**** BEGIN LOGGING AT Fri May 15 02:59:58 2009
May 15 03:33:17 <Bobbe> GAN800, hack it my friend. Hack it =)
May 15 03:33:41 <Bobbe> GAN800: is your firmware new?
May 15 03:37:01 <GAN800> 5800
May 15 03:37:12 <GAN800> Haven't figured out OTA updates
May 15 03:37:47 <johnx> ha! hacking phones, an age old tradition
May 15 03:38:09 <johnx> I remember getting into debug/monitor mode on a startac
May 15 03:42:00 <Bobbe> the hack seems to be a standard-issue HelloOX hack GAN800
May 15 03:43:46 <johnx> speaking of hacks, and GAN800 here will especially enjoy this: I now run the jailbreak on ipod touches as part of my 9-5 job
May 15 03:44:05 <Bobbe> GAN800, http://www.youtube.com/watch?v=coD7XAyv1nQ
May 15 03:44:06 <Bobbe> Have fun
May 15 03:44:12 <Bobbe> johnx, o_O?
May 15 03:44:19 <johnx> yup
May 15 03:44:42 <Bobbe> elaborate
May 15 03:44:46 <Bobbe> pls
May 15 03:45:23 <johnx> we use them as an educational tool for handing out to kids to watch videos on (in a supervised area), but we don't want kids going in and mucking around when they should be watching their video
May 15 03:45:38 <johnx> so we jailbreak so we can turn off all the functions that aren't video basically
May 15 03:45:38 <GeneralAntilles> johnx, lol
May 15 03:45:51 <Bobbe> lol
May 15 03:46:29 <johnx> I turned out 6 today in the afternoon...what a *$&%$ing annoying process
May 15 03:46:50 <Bobbe> I wonder how Steve would explain to the world how it is SO DAMN WRONG that ur jailbreaking his devices to do terrible, obnoxious things, like educating children
May 15 03:47:25 <johnx> heh, and that we need more control of the device so we can give someone else less control
May 15 03:47:31 <johnx> we are jailbreaking to remove features
May 15 03:47:46 <Bobbe> exactly.
May 15 03:48:11 <johnx> and I'd have no problem convincing them to go with Nokia stuff except that the N9xx hardware isn't out and the N8x0 hardware is obviously being phased out
May 15 03:49:01 <johnx> to say nothing of the fact that a touch isn't any cheaper than an n810 ...
May 15 03:49:05 <Bobbe> talk about phasing out. I got it here in Brazil for around R$400 (conversing and adapting to the US market it would be something like 99,99)
May 15 03:49:16 <GAN800> lol, the 5800 turns its screen on to show that retarded wallwart warning
May 15 03:49:16 <Bobbe> and even got a router in the pack
May 15 03:49:29 <GAN800> and they ship with freaking AC-8Us
May 15 03:49:37 <johnx> Bobbe, that's a nice deal
May 15 03:50:26 <Bobbe> excellent, actually. I'm all perked up tinkering with mine, tho i have no idea what I'm doing sometimes. Just sad that it doesn't get overclocked
May 15 03:50:40 <Bobbe> that and hardware acceleration drivers would make it sweet for a more years of use
May 15 03:50:47 <Bobbe> *a few more years
May 15 03:51:31 <GAN800> OC would break it
May 15 03:51:34 <Bobbe> GAN800, AC-8Us are an excellent idea. I charge my E71, old 5200 and my N800 on one plug
May 15 03:51:43 <Bobbe> OC?
May 15 03:51:56 <johnx> eh, overclocking probably wouldn't break it if done sparingly
May 15 03:52:08 <Bobbe> ah, overclocking lol
May 15 03:52:11 * Bobbe dumb today
May 15 03:52:33 <Bobbe> ppl talked about going as far as 600mhz, but I don't think it would take the heat
May 15 03:52:44 <Bobbe> even 450, 500 would be excellent.
May 15 03:53:17 <Navi> I charge my bluetooth headset and N800 on one plug
May 15 03:53:30 <Navi> I need the other one for my E70
May 15 03:53:48 <GAN800> johnx, igor explained it somewhere
May 15 03:54:06 <GAN800> johnx, basically, it aint gonna happen
May 15 03:54:15 <Bobbe> is the e70 that one that the kb opens to both sides?
May 15 03:54:26 <GAN800> You'll fry the cache or something that's tied into the CPU timer.
May 15 03:54:35 <Navi> Bobbe: yes
May 15 03:54:56 <johnx> ah, right. I remember Igor saying it wouldn't happen, but I forgot that it was a "fry your hardware" kinda thing
May 15 03:55:01 <Bobbe> that's old-school =). U still use it daily?
May 15 03:55:05 <johnx> hey Navi :D
May 15 03:56:43 <Bobbe> how bizarre, ovi store is asking for a login
May 15 03:56:44 <Navi> Hello
May 15 03:56:47 <GAN800> Ugh, apparently it's not possible to update this piece of shit OTA
May 15 03:57:16 <GAN800> Whatever feature on Nokia's website existed to send you the required settings is now gone
May 15 03:57:28 <GAN800> Jesus this is some complicates bullshit
May 15 03:57:41 <Bobbe> GAN800, what do u mean by update? like get the apn info?
May 15 03:57:54 <GAN800> OTA firmware updates
May 15 03:57:59 <Bobbe> usb?
May 15 03:58:13 <johnx> GAN800, are you liking the phone otherwise?
May 15 03:58:27 <GAN800> it bitches about creating a new server profile but apparently the required info isn't available anywhere.
May 15 03:58:49 <GAN800> johnx, I guess, but this is turning my vision red at the moment so I can't really say
May 15 03:58:59 <GAN800> Bobbe, no Windows here.
May 15 04:01:42 <Bobbe> ouch
May 15 04:05:13 <GAN800> Seriously, not one goddamn useful answer on the whole internet
May 15 04:05:34 <Navi> It's Nokia
May 15 04:05:46 <luke-jr__> N900 has a 3 gpx camera
May 15 04:05:51 <luke-jr__> 64000x48000
May 15 04:05:54 <GAN800> 5mp
May 15 04:05:56 <Navi> Nokia support sucks in the far west, remember?
May 15 04:06:03 <luke-jr__> GAN800: shh, I'm starting a rumour
May 15 04:06:21 <GAN800> Navi, internet doesn't have a thing to do with geography
May 15 04:06:34 <Navi> But the Nokia site does
May 15 04:06:44 <johnx> it has everything to do with written language though
May 15 04:06:45 <Navi> and Nokia's crappy servers
May 15 04:06:47 <ShadowJK> Except nokia has different firmware for different regions :)
May 15 04:07:00 * johnx could only find answers to his Samba problems in German
May 15 04:07:05 <GAN800> Can't this infp just come from another goddamn phone.
May 15 04:07:19 <johnx> that would be too easy
May 15 04:07:41 <johnx> people would be able to upgrade their phones whenever they wanted. we can't have that
May 15 04:08:08 <GAN800> Why did I think Symbian would be a good idea?
May 15 04:08:59 <johnx> optimism about the unknown
May 15 04:09:09 <johnx> same reason people think the n8x0 is a desktop replacement
May 15 04:09:23 <luke-jr__> johnx: because it's useless as anything else?
May 15 04:09:49 <Navi> Pfft
May 15 04:09:55 <Navi> the UI shit is all third party
May 15 04:10:03 <luke-jr__> Navi: what?
May 15 04:10:07 <johnx> luke-jr, so you think GA thought that Symbian would be a good idea because it's as useless as anything else? that doesn't even make sense...
May 15 04:10:12 <luke-jr__> N8x0's default UI is not third party
May 15 04:10:16 <Navi> The UI shell for Symbian
May 15 04:10:20 * Navi smacks luke-jr__
May 15 04:10:30 <luke-jr__> johnx: what is Symbian? :p
May 15 04:11:28 <Navi> Connecting to the Nokia servers and receiving the info is all Nokia's in-house crap
May 15 04:11:57 <GAN800> I like how the 'Download themes' link takes you to a page with a half-dozen crappy backgrounds.
May 15 04:12:04 <Navi> Though, Symbian sucks in general
May 15 04:13:30 <Bobbe> GAN800, dude, I think it would take less time for you to install windows on a box and update it than trying over the air. servers must be down or something
May 15 04:13:47 <GAN800> Bobbe, I HAVE NO server infp
May 15 04:14:07 <GAN800> it's all blank and there's apparently no way to get it
May 15 04:14:20 <GAN800> and, no, installing Windows is not an option
May 15 04:14:39 <Bobbe> er... what's infp?
