:54:14 uhuh Jul 27 16:54:21 could just write 1 :p Jul 27 16:54:48 sorry, not feeling like obfuscated bash contest Jul 27 16:55:04 DocScrutinizer: just try (( 0>1 )); echo -n $?; (( 2>1 )); echo $? in both ash and bash. Jul 27 16:55:15 why? Jul 27 16:55:19 why is it obfuscated, it's completely normal Jul 27 16:55:29 it's nonsense Jul 27 16:55:33 why Jul 27 16:55:38 im not saying anything lol Jul 27 16:56:29 echo "10" Jul 27 16:56:31 done Jul 27 16:56:52 there's no defined purpose of this whole exercise Jul 27 16:57:10 what? Jul 27 16:57:11 i get 01 too Jul 27 16:58:35 mywiki.wooledge.org/ArithmeticExpression Jul 27 16:59:43 one can only assume ash lacks some of the functionaliy required to correcrlt parse that statement Jul 27 16:59:57 as the outputs are different Jul 27 17:00:00 is it even POSIX? Jul 27 17:00:15 00 under ash, 10 under bash Jul 27 17:00:26 no idea tbh Jul 27 17:01:10 I just wonder what DocScrutinizer meant Jul 27 17:01:45 prolly that this stuff is somewhere in the ash documentation maybe Jul 27 17:02:14 cant speak for Doc :) Jul 27 17:02:29 I think DocScrutinizer's point was "please give us your real use case, not a toy example of comparing two integers where we already know the answer" Jul 27 17:02:43 anyweh i need a reboot Jul 27 17:02:47 tally bai Jul 27 17:02:57 mgedmin: exactly Jul 27 17:03:06 missing the whole point Jul 27 17:03:32 it's well known that bash behaves differently to ash, that's why we use bash whenever we can Jul 27 17:03:49 even more so for using rather strange unusual syntax Jul 27 17:04:12 yes bash is better, specially with ctrl+r but ash is included in busybox Jul 27 17:04:21 yeah, I was just asking about this. and I wasn't aware that this is strange or unusual syntax as I've encountered it plenty of times Jul 27 17:04:44 actually I learned it from #bash in the first place Jul 27 17:04:56 I never heard of (( Jul 27 17:05:15 doesn't busybox have a broken ash? Jul 27 17:05:21 sure Jul 27 17:05:21 broken how? Jul 27 17:05:25 ~messybox Jul 27 17:05:26 messy... err busybox is meant for lean scripting. Regarding all the missing options and immanent limitations (see su) it's not really the interactive shell of choice. A lot of people hate busybox because a lot of system integrators don't understand the difference between busybox and a decent user interactive shell plus unix utils Jul 27 17:05:27 DocScrutinizer: now you have \o/ Jul 27 17:06:35 don't expect messybox shell understand all the more exotic syntax elements bash knows to handle Jul 27 17:07:08 when debian switched /bin/sh from bash to dash, there were lots of cries of pain Jul 27 17:07:18 esp calling subshells in messybox shell isn't exactly what you'd think it is Jul 27 17:07:41 but people learned to write shell scripts that only depend on POSIX shell features, or correctly specify #!/bin/bash when they use bash-only extensions Jul 27 17:08:39 mgedmin: the problem is messybox still replaces all the coreutils etc by useless crippled broken versions Jul 27 17:09:00 yeah, it's a pain Jul 27 17:09:07 especially for interactive use Jul 27 17:09:11 100% Jul 27 17:09:58 the whole idea to use messybox for interactive is so ill-minded... :-( Jul 27 17:10:36 (I don't think interactive shell usage was a big use case in the designer's minds when they were designing a *phone*) Jul 27 17:10:41 and debian seems have lost their mind Jul 27 17:11:01 Debian uses bash for interactive shells, in case you got confused Jul 27 17:11:13 aaah Jul 27 17:11:14 they switched to dash is for shell scripts because it's much faster Jul 27 17:11:25 it sped up things like ./configure and /etc/init.d/* Jul 27 17:11:44 