**** BEGIN LOGGING AT Fri Mar 26 02:59:57 2010
Mar 26 03:06:53 <dotblank> Is there a module-ladspa-sink built for PA on the n900
Mar 26 03:10:04 <dotblank> it shows the .so file in scratch box
Mar 26 03:56:17 <DocScrutinizer> [2010-03-26 01:28:06] <javispedro> TheNewAndy: also, install package pulseaudio-module-extra
Mar 26 03:56:19 <DocScrutinizer> [2010-03-26 01:28:15] <javispedro> TheAppleMan: it containts ladspa-sink at least
Mar 26 03:57:30 <TheNewAndy> (poor The AppleMan getting spammed by tab completion :))
Mar 26 03:57:41 * DocScrutinizer is wondering what strange TV-show might have created interest in ladspa
Mar 26 03:57:54 <TheNewAndy> javispedro and I are now chatting in #maemo :)
Mar 26 03:58:43 <TheNewAndy> I'm interested because I'm trying to do a demo of some audio processing at work, and ladspa+pulse is the biggest bang for buck
Mar 26 03:58:47 <DocScrutinizer> err, what?
Mar 26 04:00:10 <DocScrutinizer> I wasn't even aware there's a ladspa plugin for friggin PA. For ALSA there's been one since ages
Mar 26 04:00:44 <TheNewAndy> it is a bit limited (single channel only, can't adjust parameters on the fly)
Mar 26 04:01:31 <TheNewAndy> but it is great for my purposes (hacking up a demo on multiple platforms really quickly)
Mar 26 05:11:26 <LiraNuna> when my N900 receives a call and plays a custom ringtone, there is a pause when it's finished
Mar 26 05:11:34 <LiraNuna> how can I make it loop without the pause?
Mar 26 05:13:30 <crashanddie> Pray Baahl and slit the troath of a 12 year old elf virgin
Mar 26 05:13:56 <jaska> bhaaaaaal
Mar 26 05:24:08 <ponyofde1th> hi, how do i mount the hdsd card in linux seems when i enter usb storage mode i only see the internal eMMC
Mar 26 05:28:50 <MiXu-> afaik it should mount automatically unless something in the phone is using the mmc all the time
Mar 26 05:29:55 <MiXu-> in which case it will complain "MMC is in use" or something like that when you plug it in
Mar 26 05:45:07 <acidjazz> just launched my 2nd game http://apps.facebook.com/luckyjackpot/
Mar 26 05:46:54 <crashanddie_> the next person who promotes his facebook app in here gets banned for life
Mar 26 06:10:16 <Stskeeps> work mornings are not the time for my laptop to start running a forced fsck.
Mar 26 06:10:27 <Stskeeps> when will linux have ZFS.
Mar 26 06:12:58 <crashanddie_> Stskeeps: the same year as the year of the Linux desktop!
Mar 26 06:13:04 <simula> heh
Mar 26 06:13:25 * simula is happy with linux in his pocket
Mar 26 06:18:03 <Stskeeps> simula: pocket is fine, it doesn't do forced fscks :P
Mar 26 06:19:31 <simula> did you pull the power plug stskeeps?  what forced fsck?
Mar 26 06:20:59 <Stskeeps> simula: 41 times without a check
Mar 26 06:21:04 <simula> ahhhh
Mar 26 06:21:16 <Stskeeps> yes, i know i can adjust that
Mar 26 06:21:27 <Stskeeps> but no, i always shut down nicely
**** BEGIN LOGGING AT Fri Mar 26 13:06:14 2010
Mar 26 13:06:16 <ShadowJK> battery management entity, or something like that
Mar 26 13:06:28 <Lumpio-> k
Mar 26 13:06:28 <SpeedEvil> smartreflex is a cpu thing
Mar 26 13:06:38 <ShadowJK> ptl: it feels practical for my use
Mar 26 13:07:30 <ptl> I'd but it if hadn't already bought a protective case for the N900...
Mar 26 13:08:20 <ptl> ShadowJK: thanks, but maybe it's not this page, because it doesn't have any change by Stskeeps on the history...
Mar 26 13:08:32 <ShadowJK> hm
**** ENDING LOGGING AT Fri Mar 26 13:09:37 2010

**** BEGIN LOGGING AT Fri Mar 26 13:12:15 2010
Mar 26 13:14:50 <DocScrutinizer> corecode_: there's no need for a loophole. We just need the kernel drivers for usb-host protocol, then we should be fine with manually enabled unpowred hostmode (the mode that needs a powered hub or y-cable)
Mar 26 13:15:12 <corecode_> ah
Mar 26 13:15:19 <corecode_> no way to power the port?
Mar 26 13:15:40 <SpeedEvil> corecode_, I have regrettably had no time. Spending most of my time digging. http://www.flickr.com/photos/14560445@N08/
Mar 26 13:15:44 <DocScrutinizer> corecode_: if we manage to switch on the GAIA chargepump, and the schematics are somewhat wrong on a missing capacitor, we even might get power
Mar 26 13:16:48 <FIQ> i don't think that google is down so... how to kill the web browser?
Mar 26 13:16:50 <FIQ> completly
Mar 26 13:17:02 <FIQ> not just close some windows
Mar 26 13:17:05 <DocScrutinizer> corecode_: worst thing to throw stick on our legs is PHY chip not switching between pullup and pulldown resistors
Mar 26 13:17:14 <Jaffa> FIQ: Power menu > close active task?
Mar 26 13:17:17 <corecode_> SpeedEvil: what are you doing?
Mar 26 13:17:40 <corecode_> is that house too low or are you filling up the sides
Mar 26 13:17:44 <DocScrutinizer> corecode_: but that wouldn't be a major showstopper
Mar 26 13:18:14 <corecode_> FIQ: xterm + killall browser browserd
Mar 26 13:18:22 <FIQ> ty
Mar 26 13:18:31 <corecode_> DocScrutinizer: yea that would suck
Mar 26 13:18:35 <corecode_> kudos
Mar 26 13:18:42 <FIQ> jaffa: that didn't help at first
Mar 26 13:18:49 <FIQ> did it 2 times
Mar 26 13:18:53 <SpeedEvil> corecode_, I am extedndcing the lawn.
Mar 26 13:18:55 <FIQ> worked now though
Mar 26 13:19:03 <Dassu> =)
Mar 26 13:19:08 <DocScrutinizer> corecode_: I mean, we got Openmoko Frerunner charging hostmode running, where we seen same problem of pullup/down set wrong
Mar 26 13:19:24 <SpeedEvil> There was much unremoved spoil when digging pond.
Mar 26 13:19:43 <SpeedEvil> FIQ, see dsmetololl
Mar 26 13:20:04 <SpeedEvil> FIQ, just killing browserd will trigger reboots.
Mar 26 13:20:25 <DocScrutinizer> SpeedEvil: oh? never seen that
Mar 26 13:21:29 <corecode_> fyi, those fake nokia CA-101 cables from hong kong for $1 work perfectly
Mar 26 13:21:41 <corecode_> finally enough micro-usb cables around
Mar 26 13:22:32 <SpeedEvil> lol.
Mar 26 13:22:48 <SpeedEvil> just got my new screen protectors.
Mar 26 13:22:48 <corecode_> no, serious
Mar 26 13:23:00 <SpeedEvil> KY screen protectors.
Mar 26 13:23:03 <corecode_> ah, my screen protector is quite good too
Mar 26 13:23:10 <SpeedEvil> they are _slidy_
Mar 26 13:23:14 <corecode_> it is that 3m finnish sounding name
Mar 26 13:23:20 <corecode_> kiiimotsu or so
Mar 26 13:23:21 <SpeedEvil> (not really)
Mar 26 13:23:32 <corecode_> vinntatu?
Mar 26 13:23:38 <corecode_> something like that
Mar 26 13:24:14 <corecode_> SpeedEvil: how come the lawn is on the first floor level?
Mar 26 13:24:24 <corecode_> raised lawn?
Mar 26 13:24:31 <corecode_> heated from below?
Mar 26 13:24:38 <SpeedEvil> corecode_, house built on slope.
Mar 26 13:24:46 <corecode_> ah
Mar 26 13:24:48 <corecode_> cute
Mar 26 13:24:58 <corecode_> we have that a lot here
Mar 26 13:25:21 <corecode_> enter 0, exit 12, both "ground floor"
Mar 26 13:25:23 <corecode_> :)
Mar 26 13:25:56 <SpeedEvil> corecode. Only 2m difference.
Mar 26 13:26:13 <ShadowJK> bleh, now dsp video decode stopped working alltogether
Mar 26 13:26:20 <ShadowJK> more fun stuff in dmesg
Mar 26 13:26:24 <SpeedEvil> I should be finishing off toiday, but 6c and rain
Mar 26 13:26:33 <SpeedEvil> So am back in bed.
Mar 26 13:26:49 <thresh> wazd: yeah i've found his phone
Mar 26 13:26:53 <thresh> not sure i'll be there at 7pm though
Mar 26 13:27:08 <ShadowJK> but the most annoying thing is that the media player is failing silently
Mar 26 13:27:21 <ShadowJK> I want error messages if there's an error!
Mar 26 13:27:46 <wazd> thresh: cool, I'm heading there right now to meet Vlad, so we'll be wandering around for a while :)
Mar 26 13:31:51 * thresh has no idea who Vlad is
Mar 26 13:32:48 <ptl> Vlad Tepes, of course
Mar 26 13:32:54 <ptl> aka Vlad The Impaler
Mar 26 13:32:58 <ptl> aka Count Dracula.
Mar 26 13:33:11 <MiXu-> Anyone know if there's a way to easily disable prefetching Ovi Maps in bootup?
Mar 26 13:33:36 <MiXu-> I don't use it. It's just eating mah memory.
Mar 26 13:34:12 <thresh> ptl: ah so he's now in Moscow, neat!
Mar 26 13:38:37 <MiXu-> Oh wait. It's not even prefetched :o
Mar 26 13:38:41 <MiXu-> Never mind :)
Mar 26 13:42:42 * noobmonk3y just had steak n chips and is stuffed..........
Mar 26 13:43:48 <noobmonk3y> wtf is up with rmt ... like a yoyo!
Mar 26 13:50:38 <Ijan> zaheerm i would like to pm you
Mar 26 13:50:41 <Ijan> is it okay?
Mar 26 13:58:17 <Gizmokid2005> is there any way to add a .ics subscription to the N900 calendar? I know you can import events from an ics file, but is there any way to make it an actual subscription so it gets refreshed?
Mar 26 14:01:12 <noobmonk3y> sorry Gizmokid2005 - no idea
Mar 26 14:01:48 <Gizmokid2005> noobmonk3y: it's not looking promising from what I can tell...it really doesn't make sense, you'd think that would be a simple and logical feature to have.
Mar 26 14:01:56 <cvandonderen> hi
Mar 26 14:02:26 <noobmonk3y> hehe, no idea what an ics file is even, but assuming it is a load of apopintments that can be imported to the calendar?
Mar 26 14:02:36 <noobmonk3y> if so a sync would just be a deamon checking somewhere for updates?
Mar 26 14:03:48 <Gizmokid2005> noobmonk3y: http://en.wikipedia.org/wiki/ICalendar <-- yeah, basically would just be refreshing the ics based on a specific refresh interval
Mar 26 14:03:53 <Gizmokid2005> would be considering it a "subscription"
Mar 26 14:05:15 <noobmonk3y> i would assume it could be done, probably a request thread over at T.M.O?
Mar 26 14:05:59 <Gizmokid2005> noobmonk3y: that's what I was thinking, but I figured I'd check to see if maybe I was just dense and missed something that's there...
Mar 26 14:06:09 <noobmonk3y> lol doesnt sound dense at all :D
Mar 26 14:07:22 <Gizmokid2005> I just find it hard to believe that with so many people this is such an overlooked "feature"
Mar 26 14:13:34 <thresh> wazd_e63: now who's Vlad?
Mar 26 14:19:26 <Gizmokid2005> noobmonk3y: looks like http://maemo.org/community/brainstorm/view/improving_the_calendar_in_n900/ #6 & 7 are what I'm referring to.
Mar 26 14:20:01 <noobmonk3y> wow..... big ol' brainstorm!
Mar 26 14:20:14 <Gizmokid2005> Yep!
Mar 26 14:22:54 <wazd_e63> Thresh: omweather admin :)
Mar 26 14:26:49 <kyle__> hiyoo
Mar 26 14:26:59 <atiti> hi
Mar 26 14:27:30 <pronto> i wonder if anyone is porting chromium to the n900
Mar 26 14:28:18 <marcusb> Why all the complexity with /opt and scripts and what not?  Any reason one can't just "mv /usr /home; ln -s /home/usr /usr" ??
Mar 26 14:29:54 <hrw> marcusb: emmc is fscking slow
Mar 26 14:30:09 <hrw> compared to onenand flash chip
Mar 26 14:30:39 <marcusb> ok
Mar 26 14:32:35 <AntiXpucT> [17:13:12] CTCP VERSION ответ от Skim[n900] [~kvirc@n900.antixpuct.ru]: KVIrc 4.0.0 svn-4222-N900 'Insomnia' 20100125 - build 2010-03-26 12:25:27 UTC - Linux (2.6.28.10maemo-omap1)
Mar 26 14:32:53 <AntiXpucT> (i've ported KVIrc to N900) :)
Mar 26 14:33:54 <hrw> AntiXpucT: I do not trust ctcp version
Mar 26 14:34:42 <ccooke> marcusb: The other reason is that there are things in /usr needed before /home is available
Mar 26 14:35:01 <AntiXpucT> hrw, then wait some time, while i rebuild it with new paths and optify packages ;)
Mar 26 14:35:08 <hrw> AntiXpucT: today irc clients allows to set anything
Mar 26 14:35:12 <AntiXpucT> and i'll give you a links
Mar 26 14:35:46 <marcusb> ccooke: oops
Mar 26 14:35:53 <AntiXpucT> hrw, but i kill 3 days for porting, and i knew, that it is real version ;)
Mar 26 14:35:57 <hrw> AntiXpucT: client which I normally use is already built for maemo5 but I do not use it
Mar 26 14:37:13 <hrw> AntiXpucT: 15:33 < hrw> AntiXpucT: I do not trust ctcp version
Mar 26 14:37:19 <hrw> Irssi 0.8.14 (20090728) - http://irssi.org/
Mar 26 14:42:30 <Vratha_> i spy on people talking on irc using tcpdump
Mar 26 14:43:26 <corecode_> what's that "backup" app people keep talking about
Mar 26 14:43:40 <andrewfblack> What is up with some people taking the same theme changing background change hue and calling it a new theme?
Mar 26 14:43:41 * SpeedEvil dumps on people using tcp.
Mar 26 14:44:00 <corecode_> oh talking about background
Mar 26 14:44:03 <Stskeeps> andrewfblack: open source?
Mar 26 14:44:04 <Stskeeps> :P
Mar 26 14:44:09 <Vratha_> corecode_: ubuntu one.
Mar 26 14:44:11 <corecode_> http://panorama.epfl.ch/index.php?&width=1920&height=1200
Mar 26 14:44:13 <petteri> uh.. I was hoping fore pr 1.2 today
Mar 26 14:44:21 <corecode_> i wanted to use that as background
Mar 26 14:44:25 <corecode_> dynamically
Mar 26 14:44:39 <Vratha_> s/\./?/
Mar 26 14:44:42 <corecode_> Vratha_: is there a readymade app?
Mar 26 14:44:45 <andrewfblack> Stskeeps: I'm not naming a name but their is one theme designer that drives me crasy with his slight mods to the same theme over and over
Mar 26 14:44:55 <Vratha_> corecode_: on ubuntu there is
Mar 26 14:45:13 <Vratha_> but i don't use desktop linux
Mar 26 14:45:21 <corecode_> Vratha_: but people use it for maemo, or so
Mar 26 14:45:26 <corecode_> also one is suprslo
Mar 26 14:45:30 <Vratha_> so i haven't used it
Mar 26 14:45:40 <Vratha_> don't know about maemo
Mar 26 14:46:04 <Vratha_> thought you meant generally
Mar 26 14:46:42 <andrewfblack> corecode_: what you mean dynamically you want it to move like on website?
Mar 26 14:47:56 <corecode_> andrewfblack: no, the picture updates every 10 minutes or so
Mar 26 14:47:59 * andrewfblack needs inkscape on his n900 or thumb drive
Mar 26 14:48:53 <corecode_> problem is also that the screen is too small
Mar 26 14:49:01 <corecode_> can't fit the 360 deg panorama
Mar 26 14:49:25 <corecode_> and i'm only using 2 screens
Mar 26 14:50:04 <corecode_> anybody set up snowy yet?  ubuntu one notes is quite slow
Mar 26 14:51:06 <Stskeeps> TomaszD: got PLA, ta
Mar 26 14:51:50 <andrewfblack> corecode_: Your most likly going to have to wait for PR1.3 I hear we should be getting html backgrounds then and you will be able to do that.  Also their is someone working on a 800x480 widget that will cover up your entire background.  It will let you do animations and stuff.
Mar 26 14:52:25 <andrewfblack> Stskeeps: Bubble Bobble will be in extras tonight
Mar 26 14:52:34 <andrewfblack> Theme that is
Mar 26 14:52:46 <gameprograma> Hey all.
Mar 26 14:52:47 <Stskeeps> cool :)
Mar 26 14:52:51 <jatt> which chat program do you use in your n900?
Mar 26 14:52:54 <Stskeeps> andrewfblack: you did see i update the template right?
Mar 26 14:53:03 <jatt> I use x-chat on my laptop is it available for maemo?
Mar 26 14:53:05 <gameprograma> Are there instructions for updating the SDK installation?
Mar 26 14:53:08 <Stskeeps> jatt: yes
Mar 26 14:53:20 <gameprograma> I only see the instructions for installing the SDK from scratch using the GUI installer.
Mar 26 14:53:57 <andrewfblack> Stskeeps: Yep Thanks
Mar 26 14:54:18 * andrewfblack wishes everyone could see how many PMs he gets a day about theme design
Mar 26 14:54:58 <andrewfblack> I got several ohh my god I can't build theme PMs when new hildon-theme-layout-5 was updated
Mar 26 14:55:06 <corecode_> andrewfblack: no, putting a jpg there would be fine
Mar 26 14:57:27 <gameprograma> Anyone? :(
Mar 26 14:59:10 <thresh> where's wazd when you need him
Mar 26 14:59:20 <andre__> gameprograma, URL?
Mar 26 14:59:20 <thresh> gameprograma: IIRC SDK are better reinstalled then updated
Mar 26 14:59:51 <gameprograma> thresh: That means I'll have to download all of about 1GB or so all over again. :/
Mar 26 15:00:15 <gameprograma> andre__: What URL? The SDK installation instructions?
Mar 26 15:00:30 <andre__> yes.
Mar 26 15:00:32 <MohammadAG> gameprograma, my SDK works and it was updated
Mar 26 15:01:09 <MohammadAG> andre__, http://wiki.maemo.org/Documentation/Maemo5_Final_Installation :)
Mar 26 15:01:52 <gameprograma> andre__: http://wiki.maemo.org/Documentation/Maemo5_Final_Installation
Mar 26 15:02:15 <MohammadAG> Need some help optifying a python source for py2deb
Mar 26 15:02:23 <gameprograma> MohammadAG: How carry out the update? Did you use the GUI installer?
Mar 26 15:02:35 <andre__> gameprograma: true, not mentioning updating. feel free to file a ticket against Dev Platform > Documentation in bugs.maemo.org
Mar 26 15:02:44 <MohammadAG> gameprograma, it doesn't really matter, to update just use apt-get dist-upgrade
Mar 26 15:02:54 <gameprograma> Got it.
Mar 26 15:02:59 <MohammadAG> as fakeroot of course
Mar 26 15:03:38 <TomaszD> Stskeeps, k, np
Mar 26 15:04:25 <marcusb> # apt-get install diff   => "Note, selecting busybox instead of diff." "busybox is already the newest version."   and then: # diff  => -sh: diff: not foun
Mar 26 15:04:25 <gameprograma> Thanks for the help, guys.
Mar 26 15:05:30 <andre__> marcusb, bug 3032
Mar 26 15:05:33 <povbot> Bug https://bugs.maemo.org/show_bug.cgi?id=3032 missing diff utility (required by the GNU Coding Standards)
Mar 26 15:05:34 <marcusb> ok
Mar 26 15:05:36 <marcusb> thanks
Mar 26 15:05:48 <mgedmin> oh, cool
Mar 26 15:05:51 <mgedmin> avahi-daemon also needs that
Mar 26 15:07:05 <hrw> marcusb: thats nokism
Mar 26 15:07:08 <mgedmin> andre__, are you the one to ask about creating a bugzilla product for avahi in the Maemo Extras category?
Mar 26 15:07:50 <andre__> mgedmin, yes, by email: http://wiki.maemo.org/Bugs:Adding_Extra_products
Mar 26 15:07:55 <mgedmin> thanks!
Mar 26 15:08:01 <andre__> and bare with me as I just got back from vacation. long backlog :-/
Mar 26 15:08:06 <andre__> err bear
Mar 26 15:10:38 <gouverneur> can anyone help me with sending an sms via cli?
Mar 26 15:11:04 <gouverneur> is there a telepathy interface or something?
Mar 26 15:11:57 <gouverneur> looking in to opt it looks like noone ever follows a guidline (dont know if there is one)
Mar 26 15:12:38 <marcusb> geez, they can't fix this diff thing in 2 years?  *grmbl*
Mar 26 15:13:57 * mgedmin feels no hurry
Mar 26 15:14:07 <ShadowJK> battery meter is pretty confused by the mugen battery, even after a discharge-charge cycle..
Mar 26 15:15:25 <corecode_> how long does it last now?
Mar 26 15:21:06 <ShadowJK> wow, nexus 1 screen res is "fake"
Mar 26 15:21:10 <ShadowJK> olpc-like
Mar 26 15:21:47 <Corsac> ?