May 15 04:14:45 <Bobbe> (sorry)
May 15 04:14:48 <GAN800> info
May 15 04:15:06 <Navi> Oh, just to mention
May 15 04:15:07 <Bobbe> k
May 15 04:15:18 <johnx> luke-jr, it's the thing that runs on most of the phones in the world
May 15 04:15:19 <Navi> last I heard, OTA is only supported in Europe
May 15 04:15:36 <Navi> That's why you got no shit info
May 15 04:15:37 <johnx> more irony: one of my tasks for tonight is to install windows 7 on my work macbook
May 15 04:16:22 <Navi> And if it is, it's flaky at best
May 15 04:17:42 <ShadowJK> A friend at work has a 5800, and it had a firmware upgrade check thingy that said, when he showed it to me, that he had the latest firmware
May 15 04:17:46 <ShadowJK> (europe)
May 15 04:19:24 <Navi> Nokia could care less about the land dominated by LG and Motorola
May 15 04:19:38 <johnx> motorola is dominating something?
May 15 04:19:45 <johnx> a mud puddle?
May 15 04:19:50 <Bobbe> lol
May 15 04:20:03 <Navi> Motorola makes some of the best crap phones
May 15 04:20:25 <Navi> People get Motorola phones for free with contract
May 15 04:20:31 <Navi> who cares about features, right?
May 15 04:20:40 <johnx> maybe, but right now they're not pushing many
May 15 04:20:41 <johnx> LG and Samsung are eating their lunch
May 15 04:20:52 <johnx> just saw sales figures a bit ago
May 15 04:20:57 <Navi> orly
May 15 04:21:08 <Navi> They've been pushing more commercials as well
May 15 04:21:11 <johnx> let's see if I can dig it up
May 15 04:21:19 <Bobbe> Navi, I don't think it's geographic
May 15 04:21:20 <Bobbe> http://nokiabr.blogspot.com/2009/04/atualizando-o-xm-5800-via-ota-over-air.html
May 15 04:21:23 <ShadowJK> I guess also in the US most phones are sold bundled and locked to contracts, with firmware locked down by the operator..
May 15 04:21:43 <Bobbe> that might as well be the reason
May 15 04:21:47 <Navi> It's regional
May 15 04:21:51 <Bobbe> GAN800, is your phone locked?
May 15 04:22:03 <GAN800> No. . .  .
May 15 04:22:24 <GAN800> Nobody in the US carries Nokia phones
May 15 04:22:52 <Navi> Don't know why they still have the premiums on the Nokia phones in the US either
May 15 04:22:53 <GAN800> There are a grand total of maybe 3 E/N-series phones available with contract
May 15 04:24:47 <Bobbe> wasn't at&t selling the e71 at 99?
May 15 04:24:50 <Bobbe> is it one of the three?
May 15 04:24:56 <ShadowJK> Navi, what premiums :)
May 15 04:25:20 <Navi> ShadowJK: 499 for hardware that is half a decade old
May 15 04:25:25 <ShadowJK> I don't think anyone is taking pre-orders for N97 at lower price than Nokia USA :)
May 15 04:25:46 <GAN800> 480? Nope. ;)
May 15 04:28:02 <ljp> Navi: its because in the US, the operators wont subsidize Nokia phones
May 15 04:28:07 <Bobbe> GAN800, is Comes With Music working OK? Have u tried it yet?
May 15 04:28:39 <GAN800> No
May 15 04:28:44 <GAN800> Not tried it
May 15 04:29:05 <Navi> ljp: so it has to be as expensive as subsidized unlocked phones?
May 15 04:29:06 <Bobbe> Got this in nokia support discussions: "I spoke with a collegue who works for Nokia in the EU and they have just recently changed the management structure (read between lines) at their US division. The lack of US support is just one of the many failings Nokia HQ has recognzed in their US division. "
May 15 04:29:48 <Navi> Yes, the lack of support is a pretty obvious thing
May 15 04:37:30 <Bobbe> well, I'm off guys. GAN800, if you have access to a linux box, ppl have been able to use PC Suite to update their firmware over vmware. But if windows is not an option, I suppose a windows vm is not either
May 15 05:01:53 <b1ackdeath> penguinbait's  install-tools-N810e.deb dosnt work on 43-7 right?
May 15 05:27:36 <jaem> hey folks
May 15 05:29:16 <jaem> to clarify, does anyone know if Diga @Web is only free for the beta period?
May 15 06:29:13 <thux> morning
May 15 06:49:31 * Stskeeps yawns
May 15 06:51:03 * Myrtti sighs
May 15 06:51:19 * Stskeeps passes Myrtti a hello kitty coffee mug
May 15 06:51:26 <Myrtti> thankies love
May 15 06:52:46 <Stskeeps> things i never thought i'd do in my life: check up on chinese forums with translate.google.com in the morning
May 15 06:58:43 <RST38h> Sts: We all feel your pain.
May 15 07:13:19 <Macer> have to love a nutjob setting people on fire
May 15 07:15:24 <Stskeeps> mm?
May 15 07:24:25 <rkirti>  /whois Myrtti
May 15 07:24:33 <rkirti> oops...sorry for that
May 15 07:24:41 <Stskeeps> morning rkirti
May 15 07:24:53 <rkirti> morning Stskeeps
May 15 07:25:19 <Stskeeps> when is it the gsoc students are supposed to start their work btw?
May 15 07:26:11 <rkirti> Stskeeps: 23rd May. though its only for the namesake. All the people I know are racing ahead with their projects :)
May 15 07:26:16 <Stskeeps> hehe
May 15 07:28:29 <Corsac> do we know if freemantle will have some PDA capabilities (like PIM, sync etc.) which diablo doesn't really have?
May 15 07:28:57 <Stskeeps> they have a calendar backend and sync it seems
May 15 07:29:48 <Myrtti> last day of vacation *sigh*
May 15 07:31:35 <aquatix> Macer: aww
May 15 07:31:40 <aquatix> erm
May 15 07:31:43 <aquatix> tab error
May 15 07:31:45 <aquatix> Myrtti: aww
May 15 07:31:51 <aquatix> enjoy it :)
May 15 07:31:56 <biiter> hi, i wonder, how is internet table "internet" when default web browser and fennec are slow as hell? .. and gui in general... is there any alternative with less overload?
May 15 07:32:05 <Stskeeps> biiter: you will want to try out Tear
May 15 07:32:07 <Stskeeps> it's excellent
May 15 07:32:11 <Stskeeps> based on webkit engine
May 15 07:32:49 <radic_> does an replacement for nolo exist?
May 15 07:32:51 <biiter> ok.. ill try :)
May 15 07:33:36 <Stskeeps> it really has caused me to use my tablet much more for browsing
May 15 07:35:01 <Jaffa> Stskeeps: is it in a single repo yet? Last time I looked there was a repo from qole and one from Bundyo
May 15 07:35:08 <biiter> Stskeeps: and any GTK replacement?
May 15 07:35:08 <Stskeeps> not sure
May 15 07:35:25 <Stskeeps> biiter: as in newer GTK? that's a worse question :)
May 15 07:35:28 <Macer> haha
May 15 07:35:40 <Jaffa> biiter: There are Qt apps, SDL apps, EFL apps and liqbase
May 15 07:49:11 <biiter> it works like this ? (gtl / qt)<-hildon<-app ??
May 15 07:49:22 <biiter> gtk
May 15 07:50:25 <Summeli> how fast is the qt?
May 15 07:55:05 <biiter> oh yeah... tear is cool :] thx
May 15 07:58:02 <Jaffa> biiter: Hildon is a framework on top of Gtk
May 15 07:58:38 <ShadowJK> the browser is so fast browsing mailing list archives. wish all web pages were as light :-)
May 15 07:59:43 <biiter> Jaffa: and can it be on top of qt?
May 15 08:00:18 <Jaffa> biiter: No. But the Qt port has its own equivalent (and may even use the same names - I dunno, not looked at the API)
May 15 08:01:13 <Summeli> hmm. maybe I could benchmark the qt blit some day :P
May 15 08:01:18 <biiter> ok.. btw.. why maemo uses gtk when nokia owns qt
May 15 08:01:28 <Stskeeps> nokia didnt own qt when maemo started :)
May 15 08:01:47 <biiter> dammit.. ok :]
May 15 08:01:48 <biiter> thx
May 15 08:03:35 <biiter> oh, and that 2d/3d co-processor on the board is used natively or it needs to be requested by application?
May 15 08:07:09 <qgil> I'm intrigued about the reasons why a http://maemo.org/news/planet-maemo/ reader would click thumbs down to the entries about the Maemo Summit and the Danish Weekend...
May 15 08:07:57 <lbt> the qt hildon stuff aims to be comparable to the gtk and use the same dbus api
May 15 08:08:11 <lbt> they can't go and they're fed up?