for init dash/messybox is perfect Jul 27 17:12:04 unless you ponder editing the initscripts ;-P Jul 27 17:12:15 lennart pottering would disagree :) Jul 27 17:12:25 and suggest rewriting everything in C Jul 27 17:13:22 * Tronic <3 systemd Jul 27 17:13:36 mgedmin: you're aware fremantle enetered bootloop when replacing busybox by a decent shell plus coreutils, as the busybox syntax was not compatible to "regular" bash & corutils results Jul 27 17:14:13 I heard about problems from installing bash, never knew the details Jul 27 17:14:25 that's where busybox shell (incl all the builtin) is actually broken Jul 27 17:14:33 interesting; I assumed ash supported a strict subset of bash features Jul 27 17:14:41 no, it doesn't Jul 27 17:14:52 some cmds behave incompatible Jul 27 17:15:08 dash can be taken as a reasonable posix shell Jul 27 17:15:14 (don't ask me about details, freudian slip) Jul 27 17:15:16 ash from busybox... useless for everything Jul 27 17:16:04 ruskie: ACK Jul 27 17:17:05 I have a dream and in it is a Linux system with no shell scripts nor even a shell installed. Avoids the tcsh vs. bash war too. Jul 27 17:17:23 Tronic, how can I ssh into it if it doesn't have a shell? Jul 27 17:17:30 bash for scripting zsh for interactive FTW Jul 27 17:17:31 what use is a Linux system if I can't ssh into it? Jul 27 17:17:36 go for firefox OS then ;-P Jul 27 17:17:51 mgedmin, well you can do things like someone did it... used python for init and shell Jul 27 17:18:06 sounds awkward Jul 27 17:18:23 well his system was rather hosed at the time Jul 27 17:18:41 I bet it was - ROTFL Jul 27 17:19:24 maybe do it in lua, be original Jul 27 17:19:35 excel with adobe extensions for kernel!!! Jul 27 17:19:38 or ruby or perl or brainf*ck Jul 27 17:20:17 mgedmin: X forwarding and run a desktop. Jul 27 17:20:25 whitespace! what a nice *CLEAN* system Jul 27 17:20:27 :-P Jul 27 17:20:48 meh, o/ Jul 27 17:39:02 http://lurkmore.ru/images/e/ef/Attention_Norway.png Jul 27 17:41:10 texrat -> "Champion of the Month: Randy Arnold" Jul 27 17:41:34 recent Nokia developers newsletter Jul 27 17:42:26 What is a "champion"? Jul 27 17:42:34 Champion of what? What has he done? Jul 27 17:42:36 lol, dunno Jul 27 17:42:49 I gather it's something pleasant Jul 27 17:42:56 like? Jul 27 17:53:20 GPG error: https://downloads.maemo.nokia.com ./ Release: The following signatures couldn't be verified because the public key is not available: NO_PUBKEY ADB4438160A655EF Jul 27 18:25:42 isn't this a warning? Jul 27 18:28:08 DocScrutinizer: "GPG error" doesn't look like a warning Jul 27 18:28:29 btw, diablo repo is in same state for 2 years approx already :) Jul 27 18:35:15 any bugreport filed? Jul 27 18:35:59 against diablo? sure, in 2009 Jul 27 18:39:40 seems it's nonfatal anyway, just disabled all repos (including@ovi) except for Nokia*, and was able to install dataviz Jul 27 18:43:18 ~hostmode-charging Jul 27 18:43:19 it has been said that hostmode-powered is http://talk.maemo.org/showthread.php?p=921203#post921203 Jul 27 19:02:42 http://mg.pov.lt/maemo-irclog/%23maemo.2011-07-26.log.html#t2011-07-26T14:14:33 Jul 27 19:03:24 my evince install does save position :/ Jul 27 19:08:22 ?? Jul 27 19:11:23 Sicelo: aah, ok. Well probably evince actually does. seems arkenoi wasn't talking particularly about evince, but about any pdf reader that does both fullscreen and save pos Jul 27 19:11:49 Sicelo: and why :/ ? Jul 27 19:12:36 no particular reason.. guess i misunderstood him Jul 27 20:05:17 Hay, is there a way to change the Wlan settings via dbus? Jul 27 20:26:20 ruskie: is your repo down? https://repo.codemages.net/ Jul 27 