Mar 26 15:21:49 <andrewfblack> I just found my new work phone, where I work we only have access to iDEN network so I have always been stuck with really crappy phones.  I just saw the Motorola i1 and even though it wont be replacing my N900 as my primary phone its the best iden phone I have ever seen
Mar 26 15:22:18 <nidO> Corsac essentially the phone doesnt fully utilise the display's supported resolution
Mar 26 15:22:33 <Corsac> sounds like fail
Mar 26 15:24:27 <microlith> well no
Mar 26 15:24:56 <microlith> it does use the display's supported resolution, the display simply isn't RGB on every pixel
Mar 26 15:25:50 <ShadowJK> http://arstechnica.com/gadgets/news/2010/03/secrets-of-the-nexus-ones-screen-science-color-and-hacks.ars/
Mar 26 15:25:55 <MohammadAG> how long does an invitation into extras-devel take?
Mar 26 15:26:19 <ShadowJK> As far as I understand it, it's only 800x480 for green pixels :)
Mar 26 15:26:35 <microlith> pretty much
Mar 26 15:26:40 <noobmonk3y> MohammadAG:  Pretty quick during the week, but weekends are slow, did mine sat morning, got it accepted mon morning
Mar 26 15:27:02 <MohammadAG> noobmonk3y, oh, so maybe I should stop refreshing my email account :)
Mar 26 15:27:04 <noobmonk3y> done manually
Mar 26 15:27:08 <noobmonk3y> yup ;)
Mar 26 15:27:12 <MohammadAG> yeah I know lol
Mar 26 15:28:57 <MohammadAG> noobmonk3y, so I just upload the three files and the autobuilder does the rest?
Mar 26 15:29:52 <MohammadAG> http://talk.maemo.org/attachment.php?attachmentid=8443&stc=1&d=1269617198
Mar 26 15:32:33 <MohammadAG> rebooting into linux
Mar 26 15:36:13 <noobmonk3y> wb
Mar 26 15:36:36 <noobmonk3y> and yes MohammadAG - yes you upload one(Changes file) - it checks and then asks you to upload the other two,....
Mar 26 15:38:26 <Gizmokid2005> alright, I feel like an idiot, but I can't find this video, or figure out the name of the feature. In Firefox you can swipe right/left to bring up the menus on those sides, (tabs, etc), and I know there's a way to do it in Qt...but I can't think of the name or find the video that showcases this feature....Help???
Mar 26 15:38:35 <MohammadAG> ty noobmonk3y :)
Mar 26 15:47:01 <Arif_> moo
Mar 26 15:54:41 <Guest91702> silahkan masuk Guest91702 silahkan chit-chat dimari..
Mar 26 15:54:57 <Guest91702> silahkan masuk evo silahkan chit-chat dimari..
Mar 26 15:56:16 <Guest91702> silahkan masuk vpoluceno silahkan chit-chat dimari..
Mar 26 15:57:27 <wazd_e63> For those who cares:  we're near the monument :)
Mar 26 15:59:47 <ptl> which monument?
Mar 26 16:04:40 <Myrtti> ptl: cheesecake
Mar 26 16:07:53 <FIQ> hm, a question
Mar 26 16:08:03 <FIQ> the opt-thing-idea
Mar 26 16:08:06 <FIQ> why have that?
Mar 26 16:10:28 <rexjack> just one questin how do i restart the musik search in mediaplayer
Mar 26 16:10:36 <aquatix> Myrtti: the cake is a lie
Mar 26 16:15:28 <corecode_> what's the dimensions of the backdrop?
Mar 26 16:18:17 <corecode_> oh
Mar 26 16:18:26 <corecode_> i need to cut the image?
Mar 26 16:18:28 <corecode_> hm
Mar 26 16:20:23 <AntiXpucT> http://antixpuct.ru/maemo/opt/
Mar 26 16:20:31 <AntiXpucT> it is kvirc for N900
Mar 26 16:20:34 <ptl> wow
Mar 26 16:20:36 <ptl> really?
Mar 26 16:20:38 <ptl> lemme see
Mar 26 16:21:02 <AntiXpucT> but it is not fully usable
Mar 26 16:23:06 <MohammadAG> konttori, ping
Mar 26 16:24:52 <b0unc3> hello
Mar 26 16:25:05 <ptl> AntiXpucT: why not? what's the problem/
Mar 26 16:25:18 <gouverneur> rexjack: search maemo for trackerd
Mar 26 16:25:22 <aoeu> hello!
Mar 26 16:25:46 <Myrtti> FIQ: you mean optifying?
Mar 26 16:25:52 <aoeu> is there a package for the n900 that includes the 'screen' utility?
Mar 26 16:26:10 <FIQ> yes
Mar 26 16:26:31 <AntiXpucT> ptl, widgets size ;)
Mar 26 16:26:31 <FIQ> btw, why cannot we install .deb packages not coming from a repo?
Mar 26 16:26:53 <greenfly> you can (if they are compiled for ARM) but you have to be careful to not fill up the / partition
Mar 26 16:27:14 <greenfly> also there might be dependencies that you have to pull down manually
Mar 26 16:27:22 <noobmonk3y> screen utility?
Mar 26 16:27:24 <Flandry> Are there instructions somewhere for updating the SDK?
Mar 26 16:27:27 <greenfly> having packages in repositories solves this as the dependencies are handled
Mar 26 16:27:36 <FIQ> yes
Mar 26 16:27:53 <FIQ> so KVirc isn't optified?
Mar 26 16:27:58 <b0unc3> I just upgraded my SDK and now when I copy the bin on my device I get a dependencies problem with libhildon1 ,I think this issue is related with the upgrade, is there any way to fix ?
Mar 26 16:27:59 <greenfly> dunno
Mar 26 16:28:04 <nidO> it may or may not be
Mar 26 16:28:11 <aoeu> noobmonk3y: yes. http://linux.die.net/man/1/screen
Mar 26 16:28:13 <greenfly> I've been running irssi from a remote screen session for irc for years
Mar 26 16:28:33 <Myrtti> FIQ: the optify thing is because the / is quite small partition and you end up filling it unless you install software to other partitions. Filling the / is in general A Bad Idea™ as you prolly know
Mar 26 16:28:46 <FIQ> yes
Mar 26 16:28:48 <FIQ> but
Mar 26 16:28:56 <MohammadAG> why is there a New MMS string rtcom-messaging-ui.mo
Mar 26 16:28:59 <FIQ> then we come to the next thing:
Mar 26 16:29:03 <MohammadAG> +in*
Mar 26 16:29:16 <FIQ> why having a root partition even smaller than the swap?
Mar 26 16:29:33 <nidO> limitation of the size of nand memory chips
Mar 26 16:29:39 <achipa> FIQ: hardware limit
Mar 26 16:29:55 <FIQ> ok
Mar 26 16:30:19 <FIQ> but i would happily spend 10GB of MyDocs to the rootfs (if that would be possible)
Mar 26 16:30:29 <ptl> aoeu: screen is available in the tools repository
Mar 26 16:30:33 <MohammadAG> it should be
Mar 26 16:30:35 <nidO> technically it might be possible by repartitioning
Mar 26 16:30:37 <noobmonk3y> cheers aoeu  - interesting :D
Mar 26 16:30:37 <ptl> aoeu: I have it installed and it works perfectly
Mar 26 16:30:43 <nidO> but youll end up with a device that takes like 5 minutes to boot
Mar 26 16:30:50 <noobmonk3y> Arif_: You are quiet, this is not normal? ;)
Mar 26 16:30:55 <FIQ> meh
Mar 26 16:30:57 <FIQ> ok
Mar 26 16:31:08 <greenfly> there /has/ to be a rootfs / opt faq we can point people to
Mar 26 16:31:10 <nidO> nand = much much much faster than emmc
Mar 26 16:31:15 <greenfly> this question comes up every day
Mar 26 16:31:24 <nidO> its in about 50 places in the wiki
Mar 26 16:31:28 <nidO> and like 5,000 on the forums
Mar 26 16:31:31 <greenfly> that's the problem
Mar 26 16:31:33 <ptl> there is, greenfly
Mar 26 16:31:41 <Flandry> http://wiki.maemo.org/Opt_Problem
Mar 26 16:31:41 <greenfly> ptl: then let's use it :)
Mar 26 16:32:22 <Flandry> Great, now could someone point me to the wiki entry on updating SDK? :P
Mar 26 16:32:42 <nidO> its just apt-get update apt-get dist-upgrade
Mar 26 16:32:43 <nidO> as fakeroot
Mar 26 16:32:50 <Flandry> I really borked it when i tried apt-get before :/
Mar 26 16:32:54 <MohammadAG> Flandry, apt-get dist-upgrade
Mar 26 16:32:57 <MohammadAG> it doesn't need a wiki
Mar 26 16:33:00 <MohammadAG> nidO, beat me to it
Mar 26 16:33:16 <MohammadAG> Flandry, that borks the N900, not the SDK
Mar 26 16:33:18 <Flandry> do you have to disable any repos
Mar 26 16:33:33 <aoeu> ptl: much appreciated
Mar 26 16:33:36 <Flandry> I'm speaking past tense from experience lol
Mar 26 16:33:48 <Flandry> I'll try on the new image
Mar 26 16:33:54 <Flandry> Just wanted to verify
Mar 26 16:34:08 <MohammadAG> Flandry, I had all repos enabled
Mar 26 16:34:20 <Flandry> ok thanks
Mar 26 16:34:31 <Flandry> Read the changes thread, thanks for that
Mar 26 16:34:59 <MohammadAG> Flandry, np :)
Mar 26 16:35:29 <Flandry> oh
Mar 26 16:35:48 <aoeu> noobmonk3y: welcome to the dark side :)
Mar 26 16:35:49 <MohammadAG> I repeat, why is there a new MMS string?
Mar 26 16:35:49 <Flandry> does it need to be done in both targets separately?
Mar 26 16:36:25 <corecode_> so what do i use to download a file from http?
Mar 26 16:36:31 <corecode_> for scripting purposes
Mar 26 16:36:34 <corecode_> wget doesn't exist
Mar 26 16:36:36 <Flandry> wget
Mar 26 16:36:39 <Flandry> depend on it
Mar 26 16:38:13 <MohammadAG> Flandry, depends on what you want to do
Mar 26 16:38:32 <MohammadAG> if it's for programming, yes, if it's for emulation of the UI, X86
Mar 26 16:38:35 <Flandry> ok, so yes for release
Mar 26 16:39:28 <ptl> Flandry: I updated my maemo5 sdk ok, but I have some repositories on it that sometimes fail and it won't proceed if apt return an error.
Mar 26 16:39:41 <ptl> Flandry: so to run the maemo5 gui I had to disable these repositories
Mar 26 16:40:00 <ptl> Flandry: but the procedure is just that - you run the gui installer for the maemo5 sdk and all goes well
Mar 26 16:40:01 * noobmonk3y grins slightly evil'y
Mar 26 16:40:09 <nidO> so how was butlins
Mar 26 16:40:34 <noobmonk3y> MohammadAG: New MMS String - as in MMS - as in frals? - possibly something to do with nokia assisting Frals?
Mar 26 16:41:12 <marcusb> anybody has native packages for libgles2-sgx-img-dev and opengles-sgx-img-common-dev?  I want a native libqt4-dev
Mar 26 16:41:13 <b0unc3> guys, anyone get problems using .debs created after the SDK update ?
Mar 26 16:48:07 <dotblank> Is there a way to use gpsd?
Mar 26 16:53:36 <MohammadAG> noobmonk3y, wouldn't he have said something about it? And I really doubt it since it doesn't follow MMS specs
Mar 26 16:55:06 <nidO> who knows, nokia on the way to adding it themselves maybe
Mar 26 16:55:21 <ptl> is there an easy way to use gpsd on your ubuntu and connecting to the GPS on the N900?
Mar 26 16:55:45 <noobmonk3y> MohammadAG: If it is something to do with pr1.2 he would not be allowed to talk about it
Mar 26 16:56:09 <marcusb> is there a calendar app supporting caldavd?
Mar 26 16:56:11 <noobmonk3y> there must be a reason it is not there, and in my head if they are not officially creating it, it must be there to support it
Mar 26 16:57:11 <noobmonk3y> marcusb:  dont think so, but no calendar expert
Mar 26 16:57:17 <MohammadAG> noobmonk3y, live wallpapers were talked about :)
Mar 26 16:57:28 <noobmonk3y> http://talk.maemo.org/showthread.php?t=22831
Mar 26 16:57:45 <noobmonk3y> hmm true somethings have been mentioned, but i'm very certain other things have been kept hush hush :D
Mar 26 16:57:53 <petteri> ahh... does live wallpapers work with new sdk
Mar 26 16:58:21 <MohammadAG> petteri, I already pinged konttori, he doesn't seem to be ponging, so I'm assuming he's away
Mar 26 16:58:47 * noobmonk3y plays ping pong then goes home ;)
Mar 26 16:58:48 <noobmonk3y> ciao alls
Mar 26 16:59:09 <nidO> later
Mar 26 17:02:06 <dotblank> what repo is bluez-utils-test in?
Mar 26 17:13:32 <mgedmin> dotblank, try searching on http://maemo.org/packages
Mar 26 17:14:27 * MohammadAG wonders why qwerty12 never comes on here anymore :(
Mar 26 17:14:50 <Corsac> hmhm, still no news of PR1.2?
Mar 26 17:15:05 <dotblank> What exactly is the purpose of minigpsd?
Mar 26 17:15:26 <dotblank> does it actually use the gps device? or does it connect to an external one
Mar 26 17:18:03 <naxxatoe|mbp> dotblank: i wondered the same
Mar 26 17:18:55 <dotblank> Because im trying to use aircrack-ng with it and it has an option to use gpsd
Mar 26 17:20:50 <dotblank> maybe I can modify the source of airodump-ng to add support for liblocation API
Mar 26 17:28:29 <ptl> does GPSd allow remote forwarding of GPS data?
Mar 26 17:29:19 <dotblank> I think so but I can't seem to find a wrapper for gpsd in maemo
Mar 26 17:29:32 <dotblank> its been replaced by liblocation
Mar 26 17:30:33 <ptl> there's minigpsd for it
Mar 26 17:30:36 <ptl> would it work?
Mar 26 17:30:37 <ptl> hmm
Mar 26 17:30:48 <ptl> liblocation reserves the GPS and doesn't allow it to be used?
Mar 26 17:31:48 <gouverneur> b0unc3: have you thought of that devices are still !=PR1.2 before you asked that question?
Mar 26 17:39:10 <TheOne> hi @all
Mar 26 17:39:17 <aziwoqpd> hi @one
Mar 26 17:40:07 <TheOne> my new n900 will be deliver tomorrow
Mar 26 17:40:20 <TheOne> is there a tutorial with first steps?
Mar 26 17:40:54 <MohammadAG> Noobmonk3y, just got the invitation :)
Mar 26 17:41:03 <MohammadAG> TheOne, there's a Get started app installed
Mar 26 17:41:15 <Lumpio-> Put battery in, push button.
Mar 26 17:41:16 <TheOne> ah cool
Mar 26 17:41:16 <MohammadAG> and there are hundreds of videos on youtube
Mar 26 17:41:45 <b0unc3> gouverneur, yes, but so, if I release a deb (builded with pr1.2, since also autobuilder is updated to 1.2) nobody would be able to test it ?
Mar 26 17:41:46 * disco_stu me eclipso.. hasta lueguines
Mar 26 17:43:12 <Noobmonk3y> MohammadAG, yay
Mar 26 17:43:26 <TheOne> i am really happy - can't wait to get that device in my fingers ^^
Mar 26 17:43:34 <ptl> TheOne: there's a quickstart guide with it, but if you really want a manual, you can download the (excellent) manual in PDF format from the Nokia site
Mar 26 17:43:51 <TheOne> ah okay, thx
Mar 26 17:44:30 <TheOne> some of you guys involved with MeeGo?
Mar 26 17:45:36 <Khertan> hi !
Mar 26 17:45:57 <ptl> hello again
Mar 26 17:46:09 <ptl> TheOne: join #meego
Mar 26 17:51:10 <Khertan> does there is a dvorak keyboard layout available ?
Mar 26 17:52:28 <hrw> hi
Mar 26 17:52:35 <hrw> Khertan: ukeyboard had one iirc
Mar 26 17:52:59 <hrw> b0unc3: exactly
Mar 26 17:53:14 <Khertan> hi hrw and thx ... i ll take a look at it
Mar 26 17:56:48 <TheOne> can someone tell me how much internet traffic volume he/she has with the n900 per month?
Mar 26 17:56:58 <TheOne> circa
Mar 26 17:57:10 <TheOne> more than 1GB ?
Mar 26 17:58:52 <MohammadAG> Noobmonk3y, does autobuilder build applications for both i368 and armel?
Mar 26 17:59:13 <Noobmonk3y> MohammadAG,  yup think so
Mar 26 17:59:25 <lcuk> yes
Mar 26 17:59:31 <Noobmonk3y> armel always takes longer
Mar 26 18:00:05 * lcuk finds a quiet little patch in the corner and curls up to go to sleep
Mar 26 18:00:13 <Noobmonk3y> lol
Mar 26 18:00:16 <Noobmonk3y> evening
Mar 26 18:00:39 <Noobmonk3y> lcuk, me thinks u will like the next healthcheck :)
Mar 26 18:01:03 <Khertan> hrw: ukeyboard is only a virtual keyboard not for mapping the n900 keyboard, isn t it ?
Mar 26 18:01:11 <lcuk> does it check blood pressure and administer amphetamines?
Mar 26 18:01:20 <hrw> Khertan: in n810 times it was both
Mar 26 18:01:33 <lcuk> oh, and prescribe those little matchsticks for eyes
Mar 26 18:01:53 <Khertan> bacon flying to verify lcd dead pixels ?
Mar 26 18:02:02 <dmj726_n9001> TheOne: it varies hugely from user to user.
Mar 26 18:02:08 <lcuk> mmmmmmmmmmmmmm bacon
Mar 26 18:02:16 <Noobmonk3y> lol
Mar 26 18:02:17 <lcuk> thanks Khertan for reminding me of the finer things in life
Mar 26 18:02:26 <dmj726_n9001> some use a couple hundred mb some use several gb
Mar 26 18:02:29 <Khertan> hrw: maybe that description isn't clear enought
Mar 26 18:02:51 <Khertan> lcu: héhé
Mar 26 18:02:56 <Khertan> lcuk: héhé
Mar 26 18:03:20 <TheOne> dmj726_n9001: i know, but i though someone can tell me some numbers
Mar 26 18:03:43 <Khertan> arg cannot be install package missing :(
Mar 26 18:04:35 <dmj726_n9001> I think there is a thread in the forums.
Mar 26 18:05:27 <MohammadAG> Noobmonk3y, cause it built the i368 and it's getting pushed to extras-devel, but I haven't received anything about the armel version
Mar 26 18:05:47 <Noobmonk3y> takes an hour at best recently a few more
Mar 26 18:09:14 <summel> can i install maemo on a touchbook?
Mar 26 18:09:52 <cehteh> mer at best
Mar 26 18:09:57 <Noobmonk3y> well u can get it on an ipad so why not
Mar 26 18:10:36 <summel> ewww ipad
Mar 26 18:10:49 <Noobmonk3y> lol
Mar 26 18:12:17 <nidO> summel it's already known to run mer
Mar 26 18:12:30 <summel> how does mer compare to the maemo on trhe n900?
Mar 26 18:12:37 <nidO> maemo support is apparently on the way
Mar 26 18:12:50 <nidO> and knowing AI, theyll probably find a way to manage it pretty well
Mar 26 18:12:57 <summel> i would use meego if it would be better then moblin
Mar 26 18:13:36 <nidO> I dont really know why you'd want either mer or maemo on a touchbook other than to say "oh look it works"
Mar 26 18:13:36 <cehteh> well you cant install nokia binaries on another device, the eula forbids that
Mar 26 18:14:04 <summel> nidO: because i love the maemo on the n900 :)
Mar 26 18:14:10 <summel> and i dont want the ai "os"
Mar 26 18:14:32 <nidO> thats because its on an n900. on a full touchbook, the main choices, ai os, ubuntu, and love it or hate it, android, are all much better choices
Mar 26 18:15:25 <summel> i would rather use gentoo with kde4 then anything else currently availeable :P
Mar 26 18:15:40 <nidO> gentoo works on it as well
Mar 26 18:15:57 <summel> but compiling takes sooo long :(
Mar 26 18:16:03 <summel> esp on 600 MhZ >_>
Mar 26 18:16:36 <lcuk> how on earth did people cope in days before 600mhz processors?
Mar 26 18:16:50 <lcuk> i remember building apps and stuff on 7mhz amiga in c
Mar 26 18:17:29 <hrw> lcuk: I used AmigaE
Mar 26 18:17:45 <luke-jr> summel: meh, works for me on N810
Mar 26 18:18:09 <luke-jr> lcuk: code was less abstracted
Mar 26 18:18:19 <luke-jr> and compiled faster
Mar 26 18:18:22 <lcuk> abstraction should make code faster
Mar 26 18:18:30 <lcuk> its not like functions are a new thing
Mar 26 18:18:36 <luke-jr> abstraction just makes everything slower
Mar 26 18:18:44 <luke-jr> I don't mean functions
Mar 26 18:18:55 <luke-jr> I mean interfaces independent from implementation
Mar 26 18:21:06 <lcuk> luke-jr, ?
Mar 26 18:21:12 <lcuk> you mean like pascal :D
Mar 26 18:21:20 <luke-jr> ...
Mar 26 18:21:26 <luke-jr> I mean APIs ;)
Mar 26 18:21:42 <luke-jr> forward/backward compat
Mar 26 18:22:03 <luke-jr> also, I think compilers are a bit faster if you use -O0
Mar 26 18:26:23 <tg> hoi
Mar 26 18:27:14 <tg> anyone knows what's the current status of qt perl bindings?