May 15 08:08:16 <suihkulokki> qgil: some enviromentalist dissaproving excessive travel? :P
May 15 08:08:21 <Jaffa> :)
May 15 08:08:40 <qgil> suihkulokki: that could be
May 15 08:10:07 <qgil> lbt: what is the question exactly?
May 15 08:10:51 <lbt> qgil: that was an answer to bitter/Jaffa :)
May 15 08:12:07 <lbt> (essentially does Qt have Hildon?)
May 15 08:12:15 <Stskeeps> it emulates hildon
May 15 08:12:51 <RST38h> qgil: Track him done and wrestle the answer from him!
May 15 08:13:09 <biiter> this could be replacement http://qt4.garage.maemo.org/
May 15 08:16:04 <biiter> http://qtablet.laginen.net/doku.php
May 15 08:16:09 <biiter> got to sit on this
May 15 08:18:29 <lbt> biiter: yes, I'm working on it too
May 15 08:18:38 <lbt> right now I'm doing scroll-bounce
May 15 08:18:49 <lbt> fingerscrolling works
May 15 08:19:15 <biiter> fu*k.. im late on exam... afk
May 15 08:19:27 <Stskeeps> tablets > exams
May 15 08:19:27 <Stskeeps> ;)
May 15 08:22:07 <thux> is there tablet degree like rhce?
May 15 08:22:37 <thux> hands on test tuning tablet
May 15 08:22:55 <ShadowJK> biiter: 3d accel not used by anything. Some 2d accel used by mplayer, liqbase, media player...
May 15 08:23:17 <ShadowJK> either through direct hw access (mplayer) or through Xvideo
May 15 08:23:35 <ShadowJK> media player also uses the dsp for mp3 :)
May 15 08:27:06 <RST38h> Current Xv only exposes YUV buffers
May 15 08:27:52 <ShadowJK> yeah well, it's designed for video after all :)
May 15 08:58:25 <Jaffa> X-Fade: ping
May 15 09:19:25 <timeless_mbp> crashanddie: ping
May 15 09:20:06 <Stskeeps> morn timeless_mbp
May 15 09:20:24 <timeless_mbp> 2am..., how's life on the other side of the world?
May 15 09:20:36 <Stskeeps> sunny for once
May 15 09:23:54 <murrayc> Where is the modest package for the Maemo 5 Beta? In extras? It doesn''t seem to be in extras-devel
May 15 09:25:09 <Stskeeps> http://repository.maemo.org/pool/fremantle/free/m/modest/ ? :P
May 15 09:25:34 <murrayc> Yeah, I just saw that, which suggests that it's in the main repository.
May 15 09:25:44 <murrayc> But "apt-get install modest" can't find it.
May 15 09:26:01 <Corsac> hmhm, did someone already tried to dist-upgrade a tablet to fremantle using the sdk repository?
May 15 09:26:08 <Corsac> that would be fun
May 15 09:26:09 <Stskeeps> Corsac: it's armv7 targeted
May 15 09:26:10 <Stskeeps> so
May 15 09:26:17 <murrayc> Corsac: That will not work.
May 15 09:26:17 <Corsac> well, for some value of fun, maybe
May 15 09:26:22 <Corsac> ha yes, forgot that
May 15 09:26:58 <Corsac> hmhm, though the ABI is still armel, so the binaries should at least run?
May 15 09:27:23 <Stskeeps> armv7 on a armv6 machine..
May 15 09:28:26 <lbt> for some definition of run that includes stationary
May 15 09:29:09 <RST38h> Sts: May run =)
May 15 09:29:16 <Stskeeps> yes, may
May 15 09:29:21 <Stskeeps> but also be crashable :P
May 15 09:29:37 <RST38h> well crashable =)
May 15 09:30:21 <Corsac> well, it the abi is changed, whouldn't the arch be changed too?
May 15 09:30:38 <Stskeeps> abi isn't changed, arch is
May 15 09:30:40 <Stskeeps> it's still armel
May 15 09:30:52 <Stskeeps> arch as in armv5, armv6, armv7 .. etc
May 15 09:30:55 <Corsac> I mean, one is supposed to be able to run _i386.deb stuff on an i386 nowadays (provided you can run the kernel)
May 15 09:30:59 <Corsac> ok
May 15 09:31:29 <RST38h> ABI and CPU architecture versions are different things
May 15 09:31:43 <RST38h> armv7 vs armv5 is like Pentium vs i386
May 15 09:32:03 <RST38h> The code will mostly run until it hits a Pentium specific instruction where it will fail.
May 15 09:37:34 <Corsac> http://www.engadgetmobile.com/2009/05/15/nokia-launching-high-end-handset-with-capacitive-touchscreen-i/ could that be for the n9xx?
May 15 09:38:14 <RST38h> No.
May 15 09:38:28 <RST38h> The keyword is "handset". It will be an S60e5 device.
May 15 09:38:59 <RST38h> An update to 5800 (supposedly named 5900) has been rumored, so that may be it
May 15 09:39:15 <Jaffa> Corsac: the future tenses on "will adopt" also seem odd when Nokians have pre-production hardware already
May 15 09:42:33 <murrayc> Are there really no modest packages that I can apt-get install for the Beta SDK?
May 15 09:42:45 <RST38h> apparently
May 15 09:42:56 <murrayc> That's a strange definition of "released".
May 15 09:43:08 <RST38h> gimme a moment
May 15 09:43:27 <RST38h> http://repository.maemo.org/pool/fremantle/free/m/modest/
May 15 09:43:36 <RST38h> murrayc: is this it?
May 15 09:43:43 <murrayc> But "apt-get install modest" can't find it.
May 15 09:43:57 <RST38h> have you got this repository in your config file?
May 15 09:44:07 <murrayc> And I can't figure out any extra sources.list line to make that work.
May 15 09:44:17 <RST38h> just wget + dpkg
May 15 09:44:30 <murrayc> The Beta SDK has this by default:
May 15 09:44:30 <murrayc> deb http://repository.maemo.org/ fremantle/sdk free non-free
May 15 09:44:30 <murrayc> deb-src http://repository.maemo.org/ fremantle/sdk free
May 15 09:44:30 <murrayc> deb http://repository.maemo.org/ fremantle/tools free non-free
May 15 09:44:30 <murrayc> deb-src http://repository.maemo.org/ fremantle/tools free
May 15 09:44:41 <RST38h> Ah I see your problem
May 15 09:44:47 <murrayc> I can never figure out those lines.
May 15 09:44:55 <RST38h> Add deb http://repository.maemo.org/ fremantle free non-free
May 15 09:45:02 <murrayc> I tried that.
May 15 09:45:06 <RST38h> And the same for deb-src
May 15 09:45:23 <murrayc> W: Couldn't stat source package list http://repository.maemo.org fremantle/free Packages (/var/lib/apt/lists/repository.maemo.org_dists_fremantle_free_binary-i386_Packages) - stat (2 No such file or directory)
May 15 09:45:27 <RST38h> then do apt-get update
May 15 09:45:44 <murrayc> So I guess the Packages file is missing.
May 15 09:45:50 <RST38h> a moment
May 15 09:45:52 <murrayc> So it's maybe an orphaned .deb.
May 15 09:47:49 <RST38h> Ok, I have looked at it
May 15 09:47:59 <RST38h> It looks like the Packages file has not been set up yet
May 15 09:48:26 <murrayc> Yeah. I'll complain in a comment on the beta SDK announcement. So much of the stuff mentioned there just doesn't exist.
May 15 09:48:39 <murrayc> Thanks for verifying that I'm not completely stupid.
May 15 09:48:57 <RST38h> Just wget+dpkg for now
May 15 09:49:23 <murrayc> I don't want to fuck up my target by doing that kind of thing.
May 15 09:49:35 <murrayc> And it has dependencis that I'd have to mess around with too.
May 15 09:49:56 <murrayc> (I developed some of the code for the Diablo version)
May 15 09:50:13 <murrayc> So, yes, I could install it from git, but I am annoyed that it's not easy for regular people.
May 15 09:50:48 <RST38h> You will not fuck up your target
May 15 09:50:58 <murrayc> Really, I can do that.
May 15 09:51:00 <Myrtti> lock in target
May 15 09:51:01 <murrayc> I'm that good.
May 15 09:51:05 <RST38h> Do not install from git, install from that pool directory
May 15 09:51:24 <RST38h> Just do it with dpkg - and it will work exactly the same as if you did it with apt-get
May 15 09:52:25 <RST38h> BTW, if you look at http://repository.maemo.org/dists/ you will see that all the maemo5 directories disallow browsing. My guess is that they do contain Packages files etc but these have references to internal stuff Nokia does not want us to see
May 15 10:30:45 <Corsac> Jaffa: so no way n9xx may have capacitive screen?