20:26:38 hmm Jul 27 20:26:40 shouldn't be Jul 27 20:26:53 it doesn't do an index though Jul 27 20:27:12 ruskie: ah okay. Jul 27 20:27:56 hmm Jul 27 20:28:04 ruskie: mh. on your wiki page, rxvt-unicode is not listed as in your repo. Is it still there? if yes, how can I download your source package without using apt? Jul 27 20:28:41 hmm I don't have all sources Jul 27 20:29:13 it's basically just a rebuild of the debian package Jul 27 20:29:36 ruskie: I wondered about that, since I can't find a debian package that doesn't have dependencies which are way beyond what maemo has. Jul 27 20:29:45 don't think I modified it at all Jul 27 20:29:52 hmm Jul 27 20:29:59 what is there that shouldn't be? Jul 27 20:30:04 note I'm not really using it Jul 27 20:31:13 hmm I might have just ignored the deps Jul 27 20:31:28 it was an early package Jul 27 20:31:41 ruskie: well even the lenny package http://packages.debian.org/lenny/rxvt-unicode requires libc >=2.7 for example, whereas maemo has 2.5.1. Jul 27 20:31:56 Depends: libc6 (>= 2.5.0-1) <-- that's in mine Jul 27 20:33:18 though sadly I don't have the sources anywhere anymore Jul 27 20:33:19 yeah that sounds about right. I can't find any older debian packages than the lenny one which requires newer stuff. Jul 27 20:33:59 ah wait, I found http://snapshot.debian.org/package/rxvt-unicode/ Jul 27 20:34:11 deb-src http://archive.debian.org/debian-archive/debian/ etch main <-- that's what I have in my apt Jul 27 20:34:13 ruskie: do you remember which urxvt version you compiled? Jul 27 20:34:25 that's easy ;) Jul 27 20:34:47 rxvt-unicode_7.9-2_armel.deb Jul 27 20:35:10 ruskie: thanks Jul 27 20:35:36 note though that changing the depend version is easy anyway Jul 27 20:36:03 I highly doubt that rxvt-unicode requires anything from a more recent glibc Jul 27 20:37:08 I see what happens. I'm quite clueless but my goal is having perl support. Jul 27 20:37:26 ahh Jul 27 20:37:59 basically you can edit debian/control file and adjust the dep versions there Jul 27 20:39:35 hi! anybody there, who can help me? Jul 27 20:42:32 i have problems with my mac address. Jul 27 20:43:01 it changes every reboot. Jul 27 20:47:40 Mh. Is there some other way to unpack a .deb other than ar? Jul 27 20:48:05 yes Jul 27 20:48:39 dpkg-deb -x filename.deb targetdir/ Jul 27 20:48:46 mgedmin: yes? I can't find any other way, everyone's suggesting ar on google. Or dpkg -x but that doesn't extract the control files. Jul 27 20:49:01 for manual peeking I like mc, which lets me enter and view a deb's contents when I press Enter on it Jul 27 20:49:24 dpkg-deb -e Jul 27 20:49:27 will extract control files Jul 27 20:49:40 no idea for my mac problem? Jul 27 20:49:43 dpkg -e/-x is the same as dpkg-deb -e/-x Jul 27 20:50:02 mgedmin: ah thank you Jul 27 20:50:21 for the record, I'd no clue about these options other than a strong suspicion that they must exist Jul 27 20:50:25 so I read the dpkg man page :) Jul 27 20:50:39 mgedmin: which doesn't exist on the n900 :( But yeah I could have googled the man page. thank you. Jul 27 20:50:51 oh, your main computer is not a linux box? Jul 27 20:50:53 oh wait. I installed the man-db on the n900. I forgot. Jul 27 20:51:12 mgedmin: yes it is, but it doesn't use deb packaging. Jul 27 20:55:00 hm, but fbreader creates a new ZLQtPaintContext in paintEvent() Jul 27 20:55:05 so it can't be using an old pixmap Jul 27 20:56:06 or does it? Jul 27 20:57:03 no, I'm wrong, it uses a long-lived paint context Jul 27 20:57:51 mh. How do I tell dpkg-deb to use /usr/bin/gnu/tar instead of busybox gnu? dpkg-deb -b gives me a funny error, tar: unrecognized option `--format=gnu' BusyBox blablabla Jul 27 21:00:23 unloading and loading the wifi module causes the mac change. Jul 27 21:02:31 Shapeshifter, try PATH=/usr/bin/gnu:$PATH dpkg-deb ... Jul 27 21:03:30 mgedmin: thanks, that worked. didn't think of PATH... Jul 27 21:10:50 Is there a way to change the wlan-power? In the connection settings you can change the power-save mode. Is there a way to change it via python/dbus? Jul 27 21:12:19 mh. that's too bad, urxvt with perl from debian got compiled with 'libafterimage' support. Jul 27 21:15:47 hi.. could someone help me with packaging app to maemo (N900)? Jul 27 21:16:24 i can make deb package, install goes successfully but app binary is missing Jul 27 21:17:22 when i try to test it with simulator, I get this: "chmod: /opt/usr/bin/App: No such file or directory" Jul 27 21:24:37 is there a way to extract WI-FI passwords from a stock N810? Jul 27 21:24:48 can anybody help me? Jul 27 21:27:04 hm, what would be trying to run chmod? Jul 27 21:27:34 No such file or directory Jul 27 21:27:47 i make this at QtCreator @ Windows Jul 27 21:30:18 ShadowJK, here's compile.. http://pastebin.com/Xgf4bJDC Jul 27 21:32:52 hi all Jul 27 21:34:01 I'm not familiar with qt creator, but it looks almost like it failed due to missing perl? That's wtf Jul 27 21:34:47 im trying to installa the last version of Kasvopus on my N900, but it request libqtm-11-declarative e mcsp . how to solve this? Jul 27 21:36:16 * ShadowJK doesn't see kasvopus in extras Jul 27 21:41:09 so I just installed the maemo sdk again. now, after the sbox_ctl start and login, it says You dont have active target in scratchbox chroot. and that I should run sb-menu. Runnig sb-menu just hangs, and doing sb-conf select FREMANTLE_ARMEL like the wiki says gives me sb-conf: No such target: FREMANTLE_ARMEL Jul 27 21:43:28 I'm not familiar with qt creator, but it looks almost like it failed due to missing perl? That's wtf Jul 27 21:43:37 Windows, think it's mingw Jul 27 21:44:47 giardia, in that case you need to manually install the dependencies with apt-get first, I guess.. Jul 27 21:48:03 The reason for having repositories is that dependencies are figured out automatically, and a convenient place to find everything :) Jul 27 21:48:33 Or is it normal that sb-menu just makes python eat some cpu for half an eternity? Jul 27 21:51:51 Oh there we go. Had to prepend TERM=xterm. They even reported it as a bug... https://bugs.maemo.org/show_bug.cgi?id=8258 terrible. Jul 27 21:51:52 04Bug 8258: sb-menu freezes with an unknown TERM Jul 27 21:54:32 argh and the wiki doesn't say anything about what to pick when creating a target Jul 27 21:54:38 and I can't even remember doing this last time. Jul 27 22:55:45 Anybody discussing the N9 seconds video clues? Jul 27 22:59:43 in<< Jul 27 23:00:12 Okay. Jul 27 23:00:23 Well, actually let's go into #n9seconds Jul 27 23:00:46 I'm sure nobody here cares. Jul 27 23:01:14 lol, i bet some do Jul 27 23:01:44 Whenver the dashboard is shown, monocle.com is always an open browser window. Jul 27 23:22:26 Hi, how should I create finger-scrollable controls(QTextEdit) with Qt? Jul 27 23:38:21 oh for crying out loud, what a piece of garbage scratchbox is Jul 27 23:41:52 a normal 'dpkg-buildpackage -rfakeroot' to build rxvt-unicode gives me some crap about '/scratchbox/tools/lib/perl5/5.8.4/i686-linux-thread-multi/auto/DynaLoader/DynaLoader.a: could not read symbols: File in wrong format'. Then I read that it has something to do with speed optimizations in crapbox... https://bugs.maemo.org/show_bug.cgi?id=1954 All they suggest, including SBOX_REDIRECT_IGNORE simply doesn't work. So it