Mar 26 18:27:24 <tg> i've just seen it listed as a gsoc project idea
Mar 26 18:28:44 <luke-jr> probably non-existent
Mar 26 18:28:53 <MiXu-> I don't see why anyone would want to write qt apps in perl =)
Mar 26 18:29:26 <MiXu-> When there's Python
Mar 26 18:30:32 * SpeedEvil wonders on qt-awk
Mar 26 18:31:19 <t-tan> if you want perl code, just run the python code through an obfuscator :)
Mar 26 18:34:45 <luke-jr> MiXu-: because Python sucks?
Mar 26 18:35:14 <MiXu-> And Perl doesn't? :)
Mar 26 18:35:22 <MiXu-> Í've done my share of perl
Mar 26 18:35:28 <luke-jr> Perl is awesome
Mar 26 18:35:38 <luke-jr> but not meant for GUIs, I admit
Mar 26 18:35:44 <MiXu-> Perl is awesome to a point
Mar 26 18:36:31 <MiXu-> But when you get to a point where you're dealing with thousands of lines of code it stops being cool.
Mar 26 18:36:51 <MiXu-> For small stuff it's very useful
Mar 26 18:37:35 <Kurppa_> Hi everybody. Maemo newbie here. :-)
Mar 26 18:37:46 <MiXu-> Hi Kurppa. Long time no see.
Mar 26 18:38:33 <Kurppa_> Oh hi. I remember you :-)
Mar 26 18:38:44 <MiXu-> :D
Mar 26 18:38:48 * luke-jr finds Perl especially useful for writing compilers
Mar 26 18:39:04 <Arif_> ~rob Noobmonk3y
Mar 26 18:39:05 * infobot strips Noobmonk3y of his nick, title and all rights & privileges and gives them to someone more worthy
Mar 26 18:39:12 * Noobmonk3y farts
Mar 26 18:39:20 <Arif_> I waz outside :O
Mar 26 18:39:27 * lcuk bottles it and releases it back up Noobmonk3y 
Mar 26 18:40:24 <Venomrush> just updated http://wiki.maemo.org/Maemo_5/PR1.2
Mar 26 18:40:28 <red> can't believe i didnt need this earlier
Mar 26 18:40:42 <red> but how on earth do you create a custom group in contacts and apply a different ringtone for those?
Mar 26 18:40:44 <MohammadAG> Noobmonk3y, one more question, when I have an updated package I just upload it again and it puts it in the same directory right?
Mar 26 18:40:56 <Arif_> Venomrush, so pr1.2 is out?
Mar 26 18:41:06 <Venomrush> more BUGS RESOLVED added
Mar 26 18:41:14 <Arif_> oh
Mar 26 18:41:16 <Venomrush> no :\
Mar 26 18:41:17 <Arif_> nm then :(
Mar 26 18:41:19 <Venomrush> i didn't say that
Mar 26 18:41:21 <hrw> Venomrush: I do not like your way of fixing bugs
Mar 26 18:41:36 <Venomrush> i don't fix bugs..
Mar 26 18:41:40 <hrw> bug 6933 for example
Mar 26 18:41:40 <povbot> Bug https://bugs.maemo.org/show_bug.cgi?id=6933 Alarm times are shown wrong
Mar 26 18:42:22 <Arif_> # 7124 media player loses cover art if song changes while screen is locked
Mar 26 18:42:22 <Arif_> # eo
Mar 26 18:42:22 <Arif_> # 7353 RDS station name should not contain album (esp. when its unknown)
Mar 26 18:42:24 <red> how important is the full battery before flashing btw?
Mar 26 18:42:24 <Arif_> huh?
Mar 26 18:42:26 <Venomrush> i wasn't able to reproduce it on my end
Mar 26 18:42:32 <Venomrush> with my N97
Mar 26 18:42:33 <Venomrush> so
Mar 26 18:42:44 <Arif_> whats the eo in there?
Mar 26 18:43:18 <Venomrush> it got truncated
Mar 26 18:43:39 <Arif_> what was it originally I wonder..
Mar 26 18:43:41 <hrw> Venomrush: now try to reproduce it again with test event
Mar 26 18:43:50 <Venomrush> then feel free to reopen :)
Mar 26 18:43:58 <hrw> Venomrush: this time I gave nearly 5 years for nokia to fix it
Mar 26 18:44:03 <hrw> Venomrush: it is reopened and confirmed
Mar 26 18:44:09 <Venomrush> cool
Mar 26 18:44:28 <GAN900> Ah, nice post from achipa.
Mar 26 18:44:28 <Venomrush> think i got most of the stuff from http://repository.maemo.org/stable/fremantle/Changelog_3.2010.02-8_vs_10.2010.09-3.txt into the PR1.2 wiki
Mar 26 18:45:05 <red> wtf
Mar 26 18:45:08 <red> no contact groups
Mar 26 18:45:15 <red> and this was labeled as a business device :)
Mar 26 18:45:39 <hrw> Venomrush: drop list of pacakges maybe from wiki
Mar 26 18:45:52 <hrw> Venomrush: or move them to end of page
Mar 26 18:46:04 <hrw> Venomrush: most users are interested in new features and fixed bugs rather
Mar 26 18:46:17 <Venomrush> from the version number
Mar 26 18:46:27 <Venomrush> it seems pr1.2 only contain fixes up until week 9
Mar 26 18:46:34 <Venomrush> so that's first week of march
Mar 26 18:47:49 <hrw> I will add new feature which rather no one found before :D
Mar 26 18:48:59 <Venomrush> i thought pr1.2 will have live wallpaper
Mar 26 18:49:07 <Venomrush> apparently not anymore from the sdk
Mar 26 18:49:46 <hrw> Venomrush: try to make better descriptions in wiki page - check my last changes
Mar 26 18:50:15 <SpeedEvil> what page is this?
Mar 26 18:50:29 <MohammadAG> SpeedEvil, http://wiki.maemo.org/Maemo_5/PR1.2
Mar 26 18:50:42 <MohammadAG> what else gets talked about other than PR1.2 :)
Mar 26 18:50:49 <Arif_> Moogo
Mar 26 18:51:12 <hrw> Arif_: PeeGo?
Mar 26 18:52:17 <Venomrush> "* Status area is now reachable in portrait mode"
Mar 26 18:52:21 <hrw> do we want bug numbers to be links?
Mar 26 18:52:42 <Venomrush> there is no 'portrait mode' you may want to be specific there
Mar 26 18:52:44 <MohammadAG> Venomrush, I noticed that, but I thought it was a bug/forced rotation, who confirmed it?
Mar 26 18:52:44 <Arif_> status area?
Mar 26 18:53:02 <MohammadAG> Arif_, the thing with the clock/availability etc..
Mar 26 18:53:11 <hrw> Venomrush: run phone app, rotate phone
Mar 26 18:53:17 <Arif_> yeah, that's so useful to have in portrait mode....
Mar 26 18:53:20 <hrw> Venomrush: you will have portrait mode right?
Mar 26 18:53:42 <MohammadAG> Arif_, actually, it is
Mar 26 18:53:42 <hrw> and in PR1.2 you will be able to use status area
Mar 26 18:53:47 <Venomrush> "* Status area is now reachable in portrait mode" - this makes it sound phone support full portrait mode
Mar 26 18:53:57 <Venomrush> *sounds like
Mar 26 18:54:08 <Venomrush> change it to phont/browser specific
Mar 26 18:54:11 <Venomrush> *phone
Mar 26 18:54:26 <MohammadAG> Venomrush, many apps have portrait support, not just phone/browser
Mar 26 18:55:14 <Venomrush> the 'official' :p
Mar 26 18:56:16 <zaheerm> blocks :)
Mar 26 18:56:23 * Arif_ wonders if other official apps will ever get portrait
Mar 26 18:56:25 <hrw> Venomrush: any app can be in portrait mode if is programmed properly
Mar 26 18:56:31 <hrw> Arif_: doubt so
Mar 26 18:56:50 <Arif_> mhh
Mar 26 18:56:57 <nidO> srsly why would they bother
Mar 26 18:56:58 <Arif_> I'll buy a Moogo device then :D
Mar 26 18:57:06 <Arif_> one with real speakers!
Mar 26 18:57:19 <zaheerm> i doubt they are going to spend the time doing the whole lifecycle UI design -> code to get portrait support for apps on maemo 5
Mar 26 18:57:39 <SpeedEvil> !bug 7160
Mar 26 18:57:40 <povbot> Bug https://bugs.maemo.org/show_bug.cgi?id=7160 Camera app hardcodes and requires FAT partition
Mar 26 18:57:43 <zaheerm> harmattan will no doubt be where they put effort into getting good portrait uis for apps
Mar 26 18:57:53 <MohammadAG> <Arif_> one with real speakers!
Mar 26 18:58:06 <MohammadAG> I think the EU doesn't allow 5800-like speakers
Mar 26 18:58:19 <MohammadAG> the same EU that forced us to use microUSBs for charging
Mar 26 18:58:20 <Arif_> they don't have a regulation about speakers
Mar 26 18:58:27 <Arif_> only headphones
Mar 26 18:59:09 <TheOne> is there a key list for the channel bot?
Mar 26 18:59:14 <MohammadAG> when a package is built, does it automatically get bumped to the latest version?
Mar 26 19:00:14 <TheOne> for the infobot
Mar 26 19:00:47 <hrw> ok, updated page a bit and reformatted
Mar 26 19:01:02 <hrw> bye
Mar 26 19:01:08 <hrw|gone> have a nice rest of day and weekend
Mar 26 19:02:20 * lcuk hides behind Noobmonk3y 
Mar 26 19:03:09 * Arif_ yawns at PR1.2
Mar 26 19:04:24 * MohammadAG hits Arif_ 
Mar 26 19:04:37 * Arif_ throws search in media palyer at MohammadAG
Mar 26 19:04:49 <DocScrutinizer51> errr ""filemanager is now able to share *ANY* file...""???
Mar 26 19:04:52 * nidO hates stupid users
Mar 26 19:05:03 <nidO> any file that it can access yes
Mar 26 19:05:15 <DocScrutinizer51> lol
Mar 26 19:06:01 <Venomrush> still can't see the extensions
Mar 26 19:06:04 <Venomrush> annoying
Mar 26 19:06:16 <DocScrutinizer51> so this wording implies a major flaw been fixed while in fact shit remains the same
Mar 26 19:06:30 <nidO> ?
Mar 26 19:06:34 <Venomrush> it's not a proper file manager to me without this
Mar 26 19:06:42 <Venomrush> and changing extensions..
Mar 26 19:07:00 <Venomrush> could be a settings like windows explorer to show extensions as an advanced option
Mar 26 19:07:21 <nidO> have you tried gpe to see if that shows extensions?
Mar 26 19:07:30 <DocScrutinizer51> WHF has none if the Nokia devel took a look at his TBP, to get inspiration from finder's "go to" function?
Mar 26 19:07:44 <nidO> in fact, gpe does show extensions
Mar 26 19:07:52 <lcuk> OMG WTF BBQ?
Mar 26 19:08:28 <MohammadAG> Venomrush, well they tried to make it easier and less risky for users to toy with their systems
Mar 26 19:08:38 <MohammadAG> suppose the filemanager showed /
Mar 26 19:09:13 <MohammadAG> "I found a directory called /usr that contained a ****load files so I deleted it and my N900 doesn't boot up. HELP!!!!!!111!!!!!!!!!!!!! \
Mar 26 19:09:24 <MohammadAG> I'm disappointed in Nokia with the crap N900"
Mar 26 19:10:26 <DocScrutinizer51> MohammadAG: you know how cute finder handles this?
Mar 26 19:10:48 <joppu> Delete your system32 to make your computer run faster
Mar 26 19:11:02 <DocScrutinizer51> MohammadAG: btw user should lack perms to delete system stuff
Mar 26 19:11:18 <MohammadAG> DocScrutinizer51, knowing nokia, the user would have perms on PR1.0 :p
Mar 26 19:11:28 * noobmonk3y waves
Mar 26 19:11:35 <MohammadAG> joppu, just delete the whole windows directory
Mar 26 19:11:44 <noobmonk3y> anyone fancy asking what should be a simple python question for me?
Mar 26 19:11:58 <nidO> asking it or answering it?
Mar 26 19:12:05 <noobmonk3y> answering lol
Mar 26 19:12:05 <MohammadAG> seems like asking
Mar 26 19:12:13 <noobmonk3y> self.textBrowser.setText ("text")    ....  sets the text in a textbrowser....
Mar 26 19:12:29 <noobmonk3y> if i want to add loads of text.... how do i append, so not to create one huuuge line
Mar 26 19:12:32 <C-S-B-N900> woohoo. maemo just had fit with gpodder.
Mar 26 19:12:40 <noobmonk3y> lol C-S-B-N900
Mar 26 19:12:45 * MohammadAG has no idea what that means
Mar 26 19:12:56 <MohammadAG> noobmonk3y, could be \?
Mar 26 19:13:08 <noobmonk3y> erm, i get the new line part
Mar 26 19:13:15 <MohammadAG> like in terminal
Mar 26 19:13:15 <andrewfblack_> would it be wrong to charge people to build their themes for them if they are to lazy to learn to do it them selfs?
Mar 26 19:13:19 <noobmonk3y> but thinking more of splitting the line in the coed?
Mar 26 19:13:27 <noobmonk3y> code*
Mar 26 19:13:31 <lizardo> noobmonk3y: "your " + "long" + "text"     , or put it on some file and use open("file").read()
Mar 26 19:13:35 * Arif_ waits for "share" in media player and falls asleep
Mar 26 19:13:36 <MohammadAG> andrewfblack_, imo, no, since they can build it alone
Mar 26 19:13:37 <lcuk> noobmonk3y, coed?
Mar 26 19:13:42 <noobmonk3y> ooo lizardo - good idea
Mar 26 19:13:44 <noobmonk3y> code*
Mar 26 19:13:48 <noobmonk3y> lcuk - I r Ill
Mar 26 19:13:54 <MohammadAG> andrewfblack_, basically, xchat do it, Windows version is paid, linux one isn't (and it's open source)
Mar 26 19:13:55 <noobmonk3y> and can't type but thats usual
Mar 26 19:13:57 <lcuk> i am tired
Mar 26 19:13:59 <MohammadAG> does*
Mar 26 19:14:06 <lcuk> heh
Mar 26 19:14:09 <lcuk> Milhouse, :D
Mar 26 19:14:09 <noobmonk3y> lcuk - thats cos u dont sleep
Mar 26 19:14:14 <lcuk> i cant type either
Mar 26 19:14:14 <andrewfblack_> MohammadAG: thas true
Mar 26 19:14:24 <lcuk> noobmonk3y, i sleep occasionally
Mar 26 19:14:35 <MohammadAG> andrewfblack_, do I get half the income for saying yes? :P
Mar 26 19:15:05 <andrewfblack_> MohammadAG: lol I don't know if I'll do it but I should lol
Mar 26 19:15:27 * lcuk considers investing in cloning companies
Mar 26 19:15:27 <andrewfblack_> AndrewFBlack's Theme Build Service lol
Mar 26 19:15:32 <lcuk> but i need money first
Mar 26 19:16:33 <Arif_> what company would you clone
Mar 26 19:16:47 <lcuk> i would clone myself
Mar 26 19:16:54 <lcuk> so one of me can sleep
Mar 26 19:16:56 * noobmonk3y eeeeeeks
Mar 26 19:16:58 <lcuk> whilst the other does job
Mar 26 19:17:03 <lcuk> and another can work on liqbase
Mar 26 19:17:05 <MohammadAG> andrewfblack_, if you've got the time why not
Mar 26 19:17:06 <noobmonk3y> takeway - lol ordered it an hour ago, better go get it!
Mar 26 19:17:23 <MohammadAG> andrewfblack_, but I can imagine someone doing it for free just to make your service look bad :p
Mar 26 19:17:53 <andrewfblack_> MohammadAG: the effect will still be the same
Mar 26 19:18:08 <FIQ> uhm
Mar 26 19:18:38 * andrewfblack_ has been temped to do some paid themes but is to nice
Mar 26 19:18:57 <FIQ> ok, rhe sms/calling features isn't the most interesting thing in the phone, but it really should be a "Respond" button somewhere at a SMS
Mar 26 19:19:59 <TheOne> bye
Mar 26 19:20:25 <DocScrutinizer51> damn, 4h of UMTS seriously drained my bat to some 40%
Mar 26 19:20:37 <lcuk> AndrewFBlack, done cleanly with a decent set of correctly sourced materials i dont think anyone would complain at you :)
Mar 26 19:20:55 <lcuk> your themes kick ass
Mar 26 19:21:02 <MohammadAG> anyway to uninstall something that was installed from a setup.py file?
Mar 26 19:21:33 <MohammadAG> FIQ, huh?
Mar 26 19:21:47 <MohammadAG> the SMS feature is the best one I've seen on any device
Mar 26 19:21:47 <andrewfblack_> lcuk: You talking about my idea to charge people to build their themes or the idea about selling some themes?
Mar 26 19:23:07 <lcuk> mm both
Mar 26 19:23:24 <lcuk> soryr i got distracted looking at laughingstok's photo album
Mar 26 19:23:58 <FIQ> the sms
Mar 26 19:24:07 <FIQ> watching an incoming sms
Mar 26 19:24:14 <FIQ> missing "respond" button
Mar 26 19:24:29 <FIQ> only delete, copy, forward
Mar 26 19:27:55 <C-S-B-N900> any reason lirc doesnt start automatically?
Mar 26 19:28:21 <MohammadAG> what's /dev/pts?
Mar 26 19:29:04 <MohammadAG> FIQ, err, why do you click on the SMS in the conversation?
Mar 26 19:29:21 <MohammadAG> just type in the white box and click the send (enter) button
Mar 26 19:29:25 <FIQ> i didn't?
Mar 26 19:29:28 <FIQ> uhm
Mar 26 19:29:35 <FIQ> ok, maaaybe i did. :P
Mar 26 19:31:01 <lcuk> are there any v4l2 examples which people have tested to work with n900
Mar 26 19:35:23 <dottedmag> Is there ssh-agent and askpass dialog for Maemo?
Mar 26 19:37:09 <MohammadAG> any tmo mods online?
Mar 26 19:41:36 <cehteh> tmo mods?
Mar 26 19:47:02 <MohammdAG_> noobmonk3y, err, bugtrackers aren't auto created?
Mar 26 19:47:35 <ptl> <question type="newbie" flavor="troll">Why doesn't maemo support MMS? Why doesn't it have videocall? How about voice guide GPS? When is PR1.2 out? Why doesn't it have a capacitive screen? Why no multitouch? Couldn't it be thinner? How can I assign a ringtone to a contact? How do I do Skype video? How do I run java programs?</question>
Mar 26 19:47:40 <ptl> Did I leave something out?
Mar 26 19:48:20 <|R> troll are very creative, so something is always left out :P
Mar 26 19:48:23 <Kurppa_> Why does the battery last so short time?
Mar 26 19:48:29 <ptl> Maybe we could do a poll, what's your preferred newbie question?
Mar 26 19:48:34 <ptl> yes, good
Mar 26 19:48:38 <MohammdAG_> ptl, can we upload porn made on the N900 on tmo?
Mar 26 19:48:47 <Kurppa_> Why is the UI so slow?
Mar 26 19:48:54 <Kurppa_> Why does it leak so much memory?
Mar 26 19:49:04 <Kurppa_> Why is it so fat?
Mar 26 19:49:05 <MohammdAG_> why is the N900 called the N900? Why not N850?
Mar 26 19:49:11 <ptl> leak memory? That does not seem like a newbie question
Mar 26 19:49:21 <ptl> lol
Mar 26 19:49:25 <Kurppa_> Ah, sorry.
Mar 26 19:49:47 <odin_> Why does my (Symbian / iPhone / Andriod) application not work on it ?
Mar 26 19:49:51 <ptl> that's ok, we just change the type to "averagely informed user"
Mar 26 19:50:28 <Funnyfacemob> probably because they found the GSM/UMTS capabilities significant enough to name it that way
Mar 26 19:50:48 <odin_> Why does it break when I drop it on the floor ?
Mar 26 19:50:53 <Funnyfacemob> but well model numbers never make any sense anyway
Mar 26 19:51:04 <Funnyfacemob> it's a sales gimmick
Mar 26 19:51:17 <Kurppa_> Why can't I see the display on sunny weather?
Mar 26 19:51:17 <andrewfblack_> MohammdAG_: No on the Porn
Mar 26 19:51:17 <andrewfblack_> < TMO Mod
Mar 26 19:51:36 <ptl> Ok
Mar 26 19:51:41 <Kurppa_> Why is youtube so slow :(
Mar 26 19:51:46 <ptl> Is there any way for me to post a poll on talk.maemo.org?
Mar 26 19:51:47 <MohammdAG_> andrewfblack_, just making fun of the thread on TMO ;)
Mar 26 19:51:48 <Funnyfacemob> my nokia 5200 and 6300 are exactly the same devices
Mar 26 19:51:51 <ptl> so that I can post this
Mar 26 19:51:59 <ptl> I'm looking some way to do it.
Mar 26 19:52:10 <MohammdAG_> andrewfblack_, http://talk.maemo.org/showthread.php?t=48336
Mar 26 19:52:11 <odin_> Why isn't there going to be a Meego version for my N900 ?
Mar 26 19:52:20 <Stskeeps> eh
Mar 26 19:52:21 <odin_> :P
Mar 26 19:52:21 <Stskeeps> :P
Mar 26 19:52:21 <Kurppa_> Haha
Mar 26 19:52:52 <andrewfblack_> MohammdAG_: I was hoping lol did you notice I told him to post a photo of his wife then we would decided if he could post video or not
Mar 26 19:53:19 <MohammdAG_> haha yeah now I remember, thought you didn't see it ;)
Mar 26 19:54:16 <odin_> Where can I get this Angry Bird expansion pack ?   ... When can I get ..... ?