May 15 10:31:14 <Stskeeps> i doubt it really
May 15 10:31:24 <Stskeeps> especially with the two-finger thing
May 15 10:37:08 <Jaffa> The part number found in the kernel also is of a resistive screen.
May 15 10:37:33 <Corsac> ok
May 15 10:41:25 <pacific> yo pples i need some help
May 15 10:41:54 <Stskeeps> better to ask the question immediately instead of asking for help
May 15 10:41:54 <Stskeeps> :P
May 15 10:42:13 <pacific> my torrents on niiiiiiiiiiit dont work
May 15 10:42:31 <pacific> im usin transmission
May 15 10:43:09 <mikkov_> that's not a question :)
May 15 10:44:29 <mikkov_> X-Fade: ping
May 15 10:45:10 <Jaffa> mikkov_: I've had an X-Fade ping outstanding since 09:57, so it might time out ;-)
May 15 10:51:44 <jeremiah> Stskeeps: Did you see I signed up for some Mer hacking while in CPH?
May 15 10:51:47 <jeremiah> :P
May 15 10:52:28 <Stskeeps> yeah - sounds great :)
May 15 10:52:36 <jeremiah> I look forward to it.
May 15 10:54:22 <Stskeeps> we'll be a nice cluster, me, you, johnx, rm_you, lbt, claudius
May 15 10:54:58 <jeremiah> Yeah, starting to get critical mass.
May 15 10:56:11 <Stskeeps> yeah.. a lot of technical preparations to be done before that - i'm glad we got OBS up and going
May 15 10:56:15 <Stskeeps> will make things a hell lot easier
May 15 10:56:27 * Jaffa sad to be missing it :(
May 15 10:56:47 <Stskeeps> better come for the summit then :P
May 15 10:57:01 <RST38h> he won't have a choice being the debmaster :)
May 15 10:58:39 * Stskeeps glances at his presentation notes
May 15 11:00:50 <jeremiah> I am going to start replacing *@debian.org email addresses in packages since I doubt debian devs are uploading to maemo
May 15 11:01:26 <Stskeeps> some might :P
May 15 11:01:28 <jeremiah> qgil: Can we also make sure that Nokia packagers also follow the Maemo packaging policy and place the _real_ email address of the packager.
May 15 11:01:51 <jeremiah> Stskeeps: Yeah, some might, and I know a couple - but this package, OpenCDK is group maintained.
May 15 11:01:58 <jeremiah> And the control file is identical.
May 15 11:02:02 <jeremiah> with upstream
May 15 11:02:12 <jeremiah> So, I am a bit suspicious here.
May 15 11:02:15 <Stskeeps> *nod*
May 15 11:02:28 <biiter_> ShadowJK: thx
May 15 11:02:34 <Jaffa> jeremiah: Isn't that what Uploader is for?
May 15 11:02:55 <Jaffa> The package *is* maintained by *@debian.org, but someone has uploaded it to another repo with no changes.
May 15 11:03:21 <Stskeeps> i wonder how many extras packages there are that are no-changes
May 15 11:03:44 <jeremiah> Jaffa: "Uploader" is actually a debian package field
May 15 11:03:58 <Jaffa> Stskeeps: Mostly the command line stuff - certainly the exception rather than the rule
May 15 11:04:13 <jeremiah> It denotes who uploaded the package into a VCS - not to debian (at least in the case of debian-perl)
May 15 11:04:40 <Jaffa> jeremiah: Doesn't Ubuntu use it for who uploaded it to the repo (I've never seen a documented definition of its use before)
May 15 11:05:40 * Jaffa is coming at this from a position of little knowledge; not one of trying to convince you I'm right, btw ;-)
May 15 11:05:54 <jeremiah> I'm not sure what Ubuntu does actually - I should find out, but debian uses it to give credit to whomever did the packaging, not who uploaded it. :/
May 15 11:05:55 <Corsac> Uploaders: are the people authorized to be in the changelog line, more or less. If you aren't in Uploaders: and still do that, that's an NMU
May 15 11:06:12 <Jaffa> Corsac: NMU?
May 15 11:06:16 <Corsac> non-maintainer-upload
May 15 11:06:28 <Jaffa> Right.
May 15 11:06:35 <jeremiah> An NMU is when some else than the package maintainer uploads a package
May 15 11:06:39 <Corsac> yes
May 15 11:06:53 <jeremiah> So for bug fixes or when the original maintainer is MIA
May 15 11:07:03 <Jaffa> Where "upload" here is "put in repo"?
May 15 11:07:06 <jeremiah> It is considered bad form in debian to do NMU
May 15 11:07:10 <Corsac> depends
May 15 11:07:15 <Corsac> it's ok to do nmu fixing bugs
May 15 11:07:16 <jeremiah> Jaffa: Unfortunately, it is not really clear.
May 15 11:07:17 <Corsac> rc ones
May 15 11:07:23 <jeremiah> It is a mess, actually.
May 15 11:07:27 <Jaffa> jeremiah: Right.
May 15 11:07:32 <jeremiah> :/
May 15 11:07:36 <Corsac> it's perfectly clear, upload means upload to debian ftp archive
May 15 11:07:40 <_berto_> jeremiah: ubuntu has Maintainer: and Original-Maintainer: for packages moved from debian
May 15 11:07:40 <Jaffa> Can we not slavishly copy it then? :)
May 15 11:07:44 <jeremiah> Corsac: Not at all.
May 15 11:07:44 <Corsac> :>
May 15 11:07:56 <Corsac> mhmh?
May 15 11:08:06 <jeremiah> If you look at my packages in debian, you will see I am in the uploader field
May 15 11:08:11 <Jaffa> Cos I'm guessing Ubuntu has faced many of the packaging issues from Debian upstream that Maemo also does
May 15 11:08:19 <jeremiah> But I am not a DD and cannot upload to debian
May 15 11:08:33 <jeremiah> So I "upload" to the svn repo, and a DD uploads to debian
May 15 11:08:34 <_berto_> jeremiah: what's the uploader field ?
May 15 11:08:45 <Corsac> well, if you are a DM and DM_Allowed:yes is present, you could :)
May 15 11:09:01 <Corsac> but basically you can be the one doing the change and have it sponsored
May 15 11:09:01 <jeremiah> _berto_: It is a field that denotes that you brought software into debian's repos, but not into their mirrors.
May 15 11:09:01 <_berto_> jeremiah: there's mentors.debian.net for maintainers to put their packages so DD can upload them
May 15 11:09:10 <Corsac> jeremiah: btw, DD I am
May 15 11:09:29 <jeremiah> Corsac: What is important to note is, uploaders often are not DDs so they don't really "upload" to debian
May 15 11:09:32 <Jaffa> "upload" as equivalent to "put in VCS" seems very counter-intuitive to me (put debian/control in VCS or the source or the whole source package?)
May 15 11:09:39 <jeremiah> Jaffa: Exactlyl.
May 15 11:09:58 <jeremiah> Or exactly even.
May 15 11:10:18 <Corsac> basically Uploaders are people who have the right to do changes to the packages. Maintainer was more adapted but already taken
May 15 11:10:26 <Corsac> Maintainer: wasn't adapted to team-maintainance
May 15 11:10:34 <_berto_> jeremiah: i've never seen that field, where is it ?
May 15 11:10:45 <jeremiah> _berto_: In the control file
May 15 11:10:54 <jeremiah> not every package has it
May 15 11:11:08 <jeremiah> It is usually used when you have group maintenance of packages
May 15 11:11:16 <jeremiah> Like in the debian python or debian perl group
May 15 11:11:20 <_berto_> ah okay
May 15 11:11:53 <jeremiah> So that way, the person who made the package gets credit for doing the work as "uploader" and the Maintainer field, which receives bug reports, etc. gets kept for the group
May 15 11:11:53 <Corsac> though you could have an Uploader: field for package not maintained in vcs, but that would make a quite hard workflow
May 15 11:12:55 <Corsac> basically, the person listed in the changelog file for that entry is the people taking responsibilty for the upload, even if he can't upload directly and have to ask someone else
May 15 11:13:05 <jeremiah> Jaffa: Maybe we can get rid of it - the policy says nothing about it.
May 15 11:13:48 <Jaffa> I think my point is that given the confusion around here, there's scope for jeremiah to start discussions on maemo-developers for a sensible usage for Maemo, that some packages (although mostly CLI stuff) will be ported with no change from Debian; and that the person who maintains a package in Maemo is - very typically - also the person who does the development (few people taking upstream tarballs and doing a *lot* of packaging effort and not coding effort)
May 15 11:14:17 <jeremiah> What I would really, really like though is for people to put their _real_ email address in the maintainer field so we can contact them if there is a problem, like now with diablo -> fremantle.