uses ... Jul 27 23:41:53 04Bug 1954: Scratchbox makes building extensions to target Perl, Python etc awkward Jul 27 23:41:58 This is incredibly messed up and useless. Jul 27 23:42:01 ... /scratchbox/tools/bin/perl which is "This is perl, v5.8.4 built for i686-linux-thread-multi", which is useless for building for armel. But there's /usr/bin/perl, which is "This is perl, v5.8.3 built for arm-linux-gnueabi-thread-multi", yet if I simply do export PATH=/usr/bin:$PATH so that this comes first, then configure complains "checking for /usr/bin/perl suitability... configure: error: no working perl found, or perl not ... Jul 27 23:42:07 ... version >= 5.8" which is bullshit anyway, as it's clearly 5.8.3 Jul 27 23:49:57 Okay, I'm here. (courtesy of docscrutinizer) Jul 27 23:51:19 ~docscrutinizer Jul 27 23:51:20 methinks docscrutinizer is jOERG, a HW-developer and engineer of Openmoko Jul 27 23:51:20 infobot: asl? Jul 27 23:51:21 Old enough, yes please, my place. Jul 28 00:04:03 good morning Jul 28 00:07:10 Termana: morning Jul 28 00:34:36 scratchbox is killing me. Jul 28 00:36:39 wait til you have the pleasure to fight aegis Jul 28 00:37:27 what's that? Jul 28 00:37:50 heh Jul 28 00:37:55 ~aegis Jul 28 00:37:56 methinks aegis is http://www.developer.nokia.com/Community/Wiki/Harmattan:Developer_Library/Developing_for_Harmattan/Harmattan_security/Security_guide , or "The purpose of this framework is: ... to make sure that the platform meets the requirements set by third party software that requires a safe execution environment.", or http://en.wikipedia.org/wiki/Trusted_Computing#Criticism Jul 28 00:38:00 Is here anybody porting python games? Jul 28 00:38:37 sounds great. Though "luckily", Nokia won't sell me an N950 and I'm not at all interested in the N9, so I'll never experience the joy. Jul 28 00:39:03 though that's great, tc in a mobile phone, that's really something we need, isn't it. Jul 28 00:39:07 Shapeshifter: it's the thing that, when you edit a python script on device that's been installed with a manifest (list of required permissions), simply nukes your system Jul 28 00:39:36 Fun Jul 28 00:39:45 Fun for many hours I assume Jul 28 00:39:49 yeah, extreme fun Jul 28 00:40:28 looks like n900 is the first and last nokia smartphone that doesnt need to be `jailbroken'... :/ Jul 28 00:40:59 though I have to say, developing for the iphone is a lot less painful Jul 28 00:41:15 well, I mean the SDK part. Jul 28 00:41:24 and even more fun there's not even a proper manpage or anything explaining if and how you could query if a file is protected that way. Just edit it and either it works or you get a "please reflash!" requester after the instant reboot Jul 28 00:41:37 DocScrutinizer: oh wow. Jul 28 00:43:42 tho that wouldnt be the first undocumented linux thing :P Jul 28 00:49:14 * nox- feels evil linking to another rant: http://www.shiningsilence.com/dbsdlog/2010/04/09/5713.html Jul 28 00:50:31 But seriously, how am I supposed to compile this now Jul 28 00:50:32 it's impossible Jul 28 00:55:27 either this happens http://pastie.org/2282229 (trying to use arm perl) or this happens http://pastie.org/2282237 (default, using host perl) Jul 28 01:00:05 Shapeshifter: what does config.log say then? Jul 28 01:01:59 urxvt apparently wants to link to libperl, so the arm perl is probably the one to work on Jul 28 01:04:09 villager: what from config.log do you need to know? It says a bunch of stuff about perl, such as PERLLIB='-Wl,-E -L/usr/local/lib /scratchbox/tools/lib/perl5/5.8.4/i686-linux-thread-multi/auto/DynaLoader/DynaLoader.a -L/scratchbox/tools/lib/perl5/5.8.4/i686-linux-thread-multi/CORE -lperl -lnsl -ldl -lm -lcrypt -lutil -lpthread -lc' Jul 28 01:04:31 in the host perl case. let me switch to the other Jul 28 01:04:33 Shapeshifter: I assume the arm libperl-dev actually is installed? Jul 28 01:04:59 Shapeshifter: since it needs to link to libperl, you can't use the host one Jul 28 01:05:05 villager: yes, everthing listed by apt-cache search extutils is installed, including that one Jul 28 01:05:43 well then config.log just says configure:11803: error: no working perl found, or perl not version >= 5.8. it doesn't get any farther than that. Jul 28 01:05:57 Shapeshifter: anything before that? perhaps you can put the config.log on that pastebin Jul 28 01:08:08 villager: http://paste.xinu.at/eAj/ here's config.log Jul 28 01:08:31 hmm Jul 28 01:08:51 here's configure http://paste.xinu.at/ejDt/ somewhere around line 11644 it starts detecting perl Jul 28 01:09:04 Shapeshifter: can I get the source code from somewhere? Jul 28 01:09:16 in 11729 it does if $PERL -MExtUtils::Embed -e "use v5.8" which is the reason it fails Jul 28 01:09:38 configure.ac is usually slightly easier to read than configure Jul 28 01:09:55 villager: I got the source from here http://packages.debian.org/lenny/rxvt-unicode (right side, all three source packages, extracting the dsc) Jul 28 01:10:08 http://paste.xinu.at/i4ED/ configure.ac Jul 28 01:10:25 Shapeshifter: okay, and if you use that command manually in scratchbox, does it say anything interesting? Jul 28 01:10:51 villager: Can't locate ExtUtils/Embed.pm in @INC (@INC contains: /etc/perl /usr/local/lib/perl/5.8.3 /usr/local/share/perl/5.8.3 /usr/lib/perl5 /usr/share/perl5 /usr/lib/perl/5.8 /usr/share/perl/5.8 /usr/local/lib/site_perl .). Jul 28 01:10:55 BEGIN failed--compilation aborted. Jul 28 01:11:15 but ExtUtils/Embed.pm is in no maemo package. Jul 28 01:11:19 or at least I can't find it. Jul 28 01:11:35 and I tried installing it through cpan but it asked me a billion questions I wasn't able to answer. Jul 28 01:11:37 hmm Jul 28 01:11:55 Shapeshifter: can't you just accept the defaults? Jul 28 01:12:07 usually, if you don't know the answer, the default works for you :p Jul 28 01:12:09 wblaze: I tried that but it doesn't work. Jul 28 01:12:18 ...strange Jul 28 01:12:21 Shapeshifter: should be in the perl-modules package Jul 28 01:12:25 nah with scratchbox, it kinda gets values from the host system and it's all messed up Jul 28 01:12:47 villager: mh you are right Jul 28 01:12:49 /usr/share/perl/5.8.3/ExtUtils/Embed.pm Jul 28 01:13:07 that explains it Jul 28 01:13:14 @INC contains: /etc/perl /usr/local/lib/perl/5.8.3 /usr/local/share/perl/5.8.3 /usr/lib/perl5 /usr/share/perl5 /usr/lib/perl/5.8 /usr/share/perl/5.8 /usr/local/lib/site_perl Jul 28 01:13:36 mh. no it doesn't. @INC contains /usr/local/share/perl/5.8.3 yet it can't find it Jul 28 01:13:50 Shapeshifter: in my scratchbox there's a symlink from 5.8.3 to 5.8, do you have it? Jul 28 01:15:09 lrwxrwxrwx 1 ovek 1000 5 Nov 21 2010 5.8 -> 5.8.3 Jul 28 01:15:10 mh, the problem is the 'local' Jul 28 01:15:32 nah, the /usr/share/perl/5.8 in the @INC should match it if that symlink exists Jul 28 01:15:36 @INC contains /usr/local/share/perl/5.8.3 but all the stuff is in /usr/share/perl/5.8.3 Jul 28 01:16:12 well /usr/local/share only contains man and no symlinks. Jul 28 01:16:26 you should be looking in /usr/share/perl, not in /usr/local Jul 28 01:16:51 I didn't define that INC Jul 28 01:17:02 I know, it's compiled-in default Jul 28 01:17:08 you can't and shouldn't change it Jul 28 01:18:08 so you should just ensure things are as it expects - i.e. like I said, that /usr/share/perl/5.8 is a