Mar 26 19:54:22 <sivang> hi all
Mar 26 19:54:32 <sivang> how do I sync all my data from N900 to N97 Mini ?
Mar 26 19:54:37 <MohammdAG_> andrewfblack_, can you please change the title of http://talk.maemo.org/showthread.php?t=44293 [ANNOUNCE]Noor (Quran app) now in extras-devel?
Mar 26 19:54:39 <sivang> how does that syncML thingo works ?
Mar 26 19:55:50 <andrewfblack_> MohammdAG_: I only Mod Themes and Design Sections talk to edgar2, Flandry about application section
Mar 26 19:56:11 <MohammdAG_> I'm assuming you just pinged them? :)
Mar 26 19:56:11 <andrewfblack_> MohammdAG_: don't ask why we don't have global power only Tim does
Mar 26 19:56:37 <MohammdAG_> andrewfblack_, I'm not haha, I mod on a PS3 site and we have the same 'limits'
Mar 26 19:56:59 <odin_> what was the outcome of the meego meeting the other day, unfortunately I had a late night at the data centre, maybe I should track down the recording
Mar 26 19:57:35 <odin_> I've been totally out of the loop on anything for the past 2 months, right about the time the announcement came
Mar 26 19:58:09 <Arif_> we're still waiting for an useful firmware update :P
Mar 26 19:58:56 <Kurppa_> It will probably be a long wait :-)
Mar 26 19:59:28 <andrewfblack_> what up GAN900
Mar 26 19:59:29 <odin_> was that comment to me Arif_ ?  I am on 3.2010.02-8  is this PR1.1.1 ?
Mar 26 19:59:47 <Arif_> yeah =}
Mar 26 19:59:52 <odin_> there have not been many software updates in the past 2 months either that I remember....
Mar 26 20:00:05 <odin_> is there a PR1.1.2 due out ?  with Qt 4.7 is it ?
Mar 26 20:00:09 * Arif_ doesn't know anything about Moogo so won't comment on that =P
Mar 26 20:00:11 <GAN900> andrewfblack_, installing door sensors and bemoaning the fact that home improvement stores never have what you need.
Mar 26 20:00:12 <Kurppa_> 4.6
Mar 26 20:00:16 <MohammdAG_> wasn't it 4..5
Mar 26 20:00:17 <MohammdAG_> 4.6*
Mar 26 20:00:19 <MohammdAG_> nvm
Mar 26 20:00:26 <Arif_> 1.2 should have 4.6.2
Mar 26 20:00:37 <MohammdAG_> 1.2 has* :)
Mar 26 20:00:45 <Arif_> there is no 1.2 :P
Mar 26 20:00:52 <MohammdAG_> there is an SDK
Mar 26 20:01:01 <odin_> ok its 4.5.3~git20090723-0maemo6+0m5 that is currently on my N900
Mar 26 20:01:32 <andrewfblack_> GAN900: of cource not why would they have it
Mar 26 20:01:44 <odin_> is the new Qt incompatible to be mixed at runtime ?  i.e. why can't 4.5.3 be put in /opt/maemo/lib for legacy apps ?
Mar 26 20:01:57 <GAN900> andrewfblack_, dunno, silly me.
Mar 26 20:03:20 <C-S-B> gmail down for anyone else?
Mar 26 20:03:41 <C-S-B> says its just me at http://downforeveryoneorjustme.com/
Mar 26 20:03:45 <andrewfblack_> My new door locks my wife called thinks I'm a nerd but I know I am http://link.schlage.com/Pages/home.aspx
Mar 26 20:03:50 <Kurppa_> Qt 4.6.2~git20100212-0maemo1 says this thing.
Mar 26 20:04:01 <odin_> ignoring the problem is memory footprint, I understand the binary linkage incompatibility (although usually thats a major version bump of the DSO version usually /usr/lib/libQtCore.so.4.5.3
Mar 26 20:04:02 <Kurppa_> New-ish.
Mar 26 20:04:48 <odin_> Kurppa_, what "thing" have you got there then ?
Mar 26 20:05:07 <Kurppa_> Never mind that.
Mar 26 20:05:56 <Kurppa_> Just ignore my comments.
Mar 26 20:05:57 <odin_> I guess you must know for sure its going to be a long wait for PR1.1.2
Mar 26 20:06:18 <odin_> if you have 4.6.2 on your "thing" already LOLs
Mar 26 20:07:01 <adisbladis> Is there any way to get your n900 internal storage be something other then vfat?
Mar 26 20:07:03 <odin_> so mixing Qt 4.5 and 4.6 at runtime is not possible because ?
Mar 26 20:07:18 <adisbladis> I want some kind of ext file system but my mkfs.ext2 is failing
Mar 26 20:07:49 <odin_> in theory sure, if you don't need Windows/Mac storage support via USB but most people like to keep that feature
Mar 26 20:08:14 <odin_> you are messing with /dev/mmcblk0p1  (as your peril) ?
Mar 26 20:08:27 <Kurppa_> If it's running in mass storage mode, I wonder if it's possible to just change eMMC to NTFS.
Mar 26 20:08:50 <Kurppa_> Then use the custom ntfs-3g kernel module to make N900 also understand the contents.
Mar 26 20:09:10 <odin_> then that would really mess up with the N900 own view of it, isn't NTFS still "not ready for production status" in kernel ?
Mar 26 20:09:22 <ptl> my eMMC is on ext3
Mar 26 20:09:28 <Kurppa_> I understand there is already some custom kernel package in maemo-devel which provides this ntfs-3g.
Mar 26 20:09:33 <ptl> I had to modify a couple of init scripts
Mar 26 20:09:36 <ptl> for it to work
Mar 26 20:09:43 <Kurppa_> But it is for PR1.1 kernel.
Mar 26 20:10:14 <Kurppa_> Can you export it using mass storage mode, so linux can see it?
Mar 26 20:10:39 <odin_> have you backup up content of  /home/user/MyDocs to /home/foobar, then unmounted it first adisbladis ?  umount /home/user/MyDocs
Mar 26 20:11:56 <adisbladis> odin_: Im trying to do it from my pc
Mar 26 20:12:01 <adisbladis> odin_: Not the phone
Mar 26 20:12:04 <ptl> Kurppa_: Yes, I can, and it automounts and works perfectly under linux
Mar 26 20:12:11 <Kurppa_> Very nice.
Mar 26 20:12:22 <adisbladis> It just destroyed my old vfat partition
Mar 26 20:13:06 <odin_> but isn't it mounted inside the device already as well as ?
Mar 26 20:13:39 <adisbladis> odin_: Nope
Mar 26 20:13:58 <adisbladis> It might be the phone trying to automount it as soon as it detects it and therefore freezing?
Mar 26 20:14:21 <ptl> odin_: when it's in mass storage mode, both MMCs (internal and external, MyDocs and my /media/mmc1) are unmounted in the phone and mounted in my Linux
Mar 26 20:14:23 <odin_> yes Im not sure of the details of how the /home/user/MyDocs is mounted when your engage mass storage mode
Mar 26 20:14:32 <ptl> odin_: and when I unplug it, it correctly remounts them in the phone
Mar 26 20:14:53 <MohammadAG> **** I turned off the N900 while installing a package
Mar 26 20:14:53 <odin_> ideally there should be an interface that provides a VFAT view of any FS for mass-storage mode
Mar 26 20:14:54 <ptl> but as I said, I modified two scripts... mainly /usr/bin/mmc-mount
Mar 26 20:15:09 <MohammadAG> hopefully dpkg --configure -a will fix it
Mar 26 20:15:21 <ptl> MohammadAG: apt-get -f install, then reinstall the package
Mar 26 20:15:30 <odin_> so the N900 never needs to unmount, I had this convo before in here, this would allow ext3 for on-disc/mmc format but provide VFAT via to USB at the same time
Mar 26 20:15:42 <MohammadAG> I'll check with dpkg first :)
Mar 26 20:16:47 <Kurppa_> I hope they will get rid of this rootfs and /home/user hassle.
Mar 26 20:17:00 <ptl> odin_: my /usr/bin/mmc-mount script is the following:
Mar 26 20:17:14 <MohammadAG> Kurppa_, nokia? how?
Mar 26 20:17:18 <ptl> oops
Mar 26 20:17:19 <ptl> sorry
Mar 26 20:17:28 <ptl> it's /usr/sbin/mmc-mount
Mar 26 20:17:30 <ptl> contents:
Mar 26 20:17:37 <ptl> #!/bin/sh
Mar 26 20:17:38 <ptl> mount -t vfat -o $3,noauto,nodev,noexec,nosuid,noatime,nodiratime,utf8,uid=29999,shortname=mixed,dmask=000,fmask=0133,rodir "$1" "$2" > /dev/null || mount -o $3,noauto,noatime,nodiratime "$1" "$2" > /dev/null
Mar 26 20:17:51 <ptl> I just added the part with the "||" and after.
Mar 26 20:18:21 <ptl> it tries to mount as vfat, if it can't, it tries again with type autodetected.
Mar 26 20:18:29 <odin_> yes the rootfs issue is a major one as to contributing for me
Mar 26 20:19:32 <odin_> i.e. I worked out how much time I would be spending maintaining that issue in order to release packages and then thought about what the reward for it was, when Im sure the next model will have a bigger OneNAND/eMMC
Mar 26 20:20:14 <ml-mobile> it might have 512M if they've already started
Mar 26 20:20:33 <Kurppa_> I mean the rootfs in N900 was so small.
Mar 26 20:20:48 <odin_> remind us of the reasons for root being on the OneNAND, was it performance ?  it has a software based wear-leveler provided by part of UBIFS's infrastructure, but remind me what benefit we get again
Mar 26 20:20:55 <cehteh> small isnt the problem, the way its set up is
Mar 26 20:20:57 <Kurppa_> They fixed it partly using that /opt, but there are still some problems.
Mar 26 20:21:11 <ml-mobile> speed
Mar 26 20:21:21 <asj> odin_: I'm guess it's a very high number of read/write cycles
Mar 26 20:21:23 <odin_> speed of what ?  read's ?  write's ?
Mar 26 20:21:29 <ml-mobile> OneNAND is much faster than eMMC
Mar 26 20:21:30 <ml-mobile> both
Mar 26 20:21:55 <ml-mobile> the thing is connected directly to the CPU as well
Mar 26 20:21:58 <odin_> and the I/O stats of the FS indicate what life expectancy of the flash ?
Mar 26 20:22:18 <odin_> so why isn't it used as a cache?  main rootfs on eMMC is loaded into OneNAND
Mar 26 20:22:39 <odin_> writes back to eMMC are journaled through OneNAND, something like that
Mar 26 20:22:50 <odin_> the /tmp and such dirs are 100% on OneNAND
Mar 26 20:23:27 <asj> odin_: eMMC may only have 10,000 writes, I  don't know the part
Mar 26 20:23:42 <ml-mobile> erases*
Mar 26 20:23:54 <ShadowJK> "something like that" isn't a viable argument you take to managers :-)
Mar 26 20:24:24 <ShadowJK> Hopefully LogFS will be ready for the next device, it should have performance almost as good as ubifs on onenand
Mar 26 20:24:25 <odin_> I thought modern flash was in the order of 100k+
Mar 26 20:24:39 <ShadowJK> (although logfs author did say it'd be slower on emmc... oh well)
Mar 26 20:24:40 <ptl> what about "something unlike that"?
Mar 26 20:24:41 <odin_> no the point I am making is
Mar 26 20:24:46 <asj> odin_: depends on the device
Mar 26 20:24:47 <ptl> It would appeal to the NIH Syndrome
Mar 26 20:24:47 <ml-mobile> SLC nand at higher lithographies
Mar 26 20:24:48 <ShadowJK> slower than logfs on onenand i mean
Mar 26 20:25:08 <ShadowJK> ml-mobile, the speed thing is not really a SLC vs MLC issue... endurance, yes
Mar 26 20:25:27 <odin_> the OneNAND resource is a resource, if its purpose it to increase life expectancy and performance because it has a high a access speeds and also high number of erases possible, also maybe at smaller graundlarity
Mar 26 20:25:58 <odin_> then treat it as such in software, make it a cache/journal for a real logical rootfs which is really back by eMMC
Mar 26 20:26:19 <odin_> *backed by eMMC, its just that some useful often read/written sector have an alias in OneNAND
Mar 26 20:26:25 <ml-mobile> sure, but what are you going to use to achieve that?
Mar 26 20:26:35 <odin_> software
Mar 26 20:26:35 <ShadowJK> There are no such drivers right now
Mar 26 20:26:42 <ml-mobile> no shit
Mar 26 20:26:46 <ShadowJK> I'd rather invest in making one of the nonsucky filesystems ready..
Mar 26 20:26:48 <Kurppa_> Cost-wise, if the rootfs is situated in eMMC, I don't think they would even add an OneNAND chip there.
Mar 26 20:27:12 <odin_> well this is the issue... I do not know the commercials on future OneNAND use
Mar 26 20:27:28 <microlith> why am I typing on my N900 when I can use a real keyboard
Mar 26 20:27:33 <odin_> are we going to get mode for less money ?  is it likely the next device(s) will use OneNAND, is it here to stay etc...
Mar 26 20:27:48 <microlith> OneNAND is here to stay, I suspect
Mar 26 20:28:12 <odin_> if the next device in 2 years will have a 2Gb OneNAND for the same price point then there is no problem staying as-is
Mar 26 20:28:20 <Kurppa_> A big priority of a big company is to push production costs down. I guess only future will tell though :-)
Mar 26 20:28:25 <MohammadAG> how do I patch a folder (with files and subdirectories) from a .diff file?
Mar 26 20:28:35 <microlith> odin_: well, the solution for that will be decided in MeeGo
Mar 26 20:28:42 <odin_> if the next device is also going to have a 256Mb OneNAND because of price/problems/whatever then a solution to the rootfs needs to be found
Mar 26 20:29:03 <ShadowJK> the solution is ./configure --prefix=/opt/maemo ;D
Mar 26 20:29:10 <microlith> I suspect the next defvice will have at least 512M
Mar 26 20:29:13 <microlith> -f
Mar 26 20:29:19 <odin_> no I've looked at that
Mar 26 20:29:40 <odin_> its too costly to maintain from my PoV, I want a regular Unix system single main rootfs logical view
Mar 26 20:29:43 <Kurppa_> There is still the problem of that 2GB partition. It's too small.
Mar 26 20:29:45 <RST38h> Steve Jobs and Eric Schmidt Spotted Together Again !
Mar 26 20:29:52 <RST38h> oh-no-they-are-gonna-mate
Mar 26 20:29:58 <odin_> I'm not going to spend hours doing "optification" for no net benifit
Mar 26 20:30:05 <odin_> Im sure others will spend that time
Mar 26 20:30:09 <microlith> odin_: optification is as easy as ShadowJK just said
Mar 26 20:30:11 <ShadowJK> odin_, you know things like *BSD don't put user installed packages in /usr/*? It's a linuxism really :)
Mar 26 20:30:14 <RST38h> odin: optification takes <30-40 mins
Mar 26 20:30:23 <lcuk> if only!
Mar 26 20:30:24 <RST38h> odin: even less if you want to be messy about it
Mar 26 20:30:33 <asj> interesting, the OneNAND part I pulled up is listed at "one year retention" after 100k prog/erase cycles
Mar 26 20:30:38 <odin_> how are you optifying locale's ?
Mar 26 20:30:43 <microlith> consider that no other devices have to deal with this issue
Mar 26 20:30:51 <asj> glad it has error correction built in
Mar 26 20:31:14 <ShadowJK> microlith, uh, they just have "ran out of space :(" instead of "do this to not run out of space"
Mar 26 20:31:31 * lcuk raises a glass to maemo
Mar 26 20:31:39 <Kurppa_> lcuk: cheers.
Mar 26 20:32:19 * microlith hopes PR1.2 resolves the annoying CPU eating behavior of the background services
Mar 26 20:32:19 <Kurppa_> Or MeeGo, as they will probably call it.
Mar 26 20:32:28 <ShadowJK> lol :D
Mar 26 20:32:35 <ShadowJK> MeeGo instance :)
Mar 26 20:33:01 <lcuk> whos seen the latest indiana jones movie
Mar 26 20:33:08 <microlith> ugh, don't remind me
Mar 26 20:33:13 <lcuk> and without teasers, is it any good
Mar 26 20:33:20 <dotblank> lcuk, there are only 3 of them
Mar 26 20:33:29 <lcuk> the "new" one
Mar 26 20:33:32 <Kurppa_> I would say it's no good.
Mar 26 20:33:33 <ShadowJK> As it is now, you just have to try download something while trying to browse the web.. emmc does not like more than one app accessing it :)
Mar 26 20:33:36 <dotblank> lcuk, What no one..
Mar 26 20:33:40 <dotblank> new one*
Mar 26 20:33:44 <Kurppa_> Just my subjective opinion.
Mar 26 20:33:51 <odin_> why is /usr on OneNAND, really it seems that /tmp /var and such should be there, i.e. putting packages into /usr isn't really the issue
Mar 26 20:33:52 <lcuk> i dunno
Mar 26 20:34:02 <lcuk> tracy has dvd downstairs and we are watching soon
Mar 26 20:34:09 <ShadowJK> And stuff that uses databases is also pretty slow.. It takes something like a second for gpodder to sync its database due to the slow I/O :/
Mar 26 20:34:11 <lcuk> i just want to know if its as good as old ones
Mar 26 20:34:12 <Kurppa_> ShadowJK: there will be a fix to that. :-)
Mar 26 20:34:12 <dotblank> It isnt no indiana jone's I ever heard of
Mar 26 20:34:16 <Kurppa_> Download I mean.
Mar 26 20:34:33 <odin_> I must confess I am certainly not at fan of Sun's way, which I guess maybe like BSDs way
Mar 26 20:34:34 <lcuk> ShadowJK, i tried doing database work
Mar 26 20:34:43 <lcuk> and got a massive sketch db created
Mar 26 20:34:56 <dotblank> did someone get jack to work?
Mar 26 20:34:56 <lcuk> but over a threshhold it lost its fluidity
Mar 26 20:35:22 <ShadowJK> Basically a 4K write on emmc is as costly as a 256k write.. And a database update is a bunch of small writes all over the place..
Mar 26 20:35:31 <ShadowJK> and while it's doing those writes all reads are blocked..
Mar 26 20:35:57 <lcuk> ShadowJK, ok
Mar 26 20:36:01 <ShadowJK> (and I assume no filesystem overhead here, which there's tons of)
Mar 26 20:36:08 <lcuk> so if there was a database planned around that sort of limitation
Mar 26 20:36:10 <lcuk> could it be fast
Mar 26 20:36:11 <Kurppa_> Sounds really slow.
Mar 26 20:36:20 <ShadowJK> lcuk, yes
Mar 26 20:36:31 <lcuk> so how would you spec something like that
Mar 26 20:36:41 <lcuk> and is sqlite or similar able to have this engine underneath
Mar 26 20:36:41 <ShadowJK> Just make it one big journal
Mar 26 20:36:47 <ShadowJK> doubt it
Mar 26 20:36:50 <lcuk> im thinking about a regular front end
Mar 26 20:36:53 <lcuk> and a speedy backend
Mar 26 20:37:02 <lcuk> for compatability
Mar 26 20:37:14 <ShadowJK> The "real solution" would be a filesystem aware of the special properties of emmc
Mar 26 20:37:36 <odin_> anyhow, what is the answer to optification of locale files ?  death by 100 sym-links ?
Mar 26 20:38:09 <ShadowJK> odin_, iirc entire locale directory gets optified in PR1.2.. but that's just a "rumor" I heard :)
Mar 26 20:38:09 <lcuk> odin_,
Mar 26 20:38:12 <Kurppa_> Let's hope the next Maemo version will be designed for that. I know it is now too late for Maemo 5 at least :-)
Mar 26 20:38:13 <lcuk> no
Mar 26 20:38:30 <lcuk> but whoever said up there about speed is likely right
Mar 26 20:38:44 <lcuk> arent the locales ok to stay where they are
Mar 26 20:39:00 <lcuk> they are relatively small, required to be fast
Mar 26 20:39:01 <lcuk> etc
Mar 26 20:39:05 <lcuk> i know theres a lot of them
Mar 26 20:39:12 <odin_> in /usr/share/locale/en_GB/LC_MESSAGES/foobar.mo
Mar 26 20:39:46 <ShadowJK> nilfs2 is really speedy on my sheevaplug's attached USB flash drives.. does my "copy 2 gigs of squid spool" test in 16 minutes.. btrfs 46 minutes, ext2/ext3 110 minutes
Mar 26 20:40:17 <odin_> making them optional would help, i.e. apt/dpkg doesn't seem to support that concept to say my system doesn't need certain locale because I disabled all by the language I want
Mar 26 20:40:37 <ShadowJK> unfortunately nilfs2 also generates huge amounts of read/write that makes it unsuitable for production use on flash storage imo
Mar 26 20:40:41 <odin_> what does BSD do with locale then ?
Mar 26 20:41:06 <odin_> is it glibc based too isn't it ?  or is that just one flavor
Mar 26 20:42:01 <ShadowJK> odin_, so on NetBSD everything user installed beyond the base system goes into /usr/pkg
Mar 26 20:42:17 <ShadowJK> So presumable locale would end up in /usr/pkg/share/locale
Mar 26 20:42:27 <ShadowJK> and the binary in /usr/pkg/bin
Mar 26 20:42:27 <ShadowJK> etc
Mar 26 20:42:49 <odin_> you mean /usr/pkg is an alternate root like --prefix in GNU autoconf
Mar 26 20:42:54 <ShadowJK> yes
Mar 26 20:43:26 <odin_> ok this I alread been doing as /opt/maemo/share/... eetc...