May 15 11:14:40 <jeremiah> (Jaffa)++
May 15 11:14:42 <Corsac> definitely
May 15 11:14:45 * Jaffa would say "*a* field" ;-)
May 15 11:15:03 <Jaffa> jeremiah: For the current problem, can you get hold of the original auto-builder/upload logs?
May 15 11:15:14 <jeremiah> Jaffa: Yeah, I can.
May 15 11:15:21 <jeremiah> Just a lot of grepping. :)
May 15 11:15:28 <Jaffa> Those tie up with maemo.org accounts andbingo.
May 15 11:15:31 <Jaffa> Cool
May 15 11:15:31 <Corsac> http://www.debian.org/doc/debian-policy/ch-controlfields.html#s-f-Maintainer might help
May 15 11:15:57 <jeremiah> Corsac: Good link.
May 15 11:16:38 <jeremiah> The maemo policy, which is a pdf so I won't link, says:
May 15 11:17:04 <jeremiah> "If an upstream package is modified for maemo, the Maintainer field MUST be change from the original."
May 15 11:17:19 <jeremiah> And the typo was min.
May 15 11:17:21 <jeremiah> e
May 15 11:18:22 <Jaffa> So, yesterday when I uploaded netkit-telnet-0.17_36 (or somesuch) straight from Debian, that was fine. But when I changed the Section and added XB-Maemo-Icon-26, I changed the Maintainer and the version.
May 15 11:18:37 <jeremiah> Jaffa: Perfect.
May 15 11:18:44 <jeremiah> I wish everyone would do that. :)
May 15 11:19:10 <jeremiah> Oooops. Looks like I am spamming the fremantle builder. =)
May 15 11:19:24 <Jaffa> But if there was a problem with the former, the Maintainer would still be @debian.org - even though that's valid.
May 15 11:19:41 <Jaffa> But then, saying, "add an Uploader" (or similar) changes the package; so the Maintainer has to change.
May 15 11:19:42 <jeremiah> Jaffa: Oh yeah, that has to get changed actually.
May 15 11:19:45 <Jaffa> Damn.
May 15 11:19:46 <lardman> morning all
May 15 11:19:54 <jeremiah> Hell lardster.
May 15 11:20:07 <Jaffa> Two more links: https://wiki.edubuntu.org/Debian/ForUbuntuDevelopers#Required%20packaging%20changes and https://lists.ubuntu.com/archives/ubuntu-devel/2008-April/025325.html
May 15 11:20:14 <lardman> hey jeremiah
May 15 11:20:58 <jeremiah> One needs to replace the Maintainer with the Maemo maintainer - and you can chose to create a XSBC-Original-Maintainer field for the debian field
May 15 11:22:23 <mikkov_> jeremiah: I hope that you wouldn't upload any of my user/* packages to fremantle
May 15 11:23:05 <jeremiah> mikkov_: Can you tell me what they are - I thought I was supposed to upload as much as possible
May 15 11:24:55 <mikkov_> jeremiah: well I can't remember them all :) but for instance openlierox which is in build queue.
May 15 11:25:45 <mikkov_> jeremiah: actuallu most of them won't build with fremantle, so actually it shouldn't be a problem
May 15 11:25:51 <Stskeeps> jeremiah: wasn't it more along lines of testing if it works and get back to developers?
May 15 11:26:35 <jeremiah> Stskeeps: Yeah, it was kinda like that. But setting up the build system somewhere else is non-trivial.
May 15 11:28:04 <Jaffa> jeremiah: That sounds like a good plan (XSBC-Original-Maintainer)
May 15 11:28:36 * Jaffa needs to make sure mud doesn't allow creating naughty packages, but makes it easy to take upstream and republish.
May 15 11:28:48 <Jaffa> Or tarballs, or whatever.
May 15 11:28:55 * RST38h requests that the number of necessary fields be minimized and, more importantly, the number of necessary package config files be minimized
May 15 11:29:20 <RST38h> So that I do not need to jump all over control, changes, files, etc every time I do a release
May 15 11:29:54 <jeremiah> RST38h: Absolutely. Note that XSBC-* is optional.
May 15 11:30:08 <mikkov_> Jaffa: according to policy netkit-telnet should have had version
May 15 11:30:27 <jeremiah> But not changing the Maintainer will be a QA failure, a la lintian error.
May 15 11:30:31 <mikkov_> Jaffa: 0.17-36maemo2
May 15 11:30:31 <Jaffa> mikkov_: "had version"?
May 15 11:30:37 <RST38h> jeremiah: Actually, I would rather add these fields to control than add them to other files
May 15 11:30:46 <Jaffa> mikkov_: True - I was rushing.
May 15 11:31:02 <jeremiah> RST38h: That is where they are.
May 15 11:31:10 <RST38h> jeremiah: like changes (it also needs email address)
May 15 11:31:11 * Jaffa is planning on making mud do that for me automatically - but hasn't yet.
May 15 11:39:40 <jeremiah> RST38h: Changing the format of the changelog and other debian files leads to cataostrophic problems in apt and the apt toolchain.
May 15 11:42:54 <Corsac> yeah, that's not really something you want to do :)
May 15 11:55:10 <timeless_mbp> hi jeremiah
May 15 11:59:01 <lbt> hmm, very interesting backlog... thanks jeremiah, Corsac and the rest...  I need to address lots of this in Mer too.
May 15 12:02:08 <Corsac> no problem
May 15 12:06:03 <lbt> and you're on my goto list for packaging advice now ;)
May 15 12:17:52 <qgil> jeremiah: http://blog.jeremiahfoster.com/?p=113 is good!
May 15 12:18:26 <qgil> now... what about putting that content in http://wiki.maemo.org in a please where your target audience will probably look at the future.  :)
May 15 12:18:33 <qgil> in a place
May 15 12:25:26 <rkirti> sorry for an extremely noobish question, but what exactly is a FIASCO image ?
May 15 12:26:01 <StsN800> rkirti, nokia's firmware format for tablets
May 15 12:27:45 <StsN800> hey wazd
May 15 12:28:16 <rkirti> StsN800: ok
May 15 12:34:57 <lcuk> lbt, \o i gather you have seen the qt-kinetic posting on itt, since you have qt working nicely, wanna get the examples built and running so we can see and feel and give proper feedback to the guys? http://talk.maemo.org/showthread.php?t=28993
May 15 12:36:23 * lcuk likes the look of the flickr viewer
May 15 12:49:11 <fiferboy> lbt: Hey, how's it going?
May 15 12:49:49 * RST38h yawns then moos
May 15 12:57:50 <jeremiah> qgil: Yeah - I will do that. :)
May 15 12:58:23 <AStorm> hello
May 15 12:58:37 <AStorm> could someone nice build us python2.5-pyprocessing package?
May 15 12:59:31 <AStorm> http://pyprocessing.berlios.de/ - and probably has an Ubuntu and Debian packages
May 15 13:04:19 <AStorm> hello?
May 15 13:04:43 <AStorm> or has everyone switched to Mer already? :>
May 15 13:05:31 <RST38h> No, we are all running WinCE
May 15 13:05:51 <RST38h> With nice C# scripting builtin!
May 15 13:06:07 <AStorm> http://packages.debian.org/lenny/python-processing
May 15 13:06:22 <AStorm> so there is a package
May 15 13:06:32 <AStorm> now I need someone to submit it to a buildbot
May 15 13:07:28 <lbt> AStorm: Mer buildbot?
May 15 13:07:34 <AStorm> oh, it requires ctypes... that has to be done too
May 15 13:07:36 <AStorm> no, Maemo one
May 15 13:07:45 <AStorm> Mer has it already (Ubuntu has that)
May 15 13:07:51 <AStorm> and Mer has Python 2.6 too
May 15 13:07:53 <lbt> 'k
May 15 13:08:29 <lbt> FYI Ubuntu seems to have python problems ... I hear lots of gnashing of teeth
May 15 13:08:38 <lbt> but you may know that already :)
May 15 13:08:42 <AStorm> not enough to cause me grief
May 15 13:08:45 <AStorm> :)
May 15 13:08:56 <AStorm> still, I need these in Maemo too :)
May 15 13:09:03 <lbt> lcuk: why not
May 15 13:09:06 <lbt> fiferboy: hi
May 15 13:09:14 <lcuk> excellent :)
May 15 13:09:25 <lbt> fiferboy: it's a pain in the finger!
May 15 13:09:41 <fiferboy> lbt: How is it working?
May 15 13:10:06 <lbt> meh... mainly OK but some sign/maths issues
May 15 13:10:56 <fiferboy> Anything I could look in on?