symlink that points where it should Jul 28 01:18:36 lrwxrwxrwx 1 shapeshifter wheel 5 Jul 28 00:06 /usr/share/perl/5.8 -> 5.8.3 yes I have that symlink Jul 28 01:18:50 okay Jul 28 01:19:55 /usr/share/perl/5.8.3/ExtUtils/Embed.pm exists, /usr/share/perl/5.8 is in INC and the symlink exists. Jul 28 01:21:12 I'll try it here, let's see Jul 28 01:22:47 villager: The only changes I made to the debian package is in debian/rules, I replaced --enable-afterimage by --disable-afterimage and I removed the afterimage dependency from debian/control Jul 28 01:23:14 I also replaced xutils by xutils-dev in debian/control Jul 28 01:25:00 it works in my scratchbox at least Jul 28 01:25:12 hmm, why not in yours then Jul 28 01:25:22 oh that is strange. Jul 28 01:25:55 http://pastie.org/2282355 something in my env? Jul 28 01:26:29 of course I might have a newer scratchbox because I installed the harmattan sdk, but I don't think it should matter that much Jul 28 01:26:51 mh, I just installed the maemo 5 sdk freshly now Jul 28 01:26:52 the fremantle target should still work as before for me I think Jul 28 01:27:12 at least I've still managed to compile stuff with it since Jul 28 01:27:35 The question is, why does perl -MExtUtils::Embed -e "use v5.8.3" give the Can't locate ExtUtils/Embed.pm in @INC error Jul 28 01:28:21 did you also set export SBOX_REDIRECT_IGNORE=/usr/bin/perl in etc/profile? Jul 28 01:28:22 Shapeshifter: let's just make sure: ls -l /usr/share/perl/5.8/ExtUtils/Embed.pm gives? Jul 28 01:28:45 lrwxrwxrwx 1 shapeshifter wheel 39 Jul 28 00:06 /usr/share/perl/5.8/ExtUtils/Embed.pm -> /opt/share/perl/5.8.3/ExtUtils/Embed.pm Jul 28 01:28:46 Shapeshifter: nah, just did it in my shell before starting perl Jul 28 01:28:57 ls: /opt/share/perl/5.8.3/ExtUtils/Embed.pm: No such file or directory Jul 28 01:28:59 Shapeshifter: ok, and that /opt thing also exists? Jul 28 01:29:12 Shapeshifter: guess not... it's there for me Jul 28 01:29:19 Shapeshifter: try apt-get install --reinstall perl-modules Jul 28 01:29:47 well... Jul 28 01:30:16 though I suppose other things might be missing too Jul 28 01:30:33 if that file's not there, who knows what else is gone Jul 28 01:30:43 I think I know why they're not there. The wiki told me to rm -rf /targets/FREMANTLE_ARMEL/opt; mkdir /targets/FREMANTLE_ARMEL/opt http://wiki.maemo.org/Documentation/Maemo5_Final_Installation#On_x86_and_x86_64_Debian-based_systems Jul 28 01:31:15 it does? Jul 28 01:31:56 find rm -rf on that wiki page, the first 2 occurences Jul 28 01:32:57 villager: thanks man. Jul 28 01:33:07 still waiting for the webpage to load... this computer is apparently heavily loaded right now Jul 28 01:33:23 or maybe the wiki is slow Jul 28 01:34:10 must be the wiki server, everything else on my computer seems to work fine Jul 28 01:34:28 hm, hitting reload seemed to fix it Jul 28 01:35:11 ah I see Jul 28 01:35:33 I'd rather they suggest using mv to move /opt rather than delete it... Jul 28 01:35:56 well it works. yay. Jul 28 01:36:21 well, maybe they ask you to reinstall the rootstraps afterward?... that sb-conf might do that Jul 28 01:36:26 now finally I can go on with some tweaks to make it usable on the n900. or rather, time to go to bed. half past 3 in the morning. Jul 28 01:36:41 yes I should reinstall everything Jul 28 01:36:43 tomorrow... Jul 28 01:36:46 you did that sb-conf rs for both x86 and armel afterwards? Jul 28 01:37:02 if not it might explain stuff Jul 28 01:37:19 afterwards? no that comes before the rm -rf Jul 28 01:37:37 first it's the whole thing for X86, then rm -rf, then again for ARMEL Jul 28 01:37:39 there's one after it too