Mar 26 20:43:36 <janneg> ShadowJK: have you tried ext4? that was around factor 5 faster handling gentoo's portage tree. on a traditionell hdd though
Mar 26 20:43:56 <ShadowJK> janneg, yeah it was only twice as slow as nilfs2 ;)
Mar 26 20:44:05 <ShadowJK> 30 minutes or so iirc..
Mar 26 20:44:06 <janneg> faster sa btrfs
Mar 26 20:44:25 <ShadowJK> yeah it was faster than btrfs
Mar 26 20:44:41 <ShadowJK> and I even had btrfs' ssd mode on.. not that it made any difference :P
Mar 26 20:45:52 <odin_> ok I have a debhelper-5.0.42.20100131.dlm.diff to fix it for optification purpose in maemo
Mar 26 20:46:41 <cpasjuste> someone to take my n64 work to continue ?
Mar 26 20:50:06 <odin_> so I also do a diff for debhelper-7.0.15 which was requested then it needs to get into the SDK repo I believe to be any use to anyone
Mar 26 20:50:40 <DocScrutinizer51> microlith: odin_: Kurppa_: oneNAND is a part of the OMAP SoC aiui. This is very unlikely to change, as TI had to produce a customized SoC for Nokia
Mar 26 20:51:23 <ShadowJK> I thought the RAM and onenand was in same PoP ontop of omap3 soc
Mar 26 20:51:36 <odin_> DocScrutinizer51, yes understood on that, my question was over the future.. as in is it getting cheaper? how fast?  how long to double size?  how long to be 4Gb which is large enough for a flat rootfs layout
Mar 26 20:51:51 <DocScrutinizer51> that's what I say
Mar 26 20:52:08 <SpeedEvil> ShadowJK, the onenand is soldered onto the omap - which also has solme internal Ram and flash
Mar 26 20:52:21 <SpeedEvil> or I may be misreading stuff
Mar 26 20:52:54 <odin_> why is swap space on eMMC when OneNAND has massive speed and erase potential ?
Mar 26 20:53:01 <ShadowJK> iirc beagle has/had same 256ram/256nand configuration too
Mar 26 20:53:15 <ShadowJK> odin_, linux doesn't know how to swap on nand
Mar 26 20:53:21 <DocScrutinizer51> odin_: you,re more likely to see a system booting from eMMC, rather than 2GB nand
Mar 26 20:53:31 <Kurppa_> OneNAND was too small to house wwap space anyway.
Mar 26 20:53:41 <odin_> well there is also the issue of 1Gb swap in N900 configuration and OneNAND total size
Mar 26 20:53:43 <Kurppa_> swap*
Mar 26 20:53:48 <ShadowJK> Well, there's a driver that emulates emmc ontop of nand, but the end result is something even slower than the emmc :)
Mar 26 20:54:31 <ShadowJK> I had this idea for a loop driver to run ontop of emmc to make swap faster.. but alas I'm no kernel coder :)
Mar 26 20:55:35 * SpeedEvil ponders going to sleep.
Mar 26 20:55:57 <Kurppa_> I'll do the same. Night folks :-)
Mar 26 20:56:13 <odin_> is it midnight there?  gnite those off
Mar 26 20:56:22 <DocScrutinizer51> basically /boot should be a partiton on Nand and be mounted to / on eMMC
Mar 26 20:57:15 <odin_> well yes the whole boot loader and the resources I agree but for N900 24Mb should give it plenty
Mar 26 20:57:33 <odin_> how much space on /var and /tmp do you need at runtime?
Mar 26 20:57:48 <ShadowJK> depends on whether you have extras-devel enabled or not :D
Mar 26 20:58:08 <odin_> sigh
Mar 26 20:58:27 <odin_> nice and scalable engineering
Mar 26 20:58:48 <ShadowJK> We didn't even have optification in maemo4 :)
Mar 26 20:58:54 * ShadowJK didn't run out of space on / either
Mar 26 20:58:58 <RST38h> but we did have data-on-mmc
Mar 26 20:59:01 <ShadowJK> so I dunno what you guys are doing ;)
Mar 26 20:59:17 <ShadowJK> RST38h, yeah, but ~ was still on nand too
Mar 26 20:59:23 <ShadowJK> now /home is on emmc
Mar 26 21:02:26 <odin_> so 128Mb for writable parts of the FS /var /tmp etc... that leaves 104Mb of OneNAND to use for rootfs journal / performance read-cache from rootfs backed pages
Mar 26 21:03:52 <odin_> I believe EXT3/4 outsource their journal support, so its possible to write a journal specific device driver, such as a journal to OneNAND or UBIFS partition
Mar 26 21:04:08 <ShadowJK> I'd want to experiment with putting a gigantic ext3/ext4 journal on emmc.. and unwind it when device is idle..
Mar 26 21:04:36 <ShadowJK> the journal due to its very nature should operate shitloads faster on emmc than the actual filesystem itself :-)
Mar 26 21:05:08 <ShadowJK> So you'd want it reversed... if the sizes of the devices were reversed too, of course :(
Mar 26 21:05:19 <odin_> why gigantic ? ext3 is designed to cope with a lot of dirty block on journal but not yet flushed?
Mar 26 21:06:08 <c-s-b-n900> wl1251: ERROR Power save entry failed, giving up
Mar 26 21:06:09 <odin_> I dont understand you logic, to me you want your journal read/write speed to be fast and you want it on media which can take the erase battering over the lifetime of the device
Mar 26 21:06:15 <c-s-b-n900> this is a pain...
Mar 26 21:06:17 <MiXu-> anyone installed the crash reporter?
Mar 26 21:06:34 <MiXu-> it's complaining about missing packages
Mar 26 21:07:04 <odin_> you want journal read/write fast because all data updates are pushed through it, you also want something to cancel out redundant journal updates when its flushed
Mar 26 21:07:34 <Arif_> http://talk.maemo.org/showthread.php?t=48336
Mar 26 21:07:34 <Arif_> lol
Mar 26 21:07:37 <odin_> and there is no pressure to flush it fast unless you are running out of journal room
Mar 26 21:07:50 <ShadowJK> odin_, I want the journal to be big because writing to journal is (presumably) entirely sequential, which is fast on emmc
Mar 26 21:07:51 <odin_> MiXu-, yes I have crash reporter installed, what is hte missing packages?
Mar 26 21:08:17 <odin_> ShadowJK, only if you write out 256MB at a time, but OneNAND granduality is much smaller isn't it ?
Mar 26 21:08:29 <ShadowJK> 256k
Mar 26 21:08:56 <odin_> ShadowJK, presumable solid state does not care for random or sequential writes the issue is the page blockage the lack of pipeline
Mar 26 21:09:08 <ShadowJK> It does care
Mar 26 21:09:14 <odin_> opps yes my magnitude a little off there well corrected
Mar 26 21:09:39 <MiXu-> odin: libhildon1, libosso1, and a few others
Mar 26 21:09:51 <ShadowJK> Basically on emmc, a random 4k write gets translated by the emmc internally to a 256k read-modify-erase-write cycle
Mar 26 21:09:55 <odin_> so you are going to hold onto 256Kb of dirty data in memory to optimised the write to journal... but we want it in persistent storage ASAP.. so the data is safe
Mar 26 21:10:25 <odin_> MiXu-, which version of crash reporter are you trying to install ?  what is your N900 software version ?
Mar 26 21:10:31 <ShadowJK> It's the "erase" function which has 256k granularity, writing onto an erased block can usually be done in 4k chunks
Mar 26 21:11:03 <ShadowJK> So doing several 4k writes sequentially is much faster than 4k random writes
Mar 26 21:11:03 <odin_> ShadowJK, yes and OneNAND page granduality is much smaller, in the order of 4Kb is it not (been a few months since I looked)
Mar 26 21:11:24 <MiXu-> odin_: I have PR1.1.1 and the reporter is 1.29-0m5
Mar 26 21:11:27 <ShadowJK> Well you have the same erase size granularity, I'd think, but you can manage it directly yourself
Mar 26 21:11:43 <ShadowJK> And take advantage of half-erased blocks, etc
Mar 26 21:11:55 <odin_> ShadowJK, I thought page grandulaity was a hardware limitation, large pages in eMCC and small in OneNAND by hardware design
Mar 26 21:12:44 <ShadowJK> Well the FTL part of emmc effectively makes it 256k for everything
Mar 26 21:12:50 <odin_> MiXu-, I have 1.27+0m5 installed myself can you not go back a version, its possible internal nokia guys have released a crash reporter for PR1.1.2 since it is a development repo after all
Mar 26 21:13:12 <ShadowJK> But with onenand you don't have a braindead FTL, the FTL is in the kernel, and in the case of ubifs, it's very clever
Mar 26 21:13:43 <ShadowJK> You know those awesome Intel X-25M SSD drives?
Mar 26 21:13:54 <odin_> MiXu-, "apt-get install crash-reporter=1.27+0m5"  I think
Mar 26 21:14:19 <ShadowJK> They have an ARM processor in it, and some 64megs of RAM to do the "clever" stuff that makes random writes fast and not stutter :)
Mar 26 21:14:26 <odin_> X-25M nope
Mar 26 21:14:28 <ShadowJK> and the ram isn't for caching
Mar 26 21:14:57 <odin_> well it sounds good but no use for the mobile market
Mar 26 21:15:05 <ShadowJK> It's easiest if you think of flash as a tape recorder when you write to it, and as RAM when you read from it ;p
Mar 26 21:15:58 <odin_> but I still don't get your big ext journal to eMMC, the file system isn't designed for huge journals and putting the write page wear on eMMC is just one thing that needs to be aboived?
Mar 26 21:16:01 <odin_> avoided
Mar 26 21:16:20 <ShadowJK> write wear is overrated I think
Mar 26 21:16:46 <odin_> eMMC = slow performance read/write, big granularity, low max write/erase per segment
Mar 26 21:17:03 <odin_> OneNAND = fast performance read/write, fine/small granularity, high max write/erase per segment
Mar 26 21:17:18 <MiXu-> odin_: it worked. thanks :)
Mar 26 21:17:30 <odin_> no contest where an ext3/ext4 journal should go
Mar 26 21:17:32 <ShadowJK> The way I've abused my flash drives.. if wear was an actual concern they'd be dead by now :)
Mar 26 21:18:22 <ShadowJK> constant 3megabytes/sec write for ~2 months now, still alive :P
Mar 26 21:18:27 <odin_> Im supprised they are not like slow dynamic RAM, needing a "refresh" within some many days/months to stay alive
Mar 26 21:19:20 <ml-mobile> flash is getting to that point
Mar 26 21:19:25 <odin_> does linux kernel have any logic to optimize journal playback? for example, data blocks that have been overwritten multiple times
Mar 26 21:19:49 <ShadowJK> not sure :)
Mar 26 21:20:15 <odin_> would show up in an uncommitted journal.. we only want perform a single write at the end, and optimize out the previous ones
Mar 26 21:20:23 <ShadowJK> odin_, the idea about having big journal is that when you use the device, writes go to journal sequentially, which makes it fast and low-latency.. then when it's idle it could play back the journal and do all the slow high-latency random writes :)
Mar 26 21:21:00 <odin_> what I'm trying to get at is OneNAND is ideal for that, to take the hits reducing the load on the eMMC all the kernel needs to track is the device/block of the dirty data in the journal
Mar 26 21:21:43 <odin_> well take the example of a "directory" or "inode" page, this must get hit multiple times as a normal file grows
Mar 26 21:21:47 <ShadowJK> under heavy random write load I've seen emmc take 500msec PER REQUEST, the user complains about "choppy" graphics on screen, then the I/O queue grows to some 10 seconds in length and the user complains the device hangs :)
Mar 26 21:22:18 <odin_> so does Linux kernel already optimize out the multiple writes to the same sector(s) or merge them into 256Kb page chunks possible with scatter-gather scheme to fire to eMMC
Mar 26 21:22:40 <odin_> ok thats a kernel IO scheduling issue
Mar 26 21:24:16 <odin_> what is the read performance issue from eMMC, does a 1Kb read happen any faster than a 64Kb read ?
Mar 26 21:24:31 <ShadowJK> you'd want to try teach the kernel that a 4k write request is as costly as a 256k write request, or almost as costly as 64 4k writes in a row.. and for it to scale the length of it's i/o queue based on maximum time instead..
Mar 26 21:25:05 <odin_> yes you can in the IO scheduler, its pluggable the strategy, it also deals with reordering and regrouping IO requests
Mar 26 21:25:10 <ShadowJK> I think it's slightly faster
Mar 26 21:25:40 <ShadowJK> mmc command overhead becomes the main factor which makes small reads slower than big reads
Mar 26 21:25:44 <odin_> so it can make the same IO request within the same 256Kb chunk be groups together
Mar 26 21:25:51 <ShadowJK> it's still 4 orders of magnitude faster than writing, of course :)
Mar 26 21:25:59 <odin_> even if they were submitted to the IO subsystem out-of-order with request in between
Mar 26 21:26:28 <odin_> ok so there is no real issue with 4Kb reads from eMMC for the purpose of page-faults ?
Mar 26 21:27:00 <ShadowJK> the nilfs2 filesystem is a "Log structured" filesystem.. it just writes the log of stuff that happens... so it's written sequentially.. almost like a tape drive.. that's why it's so fast in my test.. and since random reads don't suffer too much on emmc, it's even fast for reading it back too..
Mar 26 21:27:19 <ShadowJK> in "normal" benchmarks it excells on write performance on harddrives too, but falls flat down when it comes time to read :)
Mar 26 21:27:26 <odin_> or maybe some applications like X11 server and other important system components can be aliased in OneNAND
Mar 26 21:27:36 <GAN900> What was Nokia thinking with that stupid location selection dialog
Mar 26 21:27:40 <GAN900> So confusing
Mar 26 21:28:01 <ShadowJK> Yes, as long as there are no outstanding writes blocking the queue, page-faulting from emmc is just happy happy joy joy
Mar 26 21:28:20 <ShadowJK> I'd like to see a dedicated emmc for OS or swap for this reason :)
Mar 26 21:28:39 <ShadowJK> It might mean the removal of microsd slot though..
Mar 26 21:28:49 <odin_> well reads are usually given priority over writes, because everything blocks when a read blocks but a write is usually buffer cache being flushed async
Mar 26 21:29:35 <ShadowJK> yeah.. though once you've been feeding writes to the emmc for awhile, they start taking .5 sec each, so even if you prioritize reads you still have awful latency
Mar 26 21:29:41 <DocScrutinizer51> ShadowJK: the FTL doesn,t know of such thing like "half-empty blocks", at least not for non-fat fs
Mar 26 21:29:56 <odin_> I don't get that metric, after a while, 500ms per write
Mar 26 21:29:57 <ShadowJK> DocScrutinizer, I was referring to ubi(fs) there
Mar 26 21:30:05 <DocScrutinizer51> ooh k
Mar 26 21:30:42 <odin_> I can understand the silicon FTL is it, it doing extra/other work around our I/O
Mar 26 21:31:04 <odin_> is this what you are claiming
Mar 26 21:31:05 <ShadowJK> odin_, dunno what casuses it, but the svctm value in iostat grows to about 500 or so after awhile :)
Mar 26 21:31:22 <ShadowJK> and await to 5000
Mar 26 21:31:32 <ShadowJK> under heavy io
Mar 26 21:31:44 <odin_> but thats not the case we're trying to optimize
Mar 26 21:32:29 <odin_> heavy io is "heavy io", journal is full, data is being pushed out and causing back pressure to block writes at the application
Mar 26 21:32:41 <ShadowJK> Anyway, for me the solution is called LogFS :)
Mar 26 21:32:49 <ShadowJK> and in case logfs turns out to be crap, nilfs2 :D
Mar 26 21:33:45 <ShadowJK> http://mirror.linux.org.au/pub/linux.conf.au/2007/video/talks/91.ogg <- talk by logfs author on the nature of flash and logfs' garbage collection algorithm :)
Mar 26 21:33:49 <odin_> the main case is normal usage patterns, bootup, background stuff going on, system idling and then being used for a short time and so on
Mar 26 21:33:56 <ShadowJK> it's a good presentation :)
Mar 26 21:34:58 <lcuk> can someone confirm, does it really take >2 mins for a crackberry to boot?
Mar 26 21:35:02 <lcuk> or isthat just older ones
Mar 26 21:35:09 <DocScrutinizer51> ~weather EDDN
Mar 26 21:35:11 <odin_> but you seems to be solving some kind of write issue ?  and I don't see that being so important for a mobile device, it is important for loading up maybe via USB or internet with content but this is not a often used case, unlike say a database server
Mar 26 21:35:30 <DocScrutinizer51> ~ping
Mar 26 21:35:39 <odin_> ~seen me
Mar 26 21:35:53 <ShadowJK> odin_, well loading up via usb is pretty minimal load
Mar 26 21:36:04 <ShadowJK> odin_, the biggest annoyances come from database loads ;)
Mar 26 21:36:09 <ShadowJK> and there are a ton of them
Mar 26 21:36:25 <DocScrutinizer51> ~weather EDDN
Mar 26 21:36:29 <infobot> Nuernberg, Germany; (EDDN) 49-30N 011-03E 318M; last updated: 2010.03.26 2120 UTC; Dew Point: 41 F (5 C); Pressure (altimeter): 29.83 in. Hg (1010 hPa); Relative Humidity: 87%; Sky conditions: overcast; Temperature: 44 F (7 C); Visibility: greater than 7 mile(s); Weather: Light rain; Wind: from the W (260 degrees) at 12 MPH (10 KT)
Mar 26 21:36:34 <ShadowJK> media player/tracker, gpodder, conversations, email... some of the things I use :)
Mar 26 21:36:43 <DocScrutinizer51> that's bitter
Mar 26 21:37:09 <ShadowJK> like I said earlier, deleting a file in gpodder is a 1.5 sec operation due to the database update :)
Mar 26 21:37:21 <DrGrov> Good evening
Mar 26 21:37:22 <odin_> ShadowJK, but that is a software issue, for expecting "Winchester disk performance" metric, 512byte sectors, seek time, an IO queue
Mar 26 21:37:53 <ShadowJK> It's not even writing that much, about 1600 bytes..
Mar 26 21:37:55 <odin_> ShadowJK, solvable by using a journal based file format on top of a general purpose FS
Mar 26 21:38:11 <ShadowJK> the fdatasync() takes the most time :)
Mar 26 21:38:18 <lcuk> does it take 1..5 seconds to just delete a file?
Mar 26 21:38:31 <lcuk> or a few
Mar 26 21:38:40 <ShadowJK> one point five
Mar 26 21:38:47 <odin_> ShadowJK, I think that is one issue ext4 tried to address, but another issue is do you really need to use fdatasync() ?
Mar 26 21:38:50 <ShadowJK> plus minus point five ;D
Mar 26 21:38:52 <lcuk> what, every file takes 1.5s?
Mar 26 21:39:01 <DrGrov> I managed to figure out one cool thing today
Mar 26 21:39:04 <ShadowJK> lcuk, i think mass delete is somewhat faster
Mar 26 21:39:12 <lcuk> im on about a specific file
Mar 26 21:39:16 <lcuk> named
Mar 26 21:39:46 <lcuk> or something
Mar 26 21:39:51 <lcuk> /user/home/myanonymouspoddingapp/posts/post20100326
Mar 26 21:39:51 <ShadowJK> selecting 5 files for download, between tapping download and the UI responding again is like 5 seconds
Mar 26 21:40:27 <ShadowJK> odin_, i'm glad there was fdatasync there, gpodder db survived crash :)
Mar 26 21:40:29 <lcuk> or to rename files
Mar 26 21:40:32 <lcuk> or something
Mar 26 21:41:01 <ShadowJK> lcuk, it's slow because gpodder updates its own database to remove the file
Mar 26 21:41:09 <odin_> ShadowJK, ah I remember now the issue with fdatasync() was block allocation on ext3
Mar 26 21:41:32 <ShadowJK> odin_, i think it translates into sync() ;)
Mar 26 21:41:36 <odin_> ShadowJK, if a database pre-allocated new pages for data in the file, then the fdatasync() is faster since there is no requirement to find a new block to allocate
Mar 26 21:41:42 <lcuk> of course it does shadow
Mar 26 21:41:57 <lcuk> i wonder just whats happening
Mar 26 21:42:05 <lcuk> cos dont most databases not fold up the space
Mar 26 21:42:06 <iBob> Hello!
Mar 26 21:42:12 <lcuk> when you delete records
Mar 26 21:42:17 <odin_> ShadowJK, this is a software design issue that can be fixed/improved by extending the file length when the system is more idle
Mar 26 21:42:34 <DrGrov> I found some extra themes in some other repository, not any of the default ones. How can I access this repo and download the themes?
Mar 26 21:42:41 <DrGrov> Or add it somehow into the repo list?
Mar 26 21:43:09 <ShadowJK> odin_, ok, next issue, when downloading files from internet there's increased latency using apps ;)
Mar 26 21:43:39 <odin_> ShadowJK, over wireless or 3G/GSM/GPRS ?
Mar 26 21:43:39 <ShadowJK> the page faults take longer to complete from emmc, swap is slower, etc :/
Mar 26 21:43:40 <DocScrutinizer51> DrGrov: add the repo to HAM catalogs
Mar 26 21:43:45 <SpeedEvil> or when the tracker runs
Mar 26 21:43:51 <ShadowJK> odin_, HSPA 5Mbit/s
Mar 26 21:43:56 <DrGrov> DocScrutinizer51: aha, so how exactly do I know how to add it?
Mar 26 21:44:08 <DrGrov> DocScrutinizer51: HAM catalogs you mean in the Package Manager?
Mar 26 21:44:16 <DocScrutinizer51> yep
Mar 26 21:44:34 <DrGrov> DocScrutinizer51: ok, so I can manually put other repos there? Oh, I didn't know that. Thanks :)
Mar 26 21:44:36 <odin_> ShadowJK, when you say downloading... do you mean browsing or saving a file to eMMC ?
Mar 26 21:44:43 <ShadowJK> saving files
Mar 26 21:44:49 <iBob> is w51-1 (PR1.1) the latest firmware update ?