May 15 13:11:18 <lbt> lcuk: they appear to be commercial
May 15 13:11:29 <lbt> fiferboy: I can send you the cpp
May 15 13:11:41 <AStorm> http://packages.debian.org/squeeze/python-multiprocessing - this would be even better
May 15 13:11:46 <fiferboy> Sure
May 15 13:11:49 <AStorm> (same package, but a backport)
May 15 13:12:44 <lcuk> lbt? ocmmercial in what way
May 15 13:12:49 <lcuk> as in you cant use them on maemo>?
May 15 13:13:07 <AStorm> ;-( no one to help my case?
May 15 13:13:10 <lbt> I found a 'buy now' link for @web and stopped looking
May 15 13:13:27 <lcuk> not those :$ sorry the first link gives a few example things
May 15 13:13:30 <lcuk> like a flickr viewer
May 15 13:13:31 <lbt> AStorm: Maemo is dead, long live Mer....
May 15 13:13:34 <lcuk> and a recipe list
May 15 13:13:35 * lbt ducks
May 15 13:13:39 <AStorm> lbt, yeah yeah
May 15 13:13:49 <AStorm> but I have to support both!
May 15 13:13:55 <lcuk> jeremiah, would you kindly give AStorm some advice about his package :)
May 15 13:13:59 <AStorm> (and stationary machines too)
May 15 13:14:00 <lcuk> bbl
May 15 13:14:11 <AStorm> jeremiah, I don't have the build env on me
May 15 13:14:41 <lbt> AStorm: don't you need to have an extras project for it?
May 15 13:14:45 <Jaffa> Aren't the PyMaemo packages so wildly different to Debian/Ubuntu that sorting out the Depends line will be "fun"
May 15 13:15:11 <Jaffa> lbt: no, anything can be put in the auto-builder/extras without a garage project. Garage is a hosting service entirely separate to extras
May 15 13:15:20 <AStorm> Jaffa, it won't, as ctypes has no deps
May 15 13:15:29 <AStorm> and multiprocessing deps only on ctypes
May 15 13:15:51 <AStorm> so these go like any python packages for Maemo
May 15 13:16:06 <AStorm> although.. I don't know how to package them
May 15 13:16:07 <AStorm> :)
May 15 13:16:55 <lopz> hola
May 15 13:17:07 <lbt> fiferboy: broken src incoming
May 15 13:17:31 <fiferboy> Do you have an idea about where the problem is?
May 15 13:18:27 <jeremiah> lcuk: I certainly would. :)
May 15 13:18:58 <mikkov_> Jaffa: PyMaemo for Fremantle is following debian packaging. For diablo it was completely different
May 15 13:19:14 <lbt> 				vel1.ry() = (
May 15 13:19:16 <lbt> 					(((vel.ry() >0 )
May 15 13:19:17 <lbt> 					  ? vsb->maximum()
May 15 13:19:19 <lbt> 					  : vsb->minimum()
May 15 13:19:20 <lbt> 						) - scrollBase.ry()) / scrollFactor.y() + curr.y())/braking_cycles;
May 15 13:20:05 <fiferboy> Okay, so that is overshoot decel
May 15 13:20:22 <lbt> > or < and should I be + or - scrollBase and is it the same for top/bottom
May 15 13:20:32 <lbt> and I should be able to do simple algebra
May 15 13:20:41 <lbt> and it's embarassing!
May 15 13:21:55 <lbt> circa line 1580
May 15 13:22:14 <fiferboy> Sometimes I rush through things like that, and when it is wrong try flipping a sign and making it more wrong
May 15 13:22:36 <jeremiah> AStorm: So, if I understand it, you want Ctypes packaged for Maemo?
May 15 13:22:40 <fiferboy> It's hard to start over and look at it objectively when you think you are close and a little change could fix everything
May 15 13:22:48 <lbt> yeah... done that... the list scrolls off the screen and vanishes
May 15 13:23:20 <lbt> did that too... but it was late last night and bed caught up with me...
May 15 13:23:25 <AStorm> jeremiah, that and multiprocessing (Debian squeeze has a backport)
May 15 13:23:25 <lbt> it's much saner now
May 15 13:23:33 <fiferboy> could it be this line:  if (vel1.ry() < 1.0 and vel1.ry() > -1.0)
May 15 13:23:45 <fiferboy> Should be or?
May 15 13:23:54 <lbt> no... that just makes sure we're close to 0
May 15 13:24:04 <lbt> i -1 < vel < 1
May 15 13:24:20 <jeremiah> AStorm: It looks like Ctype is provided by python
May 15 13:24:25 <lbt> if so then we're stopped
May 15 13:24:31 <fiferboy> Right, I'm not thinking well yet
May 15 13:24:31 <lbt> (d'uh)
May 15 13:24:33 <jeremiah> AStorm: How do you know it is missing in maemo?
May 15 13:24:47 <lbt> and the cartoon elastic kicks it
May 15 13:24:54 <AStorm> jeremiah, mhm, then it's not needed :)
May 15 13:24:56 <lbt> but it's real elastic
May 15 13:24:59 <AStorm> thus, only multiprocessing
May 15 13:25:11 <lbt> so it gives an instant kick and then constant velocity!!
May 15 13:25:12 <AStorm> it has some funny libjs-iquery dep
May 15 13:25:16 <jeremiah> I'll check the package in Squeeze.
May 15 13:25:24 <AStorm> *libjs-jquery
May 15 13:25:28 <jeremiah> libjs-iquery, oy.
May 15 13:25:35 <AStorm> http://packages.debian.org/squeeze/python-multiprocessing
May 15 13:25:40 <jeremiah> That will have lots of dependencies itself I'll bet
May 15 13:25:44 <AStorm> no, it doesn't
May 15 13:25:48 <jeremiah> Ah good
May 15 13:25:51 <AStorm> (except the javascript-common)
May 15 13:26:15 <AStorm> oh, even that isn't a dep, just recommended
May 15 13:27:21 <jeremiah> Why does a multithreading library rely on a Javascript web library?
May 15 13:27:26 <jeremiah> Very weird
May 15 13:27:28 <AStorm> probably XDR
May 15 13:27:44 <AStorm> (which is in Python 2.6, but not 2.5)
May 15 13:27:48 <AStorm> or some event system
May 15 13:27:54 <AStorm> weird anyway, esp. that:
May 15 13:27:55 <fiferboy> I understand why you are using the ry() function for the assignment, but why use it elsewhere when you are just getting the value?
May 15 13:27:56 <AStorm> http://packages.debian.org/squeeze/python-processing
May 15 13:28:00 <AStorm> doesn't have that dep
May 15 13:28:08 <AStorm> and it's almost the same lib
May 15 13:28:15 <fiferboy> lbt: Oops, I should have references you on the last one, it will get lost in here
May 15 13:28:16 <AStorm> (older version, actually)
May 15 13:28:44 <lbt> it did
May 15 13:29:25 <jeremiah> AStorm: Ah - so I might try to package the python-processing from Squeeze then. :)
May 15 13:29:26 <lbt> AFAIUI the ref based one is faster as the compiler essentially just uses a pointer and does no function calling....
May 15 13:29:44 <AStorm> jeremiah, I'd prefer multiprocessing, but any will do
May 15 13:29:49 <lbt> and cut'n'paste habit
May 15 13:30:11 <fiferboy> lbt: Is the function printing the debug info properly?
May 15 13:30:56 <lbt> yes
May 15 13:31:12 <lbt> 1 sec
May 15 13:32:30 <andrewfblack> Hello
May 15 13:32:36 <lbt> it's going back too far
May 15 13:34:38 <RST38h> hello andrew! how is minimalist2 doing?
May 15 13:35:19 <andrewfblack> I've been fighting my last bug for 2 days
May 15 13:35:19 <jeremiah> I'll package python-processing :) Perhaps you can try to package multiprocessing? Packaging python apps is not hard, and having me package and maintain everything in maemo does not scale well. :P
May 15 13:35:27 <fiferboy> lbt: The function is moving in the correct direction, but going too far?
May 15 13:35:38 <lbt> yeah I need to calc how overshot are we in pixels
May 15 13:36:06 <RST38h> andrew: anything that can use external help?
May 15 13:36:33 <AStorm> jeremiah, I probably should - this means I'd have to set up a VM for Maemo scratchbox
May 15 13:36:39 <AStorm> what was that URL?
May 15 13:37:08 <AStorm> and I don't have access rights to post it
May 15 13:38:08 <andrewfblack> well i'm having trouble with the thanks showing up twice on a post and using alot of space.  I considered removing thanks and releasing now then adding thanks back later,  You would be able to thank someone just not see who thanked on a post.