Jul 28 01:38:00 I see that mkdir /targets/FREMANTLE_ARMEL/opt is the last thing in that chapter Jul 28 01:38:18 the other one is mkdir /targets/FREMANTLE_X86/opt followed by sb-conf se FREMANTLE_ARMEL which is the other target Jul 28 01:38:42 hmm, guess those instructions ought to be clarified Jul 28 01:39:18 I don't really understand the explanation for that rm -rf, so I'd be the wrong person to do it. Jul 28 01:40:16 I gotta go to bed :< good night. Jul 28 01:40:18 and thanks Jul 28 01:44:26 hmm, those instructions seem a bit messed up... things are done in what appears to be an illogical order... Jul 28 01:46:52 or... maybe they're just outdated, this was written before perl-modules started using /opt, I bet Jul 28 01:52:37 maemo@maemo-desktop:/scratchbox/users/maemo/home/maemo/workspace/sharedlib$ make Jul 28 01:52:37 make: *** No rule to make target `prpltwtr_api.c', needed by `prpltwtr_api.o'. Stop. Jul 28 01:52:50 i am about to explode from anger Jul 28 01:53:18 anyone good with makefiles here? Jul 28 01:53:21 ;( Jul 28 02:30:09 fusi: is there a prpltwtr_api.c? Jul 28 02:30:49 thanks for replying - yea there is Jul 28 02:30:54 im a bit further along now Jul 28 02:30:59 learning as i go Jul 28 02:31:00 :/ Jul 28 02:31:12 correct dir? Jul 28 02:31:31 ive got it to find the files now, tit was the dir yea Jul 28 02:31:35 had to move stuff around Jul 28 02:31:47 ok Jul 28 02:32:02 just gotta coax it into seeing glib.h now Jul 28 02:32:16 its quite happily ignoring my include dits Jul 28 02:32:18 :( Jul 28 02:32:28 heh Jul 28 02:32:41 plain make or automake? Jul 28 02:33:03 most often this are path bugs Jul 28 02:33:08 nfi, im writing a file called Makefile and executing make in its dir Jul 28 02:33:33 going the manual way :P Jul 28 02:34:03 dont know of any other o_O Jul 28 02:34:11 im a total linux noob Jul 28 02:34:14 \o/ Jul 28 02:34:44 out of my warm and cozy wwin32 comfort zone Jul 28 02:34:45 windows uses nmake :P Jul 28 02:35:07 at least i did when i used it .. which ich 10+ years ago Jul 28 02:35:07 hidden behind useable gui's Jul 28 02:35:23 and nmake is really crap Jul 28 02:35:40 mmm Jul 28 02:35:52 well all build systems suck somehow Jul 28 02:36:03 just takes time to learn Jul 28 02:36:13 is yours project a qt project? Jul 28 02:36:30 im trying to make the twitter plugins for conversations work again Jul 28 02:36:35 -s Jul 28 02:36:35 if not you may consider autoconf/automake wich is de-facto standars (but sux too) Jul 28 02:37:06 plain make is nice for small things which you dont want to be portable across platforms Jul 28 02:37:41 aye thats what i want Jul 28 02:37:42 but it easily becomes complicated and stretches the limits when it starts to grow 'historically' Jul 28 02:37:47 mmm Jul 28 02:37:51 yea thats what ive got here Jul 28 02:38:00 ok Jul 28 02:38:03 i took the old codebase that works on maemo Jul 28 02:38:14 :) Jul 28 02:38:21 well go on rtfm :P Jul 28 02:38:28 and took somme newer source from mikeage (he's updating it for pidgin) Jul 28 02:38:44 and stuck it togeter into an unholy mess of woe Jul 28 02:38:48 heheh yea Jul 28 02:38:54 have been all night Jul 28 02:39:10 prolly not a good idea to tamper to much with it while you not know what you do :P Jul 28 02:39:15 gotta give it time to soak through my thick skull Jul 28 02:39:31 youre prolly right Jul 28 02:39:38 biting off too much Jul 28 02:40:05 yea Jul 28 02:40:11 i think ill sleep on it for now Jul 28 02:40:13 im sweepy Jul 28 02:40:31 yea Jul 28 02:40:32 im off Jul 28 02:40:35 thanks for the advice :) Jul 28 02:40:39 o/ nn Jul 28 02:40:44 cu **** ENDING LOGGING AT Thu Jul 28 02:59:57 2011