Mar 26 21:44:59 <odin_> ShadowJK, does the download rate exceed the eMMC max-write speed rate ?
Mar 26 21:45:04 <ShadowJK> no
Mar 26 21:46:21 <DrGrov> DocScrutinizer51: so, can I then later on remove it so it doesn't interfer with the default repos?
Mar 26 21:46:27 <iBob> and is nokia going to keep providing updates after the release meego?
Mar 26 21:46:34 <DocScrutinizer51> DrGrov: depending on how well the site is done, klicking a link on a website might suffice to add repo and install app
Mar 26 21:46:35 <odin_> what IO scheduler is in use in the stock kernel ?  how does it manage the priority of a page-fault-in verses the download disk-write-out
Mar 26 21:46:48 <DocScrutinizer51> DrGrov: yes
Mar 26 21:47:14 <DrGrov> DocScrutinizer51: so I can go to that webpage as well and install from there?
Mar 26 21:47:16 <odin_> certainly under my scheme all writes to eMMC from the ext3 journal now on the OneNAND would be 256Kb wide, which would achieve maximum write performance ?
Mar 26 21:47:23 <ShadowJK> CFQ
Mar 26 21:47:29 <pwnguin> man. i really need my banshee generated playlist =(
Mar 26 21:47:38 <DocScrutinizer51> DrGrov: depends
Mar 26 21:47:49 <DrGrov> DocScrutinizer51: now just the problem I can't install the damn theme. Can't find it in maemo.org and the downloads section
Mar 26 21:48:04 <odin_> but there is still a priority allocated in the kernel paging system and IO system there may even be things to tweak on that
Mar 26 21:49:15 <DocScrutinizer51> DrGrov: hmm, I can't find it either :-P
Mar 26 21:50:23 <odin_> where are you writing too ?  VFAT /home/user/MyDocs ?
Mar 26 21:50:27 <DrGrov> DocScrutinizer51: damn! :)
Mar 26 21:50:51 <DrGrov> DocScrutinizer51: nah, I will get it some other day. I have the default Nseries blue which is absolutely beautiful
Mar 26 21:51:03 <DocScrutinizer51> damn theme: not found
Mar 26 21:51:40 <DrGrov> DocScrutinizer51: lol
Mar 26 21:51:48 <iBob> theres this update called maemo 5 3.2010.02-8
Mar 26 21:51:49 <ShadowJK> odin_, yes
Mar 26 21:52:03 <iBob> is this an upgrade for the OS ?
Mar 26 21:52:55 <odin_> of course the issue with my scheme is that it only affects ext3 performance, so VFAT is not affected.. however if I go back to my original intent for OneNAND to be a cache, there is no reason not to journal the entire eMMC over UBIFS partition but provide an specific interface for ext3 usage as well
Mar 26 21:52:56 <iBob> i already installed Maemo 5/PR1.1 , do i need to installed this one also ?
Mar 26 21:54:00 <odin_> FWIW there would be no need for UBIFS compression support since that is not a goal here, mainly to nick the page write counters
Mar 26 21:55:32 <lcuk> iBob, that update 3.2010.02-8 is a minor update to prepare for ovi store stuff afaik.
Mar 26 21:55:42 <lcuk> its not the pr1.2 you may have heard about more recently
Mar 26 21:55:49 <lcuk> its 1.1.1
Mar 26 21:56:11 <iBob> i have not heard if the pr1.2
Mar 26 21:56:24 <lcuk> well thats all good then :p
Mar 26 21:56:38 <iBob> i've read some stuff that seem to hint that nokia isnt going to be providing updates
Mar 26 21:56:47 <iBob> because of the deleopment of meego
Mar 26 21:56:50 <lcuk> (i only mentioned it because this last week the pr1.2 sdk was released, but its not a flashable image)
Mar 26 21:57:03 <lcuk> well :) the hints are wrong
Mar 26 21:57:09 <DocScrutinizer51> iBob: that's nonsense
Mar 26 21:57:19 <iBob> great.
Mar 26 21:57:22 <lcuk> theres fully documented lists around of the massive number of fixes and changes in the pr1.2 :)
Mar 26 21:57:32 <trem> nite all, sweet dreams
Mar 26 21:57:34 <user_> hiiii
Mar 26 21:59:03 <user_> heyy
Mar 26 22:00:06 <user_> hi people
Mar 26 22:00:42 <DocScrutinizer51> user_: hi. Any more intelectual stuff to contribute?
Mar 26 22:01:23 <iBob> i keep reading stuff regarding disabling the unstable testing repos. after installing software from them
Mar 26 22:01:36 <iBob> its safe to keep them if i dont apt-get upgrade, right?
Mar 26 22:02:04 <odin_> well it makes sense to release the 1.2 SDK earlier than the N900 images, due to the breaking ABI changes for Qt
Mar 26 22:02:04 <ShadowJK> Well they'd also give you upgrades to software you've already installed, untested upgrades :)
Mar 26 22:02:16 <adalal> hey, anyone know how to get the gps on the n900 to transfer data over bluetooth serial?
Mar 26 22:02:19 <ShadowJK> and it eats diskspace just to have them enabled
Mar 26 22:02:35 <Corsac> adalal: I think using mini-gpsd should do the trick
Mar 26 22:02:41 <odin_> I can only support this action, so by the time the flashable image is ready there is a change all the application space is too
Mar 26 22:02:49 <user_> i need an information
Mar 26 22:02:53 <lcuk> does that happen
Mar 26 22:02:57 <lcuk> ShadowJK,
Mar 26 22:03:02 <lcuk> if i install app from extras
Mar 26 22:03:06 <lcuk> then enable -devel
Mar 26 22:03:13 <iBob> ShadowJK : i gui package manger shows all the updates but i install them one by one.
Mar 26 22:03:16 <lcuk> and theres a newer/unstable version in -devel
Mar 26 22:03:21 <lcuk> does it offer an upgrade?
Mar 26 22:03:24 <user_> do u know when can we get the PR1.2 update???
Mar 26 22:03:35 <iBob> the=i*
Mar 26 22:03:50 <Corsac> user_: when it's ready is the more appropriate answer
Mar 26 22:04:00 <iBob> user_ : when its released.
Mar 26 22:04:15 <user_> yes but when :(
Mar 26 22:04:28 <adalal> Corsac: oh right
Mar 26 22:05:15 <iBob> user_ : http://talk.maemo.org/showthread.php?t=42285
Mar 26 22:05:20 <user_> one more question.... a java me gui for fremantle???
Mar 26 22:05:57 <user_> valid java me i mean not microemu or demos versions
Mar 26 22:07:49 <c-s-b-n900> hello docscrutinizer
Mar 26 22:09:17 <adalal> Corsac: it has missing depencies (bluez-util-test and python2.5-gnome)
Mar 26 22:10:16 <Corsac> adalal: I have to admit I didn't tested it myself, it was just a random idea
Mar 26 22:10:31 <adalal> right, can gpsd be transferred over bluetooth?
Mar 26 22:10:51 <ma3mo> people... one more question... will be possible to add 3g videocall features in the n900? maybe in future? what do you think boput it???
Mar 26 22:11:33 <Corsac> ma3mo: I think it's in PR1.2
Mar 26 22:11:39 <ShadowJK> I doubt it
Mar 26 22:12:01 <ShadowJK> I think 1.2 will have Google Talk video calls, and that'll be it :P
Mar 26 22:12:04 <ma3mo> corsac its voip call that
Mar 26 22:12:08 <iBob> fuck it, we dont even have such a service here , hehe
Mar 26 22:12:33 <ma3mo> yes gtalk and more videocall plugins for these servers
Mar 26 22:12:50 <ShadowJK> 3G anything is a big task, because 3g is almost everything that internet does, except it's all different and doesn't work on anything else or anywhere else... lots of effort and code to write from scratch
Mar 26 22:13:53 <iBob> i'd say they should work on getting the menus a smooth as possible
Mar 26 22:14:11 <iBob> not as good as the iphone , but as good as possible :P
Mar 26 22:15:06 <ma3mo> ok..... no 3g calls... java support??? when???
Mar 26 22:15:09 * lbt wonders why fremantle doesn't use ntp...
Mar 26 22:15:19 <DocScrutinizer51> that'd imply "good" being an absolute metric
Mar 26 22:15:43 <ShadowJK> 3g voice calls, sure :)
Mar 26 22:15:50 <Corsac> ShadowJK: I already tried jabber calls, and it worked fine, though video failed
Mar 26 22:15:53 <Corsac> (on 1.1)
Mar 26 22:15:59 <ShadowJK> I doubt anyone will do 3g video calls
Mar 26 22:16:11 <ShadowJK> and java will probably never have a user friendly solution...
Mar 26 22:16:17 <Corsac> ShadowJK: why not?
Mar 26 22:16:20 <ShadowJK> that'd require giving Sun.. err... Oracle lots of money
Mar 26 22:16:29 <ma3mo> what a shame of phone
Mar 26 22:16:30 <Corsac> (if the carrier allows it)
Mar 26 22:16:54 <DocScrutinizer51> ma3mo: stop trolling
Mar 26 22:16:54 <ma3mo> i have i stuck pixel one bright spot on display
Mar 26 22:17:03 <iBob> ma3mo : respect y
Mar 26 22:17:05 <ShadowJK> Corsac, mostly because it's behind some 5000-page spec from the 3g consortium..
Mar 26 22:17:15 <iBob> yourself*!
Mar 26 22:17:16 <Corsac> ShadowJK: eh?
Mar 26 22:17:28 <Corsac> ShadowJK: I meant jabber video call on 3g network
Mar 26 22:17:44 <ma3mo> sorry what isad???
Mar 26 22:17:48 <Corsac> ShadowJK: were you referring to something else?
Mar 26 22:18:01 <ShadowJK> Corsac, jabber is internet
Mar 26 22:18:06 <ShadowJK> Corsac, I was referring to 3g video calls
Mar 26 22:18:11 <ma3mo> shame couse its a shame device
Mar 26 22:18:19 <ma3mo> i m real not a dreamer
Mar 26 22:18:19 <iBob> ma3mo : would you like some apple pie ?
Mar 26 22:18:21 <Corsac> ShadowJK: 3g usually means data, and inet
Mar 26 22:18:29 <DocScrutinizer51> ma3mo: internal QA rules of Nokia require zero pixel-error screen for N900
Mar 26 22:18:35 <Corsac> ShadowJK: but yeah I'm not sure we'll see operator-only video stuff
Mar 26 22:18:40 <ShadowJK> Corsac, yes but I specifically meant 3g video calls
Mar 26 22:18:49 <ShadowJK> not internet video calls using 3g for internet connectivity.
Mar 26 22:18:58 <Corsac> yeah, sorry
Mar 26 22:20:07 <DocScrutinizer51> ma3mo: so instead of crying here you should either get a error free device, or be a little more constructive
Mar 26 22:20:46 <DocScrutinizer51> maybe both
Mar 26 22:20:55 <ma3mo> i have 1 dead pixel and one bright spot on lcd... no java... no jre... no 3g calls... no irda ftp between devices... no options to take pictures from frontal camera... this is constructive i m real i m talking about this device
Mar 26 22:21:12 <ma3mo> ;)
Mar 26 22:21:26 <iBob> well, i enjoy running aircrack wireshark and lighttpd on my mobile phone
Mar 26 22:21:56 <iBob> still has some bugs? yes. do I give a fuck ? no.
Mar 26 22:21:58 <ShadowJK> ma3mo, java, jre, 3g video calls and irda was never part of the N900 specs :)
Mar 26 22:22:06 <ShadowJK> the n900 doesn't even have the hardware for irda
Mar 26 22:22:13 <ma3mo> it has
Mar 26 22:22:23 <ma3mo> it has irda port
Mar 26 22:22:33 <Myrtti> irda? who the fuck uses irda for anything these days? apart from emulating a Logitech harmony
Mar 26 22:22:35 <ma3mo> disassembly and have a look
Mar 26 22:22:38 <DocScrutinizer> ma3mo: no, you're not constructive. We all *know* there's no java (or we know why you think there isn't), and we know there's no IrDA
Mar 26 22:22:48 <ShadowJK> ma3mo, it's not irda
Mar 26 22:22:56 <ma3mo> what is it???
Mar 26 22:22:58 <DocScrutinizer> so wtf should we benefit from your wining
Mar 26 22:23:04 <ShadowJK> ma3mo, it's CIR
Mar 26 22:23:08 <ma3mo> shad what is it
Mar 26 22:23:15 <ma3mo> so???
Mar 26 22:23:22 <ShadowJK> ma3mo, if you disassemble the N900 you'll notice that it has 3 IR emitters :)
Mar 26 22:23:25 <ShadowJK> None of them can do irda
Mar 26 22:23:26 <ma3mo> c??? Infra Red??
Mar 26 22:23:38 <iBob> ma3mo has to be one of those windows mobile users!
Mar 26 22:23:42 <DocScrutinizer> ma3mo: and no, it has NO irda port
Mar 26 22:23:47 <ma3mo> noooo
Mar 26 22:23:52 <ma3mo> i was symbian
Mar 26 22:23:57 <iBob> we dont need an irda port!
Mar 26 22:23:57 <ma3mo> wm too
Mar 26 22:24:22 <ShadowJK> ma3mo, one of the emitters at the top can be used to send signals that control televisions and such... with the right apps installed
Mar 26 22:24:25 <iBob> i wana see you guys port wireshark to wm :P
Mar 26 22:24:43 <Kegetys> I wonder if the proximity sensor could be used to receive IR as well :P
Mar 26 22:24:48 <Myrtti> I wana see you guys spit out uncut diamonds at my command
Mar 26 22:24:53 <Myrtti> plz?
Mar 26 22:25:01 <ma3mo> ummmm
Mar 26 22:25:03 <ma3mo> ok
Mar 26 22:25:08 <ma3mo> got it
Mar 26 22:25:12 <DocScrutinizer> Kegetys: too slow probably
Mar 26 22:25:14 <ma3mo> but java????
Mar 26 22:25:33 <ShadowJK> ma3mo, did you see java support listed on the box?
Mar 26 22:25:37 <ShadowJK> ma3mo, or in the specs?
Mar 26 22:25:59 <ma3mo> nooo i want java :)
Mar 26 22:26:20 <ma3mo> but i see umts
Mar 26 22:26:34 <DocScrutinizer> ma3mo: if you need java, go and start a project to port it. I mean that's LINUX! you know that OS don't ya?
Mar 26 22:26:35 <iBob> ma3mo : are you sure you dont want some pie ?
Mar 26 22:26:39 <ma3mo> ???? they havre to write umts signal
Mar 26 22:27:03 <iBob> lol
Mar 26 22:27:20 <ShadowJK> ma3mo, it does
Mar 26 22:27:27 <DocScrutinizer> ma3mo: please stop acting like a troll
Mar 26 22:27:28 <ma3mo> i dont want appleeeeeeeeeee i get it couse it was with qwerty etc
Mar 26 22:27:35 <iBob> DocScrutinizer : you want him to port java ?
Mar 26 22:27:46 <iBob> suggest that he reads the specs next time.
Mar 26 22:27:47 <DocScrutinizer> nah, not really
Mar 26 22:28:01 <Shapeshifter> oh noes java.
Mar 26 22:28:16 <ma3mo> be real i spent 600 euros for a brick hahahaha
Mar 26 22:28:22 <iBob> DocScrutinizer : no, its ok- he can try. :P
Mar 26 22:28:53 <iBob> ma3mo : is that in any way our problem ? :P
Mar 26 22:28:58 <Shapeshifter> ma3mo: you DID read the feature list before buying it, right?
Mar 26 22:29:04 * DocScrutinizer throws a real brick in ma3mo general direction
Mar 26 22:29:14 <ma3mo> i diiiiiid
Mar 26 22:29:30 <ma3mo> i love this device
Mar 26 22:29:39 <iBob> ok then,
Mar 26 22:29:44 <iBob> have a nice day,
Mar 26 22:29:50 <ds3> Wow... disabling 3G makes a major difference in battery life
Mar 26 22:29:50 <DocScrutinizer> *stop trolling*
Mar 26 22:29:50 <ma3mo> but it has stupid things inside
Mar 26 22:29:50 <Shapeshifter> ma3mo: and it said "java" and "irda ftp betweent devices" (whatever that exactly is supposed to be)?
Mar 26 22:29:53 <Myrtti> there's plenty of people ready to buy it off your hands if you're not satisfied with it
Mar 26 22:30:20 <DocScrutinizer> Shapeshifter: he means obex push
Mar 26 22:30:37 <ma3mo> okkk i stop to troll... can u send me a valid link for java???
Mar 26 22:30:48 <ds3> anyone remember some of the old bandwidth on demand technologies used with ISDN?
Mar 26 22:30:52 <Myrtti> ma3mo: so when are you going to stop trolling exactly?
Mar 26 22:31:07 <ma3mo> now
Mar 26 22:31:08 <Shapeshifter> ma3mo: java.sun.com?
Mar 26 22:31:11 <Myrtti> coz I didn't see you stop
Mar 26 22:31:13 <ShadowJK> ds3, 3g is bad for IM and such :)
Mar 26 22:31:37 <ma3mo> valid java app for fremantle i mean
Mar 26 22:31:40 <DocScrutinizer> ignore or kick? elections are open
Mar 26 22:31:41 <ShadowJK> There are actually some threads on the forums where people have installed java..
Mar 26 22:31:43 <ds3> ShadowJK: right...
Mar 26 22:31:54 <ShadowJK> ds3, it's all about the wake ups :)
Mar 26 22:32:21 <Myrtti> DocScrutinizer: aw, but  he's so much fun
Mar 26 22:32:26 <ds3> back then, there was ISDN routers that would use the D channesl to main a min. presence then when bandwidth demand goes up, it would add B channels
Mar 26 22:32:27 <DocScrutinizer> k
Mar 26 22:32:49 <ds3> why can't we default to 2.5G/2.0G and switch to 3G as bandwidth demand increases automatically?
Mar 26 22:32:53 <ma3mo> not with microemulator. if i m here its couse i yet read all bout that. i hate microemu is just a demo
Mar 26 22:32:53 <ShadowJK> ds3, I saw some script announced on t.m.o. that promised automatic switch between 2g and 3g :)
Mar 26 22:33:03 <ds3> oooh
Mar 26 22:33:08 * Arif_ sneezes
Mar 26 22:33:25 <ds3> 2.5G actually makes the device usable (from a battery life standpoint)
Mar 26 22:33:50 <ml-mobile> doesn't flipping between 2.XG and 3.XG kill connections?
Mar 26 22:33:51 <Arif_> 2.5G is too slow ={
Mar 26 22:34:05 <DocScrutinizer> ds3: probably mainly because there's no crystal ball to tell future bandwidth demand
Mar 26 22:34:10 <ma3mo> some help???
Mar 26 22:34:14 <Shapeshifter> my provider doesn't even do 2.5G :(
Mar 26 22:34:24 <Shapeshifter> only 2,3 and 3.5
Mar 26 22:34:36 <ds3> DocScrutinizer: that's the same issue with the ISDN stuff and it mostly worked
Mar 26 22:34:42 * ml-mobile is on 2.5G all the time, damned AT&T
Mar 26 22:34:54 <Arif_> get TMobile!
Mar 26 22:35:15 <ds3> depends on what you are used to... the N800 + tetered 2.5G phone was find for my use so...
Mar 26 22:35:17 <ShadowJK> I don't notice any difference in browsing speed between edge and 3g
Mar 26 22:35:25 <DocScrutinizer> ds3: ISDN connection establishing ois 10 times as fast as a band switch for cellular
Mar 26 22:35:39 <DocScrutinizer> maybe 100 times as fast
Mar 26 22:35:43 <Myrtti> wohey
Mar 26 22:35:43 <ds3> DocScrutinizer: oh..DOH
Mar 26 22:37:26 <DocScrutinizer> I mean, while you're at GPRS you don't keep a list of valid UMTS basestations even
Mar 26 22:38:06 <DocScrutinizer> ther's no defined roaming between GSM and UMTS afaik
Mar 26 22:38:28 <ds3> Oh... somehow I thought they were the same stations...akin to how EDGE and GPRS is
Mar 26 22:39:11 <ShadowJK> I think in dual mode it keeps track of both 2g and 3g stations though?
Mar 26 22:39:15 <DocScrutinizer> nah 3G is an entirely new BTS technology, distinct from GSM
Mar 26 22:39:24 <crashanddie_> hey guys, does MfE now actually work with GMail?
Mar 26 22:39:48 <DocScrutinizer> ShadowJK: I'm not sure but I'd guess no
Mar 26 22:39:59 <ds3> so based on what you are saying... if I am on a voice call in a 3G area and I travel into a 2.5G area, the call drops?
Mar 26 22:40:16 <crashanddie_> that's how blackberries behave
Mar 26 22:40:23 <ShadowJK> DocScrutinizer, I'd guess yes, because roaming between the two isn't up to the handset itself, and when you move from 3g area to 2g-only area your calls and data packets follow you..
Mar 26 22:40:40 <ds3> oh my
Mar 26 22:40:43 <DocScrutinizer> at least there's no such thing like a cell handover from 2G to 3G
Mar 26 22:41:07 <ShadowJK> Although when 3g was newly introduced here, most handsets acted broken when going from 3g to 2g or vice versa...
Mar 26 22:41:20 <ds3> so 3G is more like a bolting in a second V8 to get a Yugo to go faster w/removing the original Yugo engine?
Mar 26 22:41:37 <crashanddie_> Vodafone told us that we shouldn't use 3G all the time (If you have 4x4 on your SUV, you don't have it activated the whole time?)
Mar 26 22:41:47 <ShadowJK> Nokia blamed the operator, operator blamed nokia, nokia's left hand blamed nokia's right hand (which then said FU i want a divorce, and became nokia siemens networks)...