May 15 13:41:00 <RST38h> andrew: it really sounds quite acceptable to me, the thanks feature is nice but minor
May 15 14:07:14 <fiferboy> lbt: Do VMAX and DECEL_DURATION not cancel each other out?
May 15 14:08:03 <lbt> line?
May 15 14:08:19 <fiferboy> 1327 and 1328
May 15 14:08:39 <lbt> the definitions?
May 15 14:08:45 <lbt> vmax is max speed
May 15 14:08:54 <lbt> decel_duration is in milliseconds
May 15 14:09:01 <fiferboy> Yes.  Are you ever going to reach DECEL_DURATION now that you have VMAX?
May 15 14:09:01 <AStorm> :-( no jeremiah
May 15 14:09:10 <lbt> so normal 'friction
May 15 14:09:17 <fiferboy> I must misunderstand what DECEL_DURATION is doing
May 15 14:09:18 <lbt> lasts for 1.5seconds
May 15 14:09:25 <AStorm> and I'm fairly pressed for time on that (only a day or two at best)
May 15 14:09:30 <lbt> or until the deceleration hits 0
May 15 14:09:38 <fiferboy> lbt: What happens after 1.5 seconds?
May 15 14:09:39 <lbt> s/decel/velocity
May 15 14:09:42 <lbt> it stops
May 15 14:09:44 <lbt> dead
May 15 14:10:01 <lbt> its more a safety thing from before I had vmax :)
May 15 14:10:24 <fiferboy> But with VMAX and deceleration you should stop after a certain time anyway, I assume less than 1.5
May 15 14:10:44 <fiferboy> If it is more than 1.5 then DECEL_DURATION is just cutting off the movement
May 15 14:10:48 <lbt> yes
May 15 14:11:04 <lbt> now we have vmax the DECEL_DURATION isn't really needed
May 15 14:11:15 <fiferboy> Right
May 15 14:11:34 <lbt> good catch for optimising though
May 15 14:11:58 <fiferboy> Glad I could do something.  I don't understand your scrolling math enough to help with the overshoot problem
May 15 14:12:30 <lbt> :(
May 15 14:12:45 <crashanddie_> spl: ;)
May 15 14:12:57 <lbt> essentially you have pixel values (curr and vel)
May 15 14:13:19 <lbt> and the scrollbar has a range value which is not 1:1 pixels
May 15 14:13:28 <lbt> so scrollFactor converts
May 15 14:13:33 <fiferboy> Right.
May 15 14:13:48 <lbt> delta_px to delta_range only though
May 15 14:13:58 <fiferboy> So each step of the scrollbar is X pixels?
May 15 14:14:02 <lbt> so then there is a base offset for both px and range
May 15 14:14:10 <lbt> yes or 0.X
May 15 14:14:24 <fiferboy> Okay.
May 15 14:14:36 <lbt> hence floating point value too
May 15 14:14:50 <fiferboy> QPointF.  Got it.
May 15 14:15:20 <AStorm> lbt, so, you're overshooting because conversion is not exact
May 15 14:15:29 <lbt> no
May 15 14:15:38 <lbt> it's a gui 'effect'
May 15 14:15:43 <AStorm> how much?
May 15 14:15:49 <lbt> 'cos fremantle does it
May 15 14:15:52 <AStorm> ah
May 15 14:15:55 <lbt> it's kinetic scroll
May 15 14:15:57 <AStorm> hopefully toggleable
May 15 14:16:04 <lbt> yeah ;)
May 15 14:16:06 <AStorm> I know it's kinetic
May 15 14:16:16 * lbt isn't sold but it looks nice
May 15 14:16:28 <fiferboy> So with overshoot you have the velocity when it hits the end of the list
May 15 14:16:31 <AStorm> is less useful though than nice-looking
May 15 14:16:32 <lbt> yes
May 15 14:16:37 <fiferboy> How are you making the list scroll past it's limits?
May 15 14:16:54 <lbt> commented out the bounds check ;)
May 15 14:16:58 <lbt> in the core
May 15 14:17:00 <AStorm> the question is - what kind of deceleration do you want?
May 15 14:17:15 <fiferboy> lbt: Whoa, that's hardcore!
May 15 14:17:18 <lbt> factor/timestep
May 15 14:17:18 <AStorm> constant? linear? quadratic?
May 15 14:17:27 <AStorm> linear this is.
May 15 14:17:31 <AStorm> :)
May 15 14:17:33 <lbt> is it?
May 15 14:17:55 <lbt> velocity halves per 0.1 seconds
May 15 14:17:58 <AStorm> sure, dv/dt = const
May 15 14:18:02 <lbt> non-linear
May 15 14:18:14 <lbt> factor, not value
May 15 14:18:16 <AStorm> ah, so you have constant deceleration
May 15 14:18:23 <lbt> like gravity
May 15 14:18:26 <fiferboy> Well, I assume it is there for some reason.  Mostly for programs that don't check the bounds before trying to set the values.
May 15 14:18:27 <AStorm> so velocity is square root
May 15 14:18:41 <lbt> fiferboy: we'll do better
May 15 14:18:43 <AStorm> lbt, mhm, though I prefer sharper than that
May 15 14:18:59 <AStorm> maybe use log?
May 15 14:19:04 <lbt> actually I use 0.75
May 15 14:19:06 <fiferboy> Where do you comment the bounds check?
May 15 14:19:19 <lbt> AStorm: integer math?
May 15 14:19:23 <lbt> (mainly)
May 15 14:19:49 <AStorm> so v /= e^nslice
May 15 14:19:52 <lbt> qabstractslider
May 15 14:20:38 <lbt> vel1.rx() = (vel1.rx()*DECEL_PC)/100;
May 15 14:20:53 <fiferboy> Maybe instead of commenting it out, put in another function that doesn't have a bounds check?
May 15 14:21:16 <lbt> fiferboy: I will, and in fact may shortcut it out to use _hs_slide
May 15 14:21:39 <fiferboy> I'm just looking at qabstractslider source
May 15 14:21:50 <AStorm> lbt, why not v *= e^-dt ?
May 15 14:21:55 <AStorm> exp is in math lib
May 15 14:22:12 <fiferboy> You do it in the setValue functions?
May 15 14:22:46 <lbt> AStorm: simple approach 1st.... and it's pretty nice.... nb what's the equation for friction?
May 15 14:23:09 <lbt> fiferboy: yes
May 15 14:23:12 <AStorm> friction? that's "a force"
May 15 14:23:22 <AStorm> F = m*a
May 15 14:23:33 <fiferboy> I see it, value=d->bound(value)
May 15 14:23:53 <AStorm> v = 0.5 * a*t^2
May 15 14:25:39 <AStorm> k * v / t^2
May 15 14:25:55 <fiferboy> lbt: So in the code in question you have overshot and the velocity is close to 0, this is where you have to scroll back to the limit?
May 15 14:26:29 <AStorm> obviously, you need some epsilon to stop
May 15 14:26:55 <lbt> yes
May 15 14:27:21 <Stskeeps> lardman|afk: http://www.jiongtang.com/blog/html/digital-product/smartq7-unboxing.html btw
May 15 14:27:43 <fiferboy> lbt: So what is it you are setting vel1.ry to?
May 15 14:28:22 <AStorm> also, what is the timeslice? :)
May 15 14:28:37 <lbt> 50 ms
May 15 14:28:41 <AStorm> because the starting velocity is measured over a timeslice
May 15 14:29:22 <AStorm> that's too slow for decel, but decent for measuring
May 15 14:29:27 <lbt> http://pastebin.com/d40de85cf
May 15 14:29:32 <lbt> current code ^^
May 15 14:29:32 <fiferboy> (sbMax.y() - scrollBase.ry()/scrollFactor.y()+curr.y())/braking_cycles
May 15 14:29:34 <AStorm> at least, I think so
May 15 14:29:43 <lbt> fiferboy: now, I cheat here
May 15 14:29:48 <AStorm> you'll need, say, 2 timeslices to measure vstart
May 15 14:29:49 <lbt> and the rebound is linear velocity
May 15 14:30:07 <lbt> AStorm: done in the fingerscroll moveEvent code
May 15 14:30:12 <AStorm> mhm
May 15 14:30:31 <fiferboy> lbt: Is the new code closer to working?
May 15 14:30:37 <AStorm> then, on each timeslice, just apply the equation
May 15 14:30:42 <lbt> should be
May 15 14:30:54 <fiferboy> Compiling?
May 15 14:31:20 <lbt> 279 on that pastebin AStorm
May 15 14:31:43 <lbt> actually it's on the tablet but I'm chatting...
May 15 14:31:51 <AStorm> lbt, missed it?