Mar 26 22:42:17 <DocScrutinizer> ShadowJK: sounds about right
Mar 26 22:42:54 <DocScrutinizer> there's no handover criterion in 2G for roaming to 3G. figure
Mar 26 22:42:55 <ShadowJK> But I've definitely maintained IP connectivity driving from 3g to 2g to 3g to 2g to 3g...
Mar 26 22:43:22 <ShadowJK> And dozens of times my E70 has borked on that
Mar 26 22:43:44 <ShadowJK> I can't remember my E75 doing that, and I've only had N900 since December so it's a bit early to tell :)
Mar 26 22:44:14 <DocScrutinizer> well it doesn't work here
Mar 26 22:44:17 <ds3> makes me wonder more and more on how well the 900 would work in really rural areas
Mar 26 22:44:33 <ShadowJK> I think I'm in "really rural" area :-)
Mar 26 22:44:50 <DocScrutinizer> ds3: better than in urban areas, as in rural ones there's no 3G at all
Mar 26 22:44:52 <ShadowJK> the 3g is more sensitive than on E75 atleast
Mar 26 22:45:04 <ds3> how does it acquire a signal? or does it just happying bounce around depending on how you hold your phone?
Mar 26 22:45:06 <ShadowJK> I get 3g signal at 10 kilometres distance from base station :-)
Mar 26 22:45:26 <ShadowJK> ds3, dont hold hand near right side, that's where antenna is
Mar 26 22:45:30 <ShadowJK> or "bottom" in portrait mode
Mar 26 22:45:52 <DocScrutinizer> actually bottom right
Mar 26 22:45:59 <DocScrutinizer> yeah
Mar 26 22:45:59 <ShadowJK> if I keep hand there the N900 instantly loses 3g and switches to the 2g station instead (which is at 400 metres distance instead of 10km for 3g)
Mar 26 22:46:26 <DocScrutinizer> same here
Mar 26 22:46:54 <DocScrutinizer> though both 2G and 3G are max 500m away
Mar 26 22:47:02 <ds3> ShadowJK: and what happens when you keep shifting your hands to cover and on cover that area?
Mar 26 22:47:16 <ShadowJK> uncover?
Mar 26 22:47:36 <ds3> like if you move your hands over and away from that area
Mar 26 22:47:38 <ShadowJK> Well the signal meter goes up and down, and at some point N900 is like "fuck this I'm going back to the stable strong 2g"
Mar 26 22:47:48 <DocScrutinizer> well here the jump to 2G means connectivity loss
Mar 26 22:48:03 <DocScrutinizer> here==my-provider
Mar 26 22:48:14 <ds3> but let's say you are doing a long TCP transfer..does it get killed or you get into TCP retry hell?
Mar 26 22:48:16 <ShadowJK> on my provider it's seamless
Mar 26 22:48:41 <ShadowJK> ds3, there's a few seconds where no data is moved, then data starts flowing again
Mar 26 22:48:58 <ShadowJK> obviously that's just with my operator, DocScrutinizer has different experience with his operator
Mar 26 22:49:21 <DocScrutinizer> that's a setup issue with carrier's network
Mar 26 22:49:47 <ShadowJK> My operator is pretty famous for disconnecting connections randomly...
Mar 26 22:50:10 <ShadowJK> My other operator has very solid performance in this regard.. I can stay connected for months nonstop
Mar 26 22:50:27 <DocScrutinizer> see, afaik there's no clear spec how to switch 2G/3G. So carriers take assumptions. Modem builders take assumptions
Mar 26 22:50:58 <DocScrutinizer> sometimes the match, sometimes :-/
Mar 26 22:54:50 <SpeedEvil> DocScrutinizer, open-source modem fw! :)
Mar 26 22:55:28 <DocScrutinizer> it's almost like in the early day of O2 (viag intercom), back when they used T-Mobile (telekom) for gap fillin. You had to relogin with a different PIN whenever you moved from one coverage area to the other
Mar 26 22:56:11 <ShadowJK> wtf :-)
Mar 26 22:56:17 <DocScrutinizer> yeah!!
Mar 26 22:56:32 <ShadowJK> My operator used to be a full virtual operator... on the website you could click a button to pick which network to use :-)
Mar 26 22:56:37 <ShadowJK> it required powercycle of phone..
Mar 26 22:56:58 <DocScrutinizer> hehe. so same shit
Mar 26 22:57:55 <ShadowJK> It was mostly an academical excercise anyway. The two biggest operators have nearly identical geographical coverage
Mar 26 22:58:02 * noobmonk3y prods frals
Mar 26 22:58:28 <ShadowJK> Though one of them has slightly more solar/wind powered byous out in the sea and thus greater coverage offshore
Mar 26 23:00:35 <noobmonk3y> ~kick Arif_
Mar 26 23:00:37 * infobot kicks Arif_
Mar 26 23:00:50 <Arif_> ~rape noobmonk3y
Mar 26 23:00:52 * infobot takes noobmonk3y behind the WallMart and makes a few grunts and screams
Mar 26 23:01:01 * noobmonk3y squeals with joy
Mar 26 23:01:50 * DocScrutinizer kicks infobot
Mar 26 23:02:00 <DocScrutinizer> no more toy for the kids
Mar 26 23:03:59 <noobmonk3y> poor old infobot
Mar 26 23:04:58 <DocScrutinizer> that's all your fault. Told ya "don't mess with infobot" :-P
Mar 26 23:10:03 <GAN900> Haha
Mar 26 23:10:07 <GAN900> Dumbass Engadget
Mar 26 23:10:18 <DocScrutinizer> ?
Mar 26 23:10:19 <GAN900> "Entelligence" my ass.
Mar 26 23:10:25 <microlith> ain't it?
Mar 26 23:10:35 <GAN900> "Mobile Multitasking is Mostly a Myth"
Mar 26 23:10:43 <microlith> except when it's not
Mar 26 23:10:44 <DocScrutinizer>  bah
Mar 26 23:11:20 <DocScrutinizer> and margerine is healthy for you
Mar 26 23:11:21 <GAN900> http://www.engadget.com/2010/03/26/entelligence-mobile-multitasking-is-mostly-a-myth/
Mar 26 23:11:54 <microlith> actually the biggest killer on my N900 is the background processes
Mar 26 23:11:59 <DocScrutinizer> and in five years we jave implanted multimedia
Mar 26 23:12:14 <DocScrutinizer> have even
Mar 26 23:12:18 <microlith> mail daemon always lets you know, via crippling cpu usage, that a message has arrived
Mar 26 23:12:20 <DrGrov> where the hell do these Engagdet people get there shit?
Mar 26 23:12:25 <GAN900> Engadget has gone so far downhill since Block left.
Mar 26 23:12:43 <GAN900> DrGrov, Apple PR department.
Mar 26 23:12:49 <DrGrov> GAN900: these god damn fuckers
Mar 26 23:13:25 <DocScrutinizer> whatever they smoke, please do mankind a favour and exinguish it from earth
Mar 26 23:17:53 <javispedro> ohnoes talk is down
Mar 26 23:18:17 <javispedro> where i'm going to troll now.
Mar 26 23:22:09 <DocScrutinizer> pfff, raw horsepower of PC makes multitasking more important there than it is on smartphones. AHUH! o.O
Mar 26 23:22:36 <DocScrutinizer> javispedro: how'd you figure?
Mar 26 23:22:41 <DocScrutinizer> ;-P
Mar 26 23:23:27 <DocScrutinizer> don't tell me you're actually using tmo
Mar 26 23:23:33 <javispedro> I read that on tmo
Mar 26 23:23:41 <DocScrutinizer> lol
Mar 26 23:31:24 <luke-jr> DocScrutinizer: smartphones are almost always PCs
Mar 26 23:32:02 <DocScrutinizer> don't tell me
Mar 26 23:32:36 <corecode_> so how do i write a python daemon for maemo?
Mar 26 23:32:41 <corecode_> where do i look?
Mar 26 23:33:16 <DocScrutinizer> luke-jr: not my words, qiote of Mr Michael Gartenberg
Mar 26 23:34:14 <DocScrutinizer> Multitasking is far more important on the personal computer -- whose windowed UI and raw horsepower make it not just a luxury but a necessity
Mar 26 23:34:40 <DocScrutinizer> BS
Mar 26 23:34:41 <luke-jr> wtf?
Mar 26 23:35:07 <luke-jr> tbh, I think desktop WMs need to support SIGSTOP on minimized windows in some cases
Mar 26 23:35:09 <luke-jr> eg, browsers
Mar 26 23:35:21 <luke-jr> even if the app itself sends the SIGSTOP
Mar 26 23:35:25 <corecode_> and then do what?
Mar 26 23:35:35 <luke-jr> SIGCONT when it's restored
Mar 26 23:35:35 <corecode_> how is the WM supposed to find out?
Mar 26 23:35:36 <javispedro> DocScrutinizer: you're reading iPhone fan club blogs now?
Mar 26 23:35:47 <corecode_> ah
Mar 26 23:35:48 <luke-jr> corecode_: the same way 'ps' finds out
Mar 26 23:35:51 <ShadowJK> luke-jr, I used to do that with firefox
Mar 26 23:35:57 <corecode_> luke-jr: running as root? :)
Mar 26 23:36:00 <DocScrutinizer> javispedro: GAN900 lured me in
Mar 26 23:36:01 <luke-jr> corecode_: no
Mar 26 23:36:07 <javispedro> evil GAN.
Mar 26 23:36:08 <luke-jr> ShadowJK: I still do, sometimes
Mar 26 23:36:10 <corecode_> luke-jr: polling the kernel?
Mar 26 23:36:12 <ShadowJK> luke-jr, but then I bought a quad core computer and 8 gigs of ram.. and now it's barely copying with it, so I'm fine..
Mar 26 23:36:27 <luke-jr> corecode_: when you restore a window, or it needs an update, check if it's suspended and SIGCONT it
Mar 26 23:36:33 <luke-jr> ShadowJK: I should do that :)
Mar 26 23:36:49 <corecode_> luke-jr: can always cont it
Mar 26 23:36:52 <corecode_> doesn't hurt
Mar 26 23:36:54 <luke-jr> sure
Mar 26 23:37:04 * GAN900 cackles
Mar 26 23:37:09 <corecode_> but what's that good for
Mar 26 23:37:09 <ShadowJK> I tried to sigstop things in maemo4, but it kinda didn't work out that well
Mar 26 23:37:15 * javispedro tends to CLOSE firefox instead of minimizing it... after a long session it gets way too heavy 
Mar 26 23:37:17 <ShadowJK> far more things than just the app itself hung :)
Mar 26 23:37:19 <luke-jr> corecode_: high-multitasking systems w/o the RAM
Mar 26 23:37:21 <corecode_> shouldn't the software just block if there is nothing happening?
Mar 26 23:37:40 <corecode_> stop doesn't swap out the process
Mar 26 23:37:48 <luke-jr> corecode_: unfortunately, I don't think toolkits are that smart
Mar 26 23:37:54 <GAN900> javispedro, Community forum has come screetching to a halt, so I don't have to read anything.
Mar 26 23:37:59 <luke-jr> STOP prevents it from running
Mar 26 23:38:28 <lcuk> ~seen jebba
Mar 26 23:38:29 <ShadowJK> corecode_, yeah well, software is so badly written
Mar 26 23:38:30 <SpeedEvil> stop will lead to swapout under memorhy pressure
Mar 26 23:38:33 <corecode_> but what is it doing?
Mar 26 23:38:33 <infobot> jebba <~jebba@Ploegsma.cwx.net> was last seen on IRC in channel #meego, 9d 1h 21m 46s ago, saying: 'i'm sick of this shit. anyway, good luck.'.
Mar 26 23:38:37 <corecode_> ShadowJK: sure
Mar 26 23:38:51 <luke-jr> corecode_: polling crap? "just checking" that it has nothing to do?
Mar 26 23:39:13 <corecode_> then it would run at 100%
Mar 26 23:39:16 <corecode_> surely not
Mar 26 23:39:22 <luke-jr> not necessarily
Mar 26 23:39:27 <luke-jr> there's usually a select loop
Mar 26 23:39:33 <javispedro> firefox? collecting garbage
Mar 26 23:39:33 <javispedro> :)
Mar 26 23:40:07 <ShadowJK> Well, xchat was checking every second or something if there was stuff to do :)
Mar 26 23:40:08 <luke-jr> I totally agree that software should be written better than that
Mar 26 23:40:17 <luke-jr> but occasionally I simply need firefox :(
Mar 26 23:40:52 <corecode_> ShadowJK: "stuff"?
Mar 26 23:41:18 <luke-jr> certain libraires *cough curl cough* make it mandatory
Mar 26 23:41:44 <corecode_> the magic of non-multithreaded and non-callback programming
Mar 26 23:41:44 <luke-jr> libcurl won't do timeouts unless you run its curl_process or whatever every N seconds
Mar 26 23:41:57 <nthx> hi.just want to say hello from n900 and test if xchat works.. c u
Mar 26 23:42:00 <corecode_> can't you run it blocking?
Mar 26 23:42:07 <javispedro> the wonders of not having a proper timing api until this decade.
Mar 26 23:42:16 <SpeedEvil> nthx, it does.
Mar 26 23:42:17 <luke-jr> I submitted a patch to fix it like a year ago, and they were all too worried about it breaking things :/
Mar 26 23:42:34 <corecode_> javispedro: yea, and no upcalls
Mar 26 23:42:47 <luke-jr> corecode_: no. even if you could, blocking prevents other things from going
Mar 26 23:43:15 <javispedro> ... now is when you start going down into the wonderful world of threading...
Mar 26 23:43:52 <luke-jr> threading+blocking makes things so much easier :)
Mar 26 23:44:03 <corecode_> luke-jr: that's when you create a thread
Mar 26 23:44:06 <luke-jr> wish Linux supported some kind of cooperative multithreading tho
Mar 26 23:44:06 <corecode_> yea
Mar 26 23:44:18 <corecode_> what's that good for?
Mar 26 23:44:30 <ptl> wow
Mar 26 23:44:35 <luke-jr> avoiding context switching
Mar 26 23:44:37 <ptl> I had a bug storm here
Mar 26 23:44:40 <luke-jr> think 10000 threads
Mar 26 23:44:41 <ptl> everything went down
Mar 26 23:44:47 <ptl> *a big storm
Mar 26 23:44:53 <corecode_> luke-jr: you still need to switch contexts
Mar 26 23:45:02 <corecode_> luke-jr: just not preemptively
Mar 26 23:45:04 <luke-jr> corecode_: not exactly
Mar 26 23:45:08 <corecode_> of course
Mar 26 23:45:16 <luke-jr> you only switch threads from the main select loop ;)
Mar 26 23:45:30 <luke-jr> and wait until it's done processing whatever and exits back
Mar 26 23:45:37 <luke-jr> then find another thread to run, or wait for data
Mar 26 23:45:40 <corecode_> and that helps how?
Mar 26 23:45:53 <corecode_> except for only running uniprocessor
Mar 26 23:45:54 <luke-jr> when you have 10000 threads, the savings becomes significant
Mar 26 23:46:01 <corecode_> which savings?
Mar 26 23:46:09 <luke-jr> not context switching needlessly
Mar 26 23:46:17 <corecode_> why would it switch needlessly
Mar 26 23:46:28 <corecode_> if there is work, it will have to run
Mar 26 23:46:32 <corecode_> if not, it won't be scheduled
Mar 26 23:46:38 <luke-jr> if thread A is running, there is no benefit to switching to thread B sooner than thread A is done
Mar 26 23:47:38 <luke-jr> why do people write async code?
Mar 26 23:48:02 <luke-jr> if there was no overhead to context switching, everything would be multithreaded blocking
Mar 26 23:48:50 <luke-jr> cooperative multithreading would be the ease of the latter with the performance of the former
Mar 26 23:48:56 <ds3> luke-jr: aren't you describing old windows with the cooperative multitasking?
Mar 26 23:49:15 <luke-jr> ds3: to an extent
Mar 26 23:49:19 <luke-jr> except that was at an OS level
Mar 26 23:49:34 <corecode_> you can do that these days as well
Mar 26 23:49:39 <corecode_> with priorities
Mar 26 23:49:48 <corecode_> and semaphores or whatever
Mar 26 23:49:51 <luke-jr> priorities only go -19 to 20 :p
Mar 26 23:49:56 <corecode_> no, that's nice levels
Mar 26 23:50:00 <ShadowJK> luke-jr, those are nice levels
Mar 26 23:50:09 <ShadowJK> real priorities go from 0-99 iirc
Mar 26 23:50:10 <luke-jr> which is effectively the same thing
Mar 26 23:50:16 <corecode_> ...
Mar 26 23:50:20 <luke-jr> 0-99 isn't much when you're talking 10000 threads
Mar 26 23:50:29 <corecode_> thanks for making it clear
Mar 26 23:50:44 <corecode_> you don't really have a lot of threading/scheduler experience, do you
Mar 26 23:50:51 <corecode_> as in "wrote it"
Mar 26 23:50:57 <luke-jr> in the kernel, no
Mar 26 23:51:06 <corecode_> the big savings you get
Mar 26 23:51:13 <corecode_> from coroutines/userland threading
Mar 26 23:51:24 <corecode_> is reduced stack space in the kernel
Mar 26 23:51:32 <corecode_> that's the big difference
Mar 26 23:52:13 <luke-jr> userland app stacks aren't in the kernel, are they? O.o
Mar 26 23:52:16 <corecode_> context switches are not really a big deal
Mar 26 23:52:39 <ptl> I think that when PR1.2 is out I'll shout at my N900. It's taking too damn long.
Mar 26 23:53:22 <luke-jr> IIRC to create a new thread, the userland needs to allocate memory for the new stack
Mar 26 23:53:39 <ShadowJK> yawm
Mar 26 23:54:50 <corecode_> no, luke-jr yes, lazily
Mar 26 23:54:58 <corecode_> that was ambiguous now
Mar 26 23:55:01 <luke-jr> wtf?
Mar 26 23:55:04 <corecode_> :)
Mar 26 23:55:14 <corecode_> userland stack is allocated lazily
Mar 26 23:55:23 <luke-jr> ...
Mar 26 23:55:27 <corecode_> kernel stack isn't
Mar 26 23:55:39 <javispedro> those stupid operating systems with 1:1 threads...
Mar 26 23:55:46 <javispedro> ;)
Mar 26 23:55:53 <luke-jr> oh, you mean the kernel keeps a local stack for syscalls?
Mar 26 23:55:58 <corecode_> but cooperative multitasking (1:n threading) also needs a userland stack
Mar 26 23:56:03 <corecode_> yes
Mar 26 23:56:03 <luke-jr> that makes sense I suppose
Mar 26 23:56:29 <luke-jr> whatever the benefit tho, I know avoiding the kernel for the multithreading seems to perform much betters
Mar 26 23:56:45 <corecode_> i doubt that
Mar 26 23:56:56 <corecode_> especially not for multi processor systems
Mar 26 23:57:12 * luke-jr ponders
Mar 26 23:57:19 <javispedro> with 10000 threads and 2 cores, I can see luke-jr being right
Mar 26 23:57:25 <corecode_> what?
Mar 26 23:57:31 <corecode_> are they all runnable?
Mar 26 23:57:34 <corecode_> or not
Mar 26 23:57:34 <luke-jr> multiprocessor systems are a slightly different scenario
Mar 26 23:57:43 <corecode_> everythign is multiprocessor
Mar 26 23:57:44 <luke-jr> corecode_: they're all runnable at some point
Mar 26 23:57:50 <corecode_> either already or in 2 years
Mar 26 23:57:55 <luke-jr> corecode_: I don't have a single multicore system
Mar 26 23:58:11 <corecode_> that sucks
Mar 26 23:58:13 <corecode_> i guess
Mar 26 23:58:17 <javispedro> corecode_: same reason you mentioned. on single processor it's faster because it's just less overhead and there's not an extra processor to schedule threads to at all, so using entirely userspace threading works.
Mar 26 23:58:22 <luke-jr> runs fine, except for RAM shortages
Mar 26 23:58:29 <corecode_> javispedro: indeed
Mar 26 23:58:38 <javispedro> corecode_: with 10000 threads and 2 processors I can see the overhead being greater than 50%
Mar 26 23:58:42 <corecode_> javispedro: but with 2 cpus, you can get a (theoretical) speedup
Mar 26 23:58:48 <javispedro> and the benefits being much smaller than 50%
Mar 26 23:58:51 <ShadowJK> there was this library that did posix threading without kernel :P
Mar 26 23:58:54 <ShadowJK> it didn't work that well
Mar 26 23:59:11 <corecode_> you mean "all unices"?
Mar 26 23:59:12 <corecode_> :D
Mar 26 23:59:17 * luke-jr ponders if he could abuse setjmp to do a proof of concept
Mar 26 23:59:24 <luke-jr> actually, I already have a semi-POC
Mar 26 23:59:33 <corecode_> luke-jr: yes, that's what all 1:n threading libs use(d)
Mar 26 23:59:41 <javispedro> userspace threads don't need any more proofs of concept :)
Mar 26 23:59:52 <corecode_> yea, that's 90ies tech
Mar 26 23:59:52 <luke-jr> corecode_: really? the spec for setjmp doesn't make that possible :P
Mar 27 00:00:01 <corecode_> wuh?
Mar 27 00:00:05 <corecode_> set/getcontext
Mar 27 00:00:12 <luke-jr> http://gitorious.org/geneticchat
Mar 27 00:00:21 <luke-jr> actually, nm
Mar 27 00:00:23 <corecode_> since you also need to preserve the fp state etc
Mar 27 00:00:27 <luke-jr> that's fork vs single-thread
Mar 27 00:00:56 <corecode_> javispedro: i can't see which overheads would build up
Mar 27 00:01:05 <corecode_> javispedro: for any reasonable workload
Mar 27 00:01:37 <javispedro> exactly. 10000 is not a reasonable workload IMHO.
Mar 27 00:01:53 <corecode_> well, let's say we have enough ram for the stacks
Mar 27 00:01:59 <javispedro> probably would be i/o boun
Mar 27 00:02:22 <corecode_> i don't know what the application at hand would be
Mar 27 00:02:26 <javispedro> (each task, ie the only reason I can think for having 10000 threads is serving 10000 i/o requests)
Mar 27 00:02:27 <corecode_> a web server?
Mar 27 00:02:54 <corecode_> 10000 chat client connections != 10000 ftp data connections
Mar 27 00:03:04 <corecode_> etc.
Mar 27 00:03:30 <corecode_> but event vs threat has been beaten to death
Mar 27 00:03:36 <corecode_> and i'm on the event side as well
Mar 27 00:03:45 <corecode_> although it makes programming more difficult
Mar 27 00:05:02 <corecode_> now i still didn't find out how the suggested way of writing a daemon for maemo is
Mar 27 00:05:18 <ptl> by 'event' you mean 'select()'?
Mar 27 00:05:23 <corecode_> no
Mar 27 00:05:29 <corecode_> generic term
Mar 27 00:05:38 <corecode_> usually something superior
Mar 27 00:05:58 <corecode_> like kqueue
Mar 27 00:06:05 <corecode_> or the linux equivalent
Mar 27 00:06:07 <ptl> ok
Mar 27 00:06:21 <corecode_> select is the worst interface
Mar 27 00:06:32 <javispedro> event queues, of which I'm a fan too
Mar 27 00:07:51 <corecode_> daemon, anyone?
Mar 27 00:08:01 <corecode_> how do you start them, how should they behave?
Mar 27 00:10:41 <DocScrutinizer> ShadowJK: xchat was updating 'meters' every second afaik
Mar 27 00:11:12 <ShadowJK> DocScrutinizer, and checking dcc timeout
Mar 27 00:11:21 <ShadowJK> even when no dccs active
Mar 27 00:11:38 <DocScrutinizer> that's realy genius
Mar 27 00:12:06 <corecode_> heh
Mar 27 00:12:28 <javispedro> well
Mar 27 00:12:37 <javispedro> a few years ago who would have though you'd have things like no_hz
Mar 27 00:12:51 <javispedro> from all the major platforms only palmos did something like that
Mar 27 00:12:52 <corecode_> no
Mar 27 00:12:59 <corecode_> right
Mar 27 00:13:05 <DocScrutinizer> well, as long as you got to wake up every second anyway, it's probably shorter code / less overhead to simply check, rather than to wind a timed alarm
Mar 27 00:13:06 <corecode_> and it makes code much easier
Mar 27 00:13:08 <corecode_> often
Mar 27 00:13:15 <ptl> I didn't even know kqeue
Mar 27 00:13:18 <ptl> *kqueue
Mar 27 00:13:24 <corecode_> ptl: that's a freebsd thing
Mar 27 00:13:30 <corecode_> don't know the linux equiv
Mar 27 00:13:40 <ptl> yes but it seems that at least two linux libraries support it
Mar 27 00:13:43 <javispedro> workqueue?
Mar 27 00:13:47 <ptl> libev and libevent0
Mar 27 00:13:59 <ptl> Alan Cox said you could use epoll instead of kqueue
Mar 27 00:14:13 <corecode_> epoll
Mar 27 00:14:14 <corecode_> yes
Mar 27 00:14:16 <corecode_> that's it
Mar 27 00:14:31 <javispedro> oops, workqueue is something entirely different
Mar 27 00:15:16 <corecode_> so, nobody developing for maemo here?
Mar 27 00:15:22 <corecode_> am i in the wrong channel?
Mar 27 00:15:24 <ptl> not yet.
Mar 27 00:15:32 <ptl> not yet --> me
Mar 27 00:15:39 <ptl> but there is also #maemo-devel
Mar 27 00:15:43 <corecode_> aha
Mar 27 00:15:46 <ptl> anyway
Mar 27 00:16:15 <ptl> I forced myself not to developer anything but simple shell scripts until I finish my master degree dissertation's fixes.
Mar 27 00:16:18 <ptl> I am too lazy. :(
Mar 27 00:16:24 <ptl> *not to develop anything
Mar 27 00:17:47 <corecode_> bleh
Mar 27 00:17:50 <corecode_> doesn't work
Mar 27 00:17:56 <corecode_> just means that you're doing less in total
Mar 27 00:17:59 <corecode_> at least for me
Mar 27 00:19:16 <ptl> ?
Mar 27 00:19:22 <SpeedEvil> corecode_, pay him so he doesn't need to continue study!
Mar 27 00:19:33 <SpeedEvil> or her
Mar 27 00:19:36 <corecode_> if i stall life because i need to finish something
Mar 27 00:19:37 <javispedro> a daemon? I'd first ask if you really, really, really need it
Mar 27 00:19:43 <javispedro> autospawned d-bus service maybe?
Mar 27 00:19:43 <ptl> oh
Mar 27 00:19:49 <ptl> I am too compulsive, corecode_
Mar 27 00:19:56 <ptl> I can't stop once I start programming anything...
Mar 27 00:20:02 <corecode_> javispedro: i want to download a new backdrop every 5/10 minutes
Mar 27 00:20:02 <ptl> been sick a few times recently because of that
Mar 27 00:20:07 <corecode_> heh
Mar 27 00:20:14 <javispedro> alarmd
Mar 27 00:20:28 <corecode_> # alarmd - daemon process that keeps track of current alarms
Mar 27 00:20:30 <corecode_> that one?
Mar 27 00:20:33 <javispedro> yep
Mar 27 00:20:37 <ptl> I've once lost the whole vacations days due to this obsessive behaviour.
Mar 27 00:20:49 <corecode_> ah, so not just BEEP BEEP alarm?
Mar 27 00:20:52 <javispedro> yep
Mar 27 00:20:53 <corecode_> but also timers
Mar 27 00:21:03 <corecode_> that wasn't clear from that description
Mar 27 00:21:04 <javispedro> for example, the rss app uses it to refresh stuff every 15 min
Mar 27 00:21:14 <ptl> I simply stood at home, programming.
Mar 27 00:21:17 <corecode_> ptl: you mean you enjoyed hacking
Mar 27 00:21:30 <ds3> what time base is the alarm system running off of?
Mar 27 00:21:32 <corecode_> how can you call that losing vacation?
Mar 27 00:21:37 <ptl> No, I mean I am kidnapped by it.
Mar 27 00:21:43 <ptl> well
Mar 27 00:21:58 <ptl> I did a good code, even upload to a brazilian site similar to sourceforge
Mar 27 00:22:17 <ptl> a few months later, I went too busy and couldn't update the code or the site on which I used it
Mar 27 00:22:31 <Lynoure> ptl: sounds like coding is not something you want to do for living?
Mar 27 00:22:39 <ptl> and then a few further months and I decided to simply close the site.
Mar 27 00:22:46 <corecode_> so?
Mar 27 00:22:56 <corecode_> natural evolution of code
Mar 27 00:23:20 <ptl> Lynoure: at that time that was what I did. Today, it's not, I am a support sysadmin at IBM. I earn the same money, but get less obsessed about stuff.
Mar 27 00:23:37 <ptl> I like programming once in a while, for fun.
Mar 27 00:23:39 <corecode_> javispedro: keeping the process around might mean less power used
Mar 27 00:23:48 <ptl> I do not make it an obligation anymore.
Mar 27 00:23:49 <corecode_> javispedro: because python doesn't have to start all the time
Mar 27 00:24:00 <Lynoure> ptl: the fine line between great motivation and unhealthy strain, I guess.
Mar 27 00:24:04 <gouverneur> re
Mar 27 00:24:06 <javispedro> ah, python...
Mar 27 00:24:16 <corecode_> javispedro: yea, i won't code that in C
Mar 27 00:24:28 <ptl> Lynoure: yeah, I think so...
Mar 27 00:25:18 <corecode_> i think there is nothing wrong with hacking for 15 hours in a 24 hour window
Mar 27 00:25:42 <Lynoure> corecode_: depends on what else one would need to achieve in that 24h window.
Mar 27 00:25:55 <corecode_> philosophically, nothing
Mar 27 00:26:03 <corecode_> does that make me a nihilist?
Mar 27 00:26:17 <Lynoure> Throw in 8h of job, and other responsibilities, and suddenly that amount of hours can become a problem.
Mar 27 00:26:32 <Lynoure> corecode_: mostly just blessed, I think
Mar 27 00:26:55 <corecode_> i used to do that as student
Mar 27 00:27:16 <crashanddie> Lynoure: 8h of work, 15h of hacking, 1h of sleep. Sounds about right.
Mar 27 00:27:26 <corecode_> what
Mar 27 00:27:29 <corecode_> 9 hours of sleep
Mar 27 00:27:57 <crashanddie> statistically speaking, people who sleep 9h every night die younger than those who sleep 6h every night
Mar 27 00:27:59 <corecode_> you must have wrong jobs
Mar 27 00:28:12 <Lynoure> crashanddie: Short term, sure :) I'd add in eating and shower, but that's just me :)
Mar 27 00:28:13 <corecode_> crashanddie: yea, i still challenge that study
Mar 27 00:28:13 <crashanddie> I figure if I only sleep 1h per night, I'll live to be 200
Mar 27 00:28:19 <ptl> lol
Mar 27 00:28:37 <firestorm> mms on the phone is there any 3rd party app for this.
Mar 27 00:28:44 <corecode_> i think they did some wrong selection towards sick/sleeps long
Mar 27 00:29:15 <ptl> firestorm: fMMS
Mar 27 00:29:25 <ptl> firestorm: it's on extras-devel.
Mar 27 00:29:27 <firestorm> ty.
Mar 27 00:29:42 <ptl> yw
Mar 27 00:29:47 <corecode_> javispedro: i can't find info on alarmd
Mar 27 00:30:08 <crashanddie> corecode_: actually, a few months ago I was doing an average of 15-16h of work on a daily basis, weekends included ;)
Mar 27 00:30:13 <javispedro> http://wiki.maemo.org/Documentation/Maemo_5_Developer_Guide/Using_Generic_Platform_Components/Alarm_Framework
Mar 27 00:30:25 <corecode_> crashanddie: for money or for fun?
Mar 27 00:30:35 <crashanddie> both
Mar 27 00:30:41 <corecode_> i was searching in the api docs :/
Mar 27 00:30:42 <DocScrutinizer> corecode_: believe me, you'll feel like that :-P
Mar 27 00:30:49 <corecode_> crashanddie: that's the best
Mar 27 00:30:55 <crashanddie> indeed it is
Mar 27 00:30:59 <DocScrutinizer> crashanddie: ^^^
Mar 27 00:31:34 <crashanddie> DocScrutinizer: ?
Mar 27 00:31:56 <DocScrutinizer> crashanddie: you'll feel like 200, even while you're just 45
Mar 27 00:32:02 <crashanddie> haha
Mar 27 00:32:34 <firestorm> pti as in the app manager?
Mar 27 00:32:38 <ptl> corecode_: there's also the API for libalarm, under here: http://maemo.org/api_refs/5.0/5.0-final/
Mar 27 00:32:48 <ptl> firestorm: yes, you have to add the extras-devel repository
Mar 27 00:33:06 <corecode_> javispedro: how do you even execute the application on startup?
Mar 27 00:33:12 <firestorm> ah n this may sound dumb how lol
Mar 27 00:33:21 <corecode_> ptl: yes, i had a look at that
Mar 27 00:33:55 <javispedro> corecode_: alarmd is persistent
Mar 27 00:34:16 <corecode_> ah, maybe that's not right then
Mar 27 00:34:28 <corecode_> i need some sort of best effort every 5 minutes
Mar 27 00:34:29 <corecode_> or 10
Mar 27 00:34:42 <corecode_> but how do i ever start an application
Mar 27 00:35:41 <javispedro> does it matter? you can set it up for a 5 minute repeating event without waking up the system
Mar 27 00:35:55 <corecode_> ok
Mar 27 00:36:04 <corecode_> still, i need to set it up
Mar 27 00:36:10 <corecode_> that means i have to run it
Mar 27 00:36:18 <corecode_> or some other application
Mar 27 00:36:31 <corecode_> i mean obviously i could do that in xterm
Mar 27 00:36:44 <javispedro> or at install time, or as part of a cpanel applet, ...
Mar 27 00:37:00 <corecode_> ok
Mar 27 00:37:07 <corecode_> now i just need to find out how to do that in python
Mar 27 00:37:08 <javispedro> cause it should be configurable, so it'll have a gui.
Mar 27 00:37:10 <corecode_> or shell
Mar 27 00:37:16 <corecode_> actually no
Mar 27 00:37:26 <corecode_> the picture i'm loading only appears every 10 minutes
Mar 27 00:37:28 <corecode_> fixed
Mar 27 00:37:30 * DocScrutinizer points at alarmed app
Mar 27 00:37:49 <corecode_> DocScrutinizer: hum?
Mar 27 00:37:50 <DocScrutinizer> it's python ;-P
Mar 27 00:37:50 <firestorm> ptl: how do i dot hat :p sorry for the dumb questions it's all new to me, or point me to some where
Mar 27 00:38:33 <DocScrutinizer> maemo's crond
Mar 27 00:38:34 <ptl> firestorm: http://wiki.maemo.org/Extras
Mar 27 00:39:06 <ptl> firestorm: see the bottom of the pages, extras-testing and extras-devel
Mar 27 00:39:09 <firestorm> ty :)
Mar 27 00:39:11 <firestorm> i see it :)
Mar 27 00:39:34 <firestorm> basically a beta... then
Mar 27 00:39:38 <ptl> yep
Mar 27 00:39:44 <firestorm> i can't belive they make a epic phone but don't do mms :S
Mar 27 00:39:58 <ptl> but there are lots and lots of applications. It's like a russian roulette: one of them will break your phone :P
Mar 27 00:40:18 <firestorm> cool we liikw russians :p
Mar 27 00:40:19 <SpeedEvil> none have broken mine so far
Mar 27 00:40:31 <corecode_> i need a python-alam package?
Mar 27 00:40:31 <corecode_> :/
Mar 27 00:40:33 <firestorm> like even
Mar 27 00:40:53 <firestorm> do you know if they plan to release a mms app for the phone in a later update as that is annoying
Mar 27 00:41:05 <nthx> ptl: or fill rootfs till u have no space left :/
Mar 27 00:41:24 <lcuk> mmm firestorm ?
Mar 27 00:41:26 <ptl> firestorm: if you search for mms on talk.maemo.org you'll see lots and lots of discussion and whining about it.
Mar 27 00:41:26 <DocScrutinizer> SpeedEvil: phototranslate (or similar) seems to break flashlight
Mar 27 00:41:43 <ptl> firestorm: basically, they did it for technical and practical (i.e. priorities) reasons.
Mar 27 00:41:52 <firestorm> yer seen it before but it's not good enough
Mar 27 00:41:53 <ptl> I mean, they didn't do it.
Mar 27 00:41:54 <corecode_> javispedro: can't be that everybody installing alarms has to write a C program?
Mar 27 00:41:58 <DocScrutinizer> even though it doesn't work on taking photos itself
Mar 27 00:42:07 <firestorm> it's like take a dummy from a baby
Mar 27 00:42:11 <ptl> nthx: yes, that's one of the things that can 'break' your phone
Mar 27 00:42:21 <DocScrutinizer> well, it looks somewhat 'commercial' ;-P
Mar 27 00:42:21 <javispedro> corecode_: DocScrutinizer was telling you that alarmed, which is an alarmd gui frontend, is written in python...
Mar 27 00:42:54 <DocScrutinizer> probably I'm on his ignorelist XP
Mar 27 00:43:09 <firestorm> lcuk: ?
Mar 27 00:43:15 <corecode_> no
Mar 27 00:43:34 <corecode_> i was more searching for a shell only solution
Mar 27 00:43:47 <corecode_> since i only want to program one repetitive alarm
Mar 27 00:44:05 <DocScrutinizer> corecode_: somebody nagged the author to implement a cmdline interface ;-P
Mar 27 00:44:10 <ptl> you mean you don't want 'cron-like' functionality, only 'at-like'.
Mar 27 00:44:13 <DocScrutinizer> to alarmed
Mar 27 00:44:22 <ptl> at now +5m...
Mar 27 00:44:27 <corecode_> ptl: cron is good enough
Mar 27 00:44:29 <ptl> *5 minutes
Mar 27 00:44:38 <corecode_> i just want to set it up
Mar 27 00:44:41 <DocScrutinizer> there's no cron on maemo
Mar 27 00:44:46 <ptl> I know
Mar 27 00:44:51 <ptl> that's why I said cron-like
Mar 27 00:44:54 <corecode_> without having to write a c program to do so
Mar 27 00:45:38 <lcuk> corecode_, setup an alarm or make a daemon autostart?
Mar 27 00:45:49 <corecode_> either
Mar 27 00:45:53 <DocScrutinizer> corecode_: would you evetually consider to give a try to alarmed?
Mar 27 00:46:15 <corecode_> DocScrutinizer: yes, my sdk doesn't seem to have these things yet
Mar 27 00:46:20 <corecode_> the extras repos
Mar 27 00:47:24 <ptl> are you using python? or do you really need shell, corecode_?
Mar 27 00:47:33 <ptl> because there are binding for python
Mar 27 00:47:33 <ptl> python-alarm - Python bindings for libalarm library
Mar 27 00:47:40 <corecode_> my program is python
Mar 27 00:48:41 <ptl> so just use python-alarm, it's on extras... won't it work the way you want?
Mar 27 00:50:17 <DocScrutinizer> corecode_: http://paste.debian.net/66069/
Mar 27 00:52:24 <ptl> I don't think it's a good idea to set a dependence on a whole app like alarmed... it's better to use libalarm
Mar 27 00:52:25 <firestorm> any one tried running a mail server on there phones being linux based?
Mar 27 00:52:51 <ptl> also, there are two apps that do the same as alarmed, fcron and some other whose nome I forgot
Mar 27 00:54:07 <ptl> firestorm: now THAT's something that would be difficult to handle. mail servers do not work well with dynamic IPs, also for the cellphone network, usually the IP's are under NAT
Mar 27 00:54:16 <corecode_> so what if i'd simply like to run a daemon on boot?  init.d or something else?
Mar 27 00:54:28 <ptl> corecode_: event.d (upstart)
Mar 27 00:54:29 <corecode_> firestorm: to receive or to send?
Mar 27 00:54:31 <ptl> /etc/event.d
Mar 27 00:54:33 <firestorm> i know pti, just thought someone would given it ago lol
Mar 27 00:54:40 <firestorm> both :p
Mar 27 00:54:44 <ptl> the same as ubuntu, but ubuntu used /etc/init directory instead of /etc/event.d
Mar 27 00:55:42 <corecode_> i see i see
Mar 27 00:55:51 <ptl> corecode_: there is the /etc/init.d directory but if you put the scripts there it doesn't work, i.e. they are not started automatically (/etc/rc2.d or whatever)
Mar 27 00:56:12 <ptl> corecode_: there is some package with sysv-something, I didn't test it, maybe it enables sysv compatibility mode for upstart.
Mar 27 00:57:19 <ptl> hmmm
Mar 27 00:57:38 <ptl> no, it's sysvinit-utils and it's nothing to do with sysv init.
Mar 27 01:00:08 <ptl> btw
Mar 27 01:00:51 <ptl> DocScrutinizer: I installed alarmed on my N900, but when I go to it, it shows only the text, and the whole background (with probably with the boxes for entering data) in black.
Mar 27 01:01:04 <ptl> DocScrutinizer: do you know anything about this?
Mar 27 01:01:39 <DocScrutinizer> never seen that
Mar 27 01:01:46 <DocScrutinizer> works like a charm here
Mar 27 01:01:56 <DocScrutinizer> both cmdline and gui
Mar 27 01:02:20 <DocScrutinizer> ask Shapeshifter
Mar 27 01:02:50 <ptl> ok
Mar 27 01:04:42 <corecode_> wow why is this such an old python version :/
Mar 27 01:05:01 <ptl> DocScrutinizer: http://patola.org/ss.png if you want to take a look though
Mar 27 01:05:25 <ptl> tried with two themes already
Mar 27 01:06:34 <DocScrutinizer> hmm, theme would have been my only guess... sorry
Mar 27 01:10:40 <DocScrutinizer> as I said, as Shapeshifter - he's the author
Mar 27 01:12:04 <ptl> Shapeshifter: are you here?
Mar 27 01:13:33 <ds3> are there binaries for SLiRP on any of the repos?
Mar 27 01:13:50 <ds3> think I've had enough of fighting pnatd
Mar 27 01:14:45 <pupnik> ~gronmayer.it
Mar 27 01:14:56 <DocScrutinizer> fighting pnatd? sounds lke fun... until you learn about the missing features
Mar 27 01:15:37 <ptl> [user@n900 user]% apt-cache search slirp
Mar 27 01:15:38 <ptl> [user@n900 user]%
Mar 27 01:15:42 <ptl> no luck
Mar 27 01:18:29 <DocScrutinizer> ptl: fyi - default theme here
Mar 27 01:20:07 <DocScrutinizer> though I'd guess alarmed.py is using standard widgets which should be tested and working with all commonly used themes, no?
Mar 27 01:20:27 <ds3> DocScrutinizer: pnatd doesn't dump debugging info and the other end is equally tightlipped
Mar 27 01:20:41 <DocScrutinizer> yo
Mar 27 01:20:46 <DocScrutinizer> lol
Mar 27 01:20:53 <ds3> cross compiling slirp... I know the remote end is fine with a raw pppd so slirp 'should' just drop in and work
Mar 27 01:21:36 <ptl> hmmm
Mar 27 01:21:40 <ptl> it uses pyside. qt
Mar 27 01:21:56 <ptl> I've seen on the changelog of PR1.2 something about QT not getting the theme colors right.
Mar 27 01:22:02 <ptl> It might be this case.
**** ENDING LOGGING AT Sat Mar 27 02:59:56 2010