May 15 14:32:04 <lbt> http://pastebin.com/d40de85cf
May 15 14:32:59 <lbt> need to go back to  0  from  -29  ScrollFactor  1  so  -251  px for  7  iterations
May 15 14:33:07 <lbt> yeah, that's sounds wrong...
May 15 14:33:25 <AStorm> rx *= 1/2 * k
May 15 14:34:32 <AStorm> or 1/2 * 1/k - for speed factor as opposed to slowness
May 15 14:34:38 <lbt> AStorm: what would you replace line 279 with ?
May 15 14:35:11 <fiferboy> lbt: Do 0 and -29 seem right?
May 15 14:35:16 <AStorm> ry = ry / 2 * somefactor
May 15 14:35:50 <lbt> that's the same code
May 15 14:35:52 <AStorm> then curr.y += ry
May 15 14:36:36 <AStorm> no, you're not changing velocity
May 15 14:36:45 <AStorm> which is wrong - it has to halve
May 15 14:36:50 <lbt> 283
May 15 14:37:09 <lbt> OVERSHOOT_DECEL_PC is 0.5 * somefactor
May 15 14:37:09 <AStorm> ugh
May 15 14:37:29 <lbt> ugh?
May 15 14:37:34 <lbt> it's the same maths
May 15 14:37:37 <AStorm> I see what you mean
May 15 14:37:54 <AStorm> how is vell.y and vell.ry related?
May 15 14:38:15 <lbt> originally I had the 100 in there to avoid integer maths results of 0.4 -> 0
May 15 14:38:21 <lbt> same value
May 15 14:38:25 <lbt> literally
May 15 14:38:34 <lbt> ry is a reference to y
May 15 14:38:49 <lbt> allows private data to be directly modified
May 15 14:39:11 <lbt> fiferboy: yes, 0 and 29 are spot on
May 15 14:39:15 <lbt> -29
May 15 14:40:04 <AStorm> I take it that ry is actually dy - delta?
May 15 14:40:04 <AStorm> dy = delta y I mean :)
May 15 14:40:04 <AStorm> so it's then that vell.y() is to halve
May 15 14:40:07 <AStorm> not vell.ry which is the delta
May 15 14:40:09 <AStorm> delta would be constant 1/2 * somefactor
May 15 14:40:19 <AStorm> :)
May 15 14:40:21 <AStorm> ok
May 15 14:40:25 <lbt> no
May 15 14:40:37 <lbt> vel is a QPoint object
May 15 14:40:43 <lbt> holding x and y
May 15 14:40:44 <fiferboy> So, shouldn't it be "need to go back to 0 from -29 ScrollFactor 1 so (scrollBase.ry() + curr.y() * scrollFactor.y()) px for 7 iterations
May 15 14:40:59 <lbt> yes... but I'm chatting again
May 15 14:41:02 <lbt> heh
May 15 14:41:35 <AStorm> hmmh
May 15 14:41:44 <Macer> i am watching a movie called conspiracy about the nazi meeting about the final solution
May 15 14:41:47 <AStorm> so do calculate actual velocity, that is dx
May 15 14:41:50 <AStorm> and dy
May 15 14:41:52 <AStorm> and not cheat
May 15 14:42:06 <AStorm> then halve that and add to curr.y
May 15 14:42:08 <Macer> and one of the actors is jewish :)
May 15 14:42:15 <lbt> no need
May 15 14:42:21 <AStorm> (after multiplying by the scale factor)
May 15 14:42:23 <lbt> and anyway
May 15 14:42:35 <lbt> working at pixel level loses accuracy
May 15 14:42:41 <lbt> and you get nasties
May 15 14:43:02 <AStorm> sure it does, but dx is not in whole pixels
May 15 14:43:43 <lbt> I should have said vel is a QPointF object (float) so that keeps things honest
May 15 14:44:10 <AStorm> again, what is vel. velocity? so vel.x is dx?
May 15 14:44:27 <lbt> no vel is velocity vector
May 15 14:44:33 <lbt> vel.x is x component
May 15 14:45:25 <AStorm> ...
May 15 14:45:35 <AStorm> so vel.x is dx, you.
May 15 14:46:28 <AStorm> KINETIC_REFRESH/VSCALE
May 15 14:46:40 <AStorm> I know what is VSCALE - but kinetic refresh?
May 15 14:47:46 <AStorm> is it dt?
May 15 14:49:22 * Stskeeps is kinda impressed
May 15 14:49:30 <RST38h> Makesyour download go faster if you continue circling the screen with your finger?
May 15 14:49:38 <Stskeeps> the public services with digital signature etc and so on work on ubuntu
May 15 14:49:59 <AStorm> RST38h, looks like it, I don't know what this crud is
May 15 14:50:10 <AStorm> the screen should be refreshed every some ms
May 15 14:50:18 <AStorm> e.g. 50 ms
May 15 14:50:22 <lbt> vel.x is x component
May 15 14:50:31 <lbt> sorry wrong window
May 15 14:51:17 <lbt> AStorm: yes
May 15 14:51:28 <AStorm> but then, you need to calculate delta for that time, so vel.x = (cur.x-old.x)/TIMESLICE
May 15 14:51:41 <AStorm> assuming you measure over one timeslice
May 15 14:51:47 <lbt> I'm not calculating vel
May 15 14:51:51 <AStorm> ?
May 15 14:51:54 <lbt> I'm calculating dx using vel
May 15 14:51:59 <AStorm> wth?
May 15 14:52:14 <AStorm> no, you have to compute starting velocity somehow
May 15 14:52:17 <lbt> this is overshoot?
May 15 14:52:26 <lbt> we know start velocity
May 15 14:52:33 <AStorm> do don't
May 15 14:52:38 <lbt> line 190 ish
May 15 14:52:39 <AStorm> *you don't
May 15 14:53:06 <lbt> we also calculate it twice
May 15 14:53:29 <lbt> to avoid some glitches in mouse events being duplicated on some systems
May 15 14:53:29 <AStorm> I see *VSCALE there
May 15 14:53:32 <AStorm> why?
May 15 14:53:43 <lbt> const int QAbstractScrollAreaPrivateScroller::VSCALE          = 100;  // velocity scaling to ensure not lost in int rounding
May 15 14:54:07 <AStorm> mhm, ok
May 15 14:54:31 <AStorm> last - curr = FAIL
May 15 14:54:36 <AStorm> curr - last
May 15 14:54:47 <AStorm> wrong direction
May 15 14:56:57 <AStorm> again, what is KINETIC_REFRESH
May 15 14:57:14 <AStorm> think about it
May 15 14:57:16 <lbt> hmm, good point... I had something on this way back... it's because the scrollbars are inverted IIRC
May 15 14:57:19 <AStorm> why do you multiply by it
May 15 14:57:29 <lbt> timeslice
May 15 14:57:46 <AStorm> yes, but your speed is in pixels
May 15 14:58:01 <AStorm> not pixels/second
May 15 14:58:11 <lbt> that's not a speed
May 15 14:58:13 <AStorm> actually, pixels/ms
May 15 14:58:32 <AStorm> or whatever your dt base unit is
May 15 14:59:19 <AStorm> is cur_time Unix timestamp?
May 15 14:59:53 <AStorm> event_time.elapsed() - what's the unit?
May 15 15:01:55 <AStorm> OVERSHOOT_DECEL_PC should be 50
May 15 15:02:30 <lbt> I think you could really help point out inconsistencies in the code... it may be worth reading through it a bit to get the structure and find where some of the points you raise are commented... Then I'd really appreciate patches, comments or suggestions :)
May 15 15:02:45 <AStorm> or if you really want it accurate, then -
May 15 15:02:48 <lbt> I do
May 15 15:03:04 <lbt> but I'm very close to it
May 15 15:03:13 <lbt> an independent review would help
May 15 15:03:21 <AStorm> DECEL_PC/100 / 2
May 15 15:03:30 <AStorm> the whole value
May 15 15:03:40 <AStorm> the default should be halving the speed
May 15 15:03:44 <RST38h> What program are you all looking at anyway???
May 15 15:03:52 <lbt> http://pastebin.com/d40de85cf
May 15 15:03:52 <RST38h> What is all this kinetic stuff?
May 15 15:03:54 <AStorm> now, the factor is slowness factor
May 15 15:04:01 <fiferboy> RST38h: Qt finger scrolling
May 15 15:04:13 <lbt> RST38h: extension to qAbstractScrollArea to allow fingerscroll + fremantle bounce
May 15 15:04:28 <AStorm> or rather, 0.5 * (DECEL_PC/100)
May 15 15:04:36 <AStorm> then rounded to int
May 15 15:04:43 <fiferboy> David has implemented it in such a way that all Qt programs running in Hildon automatically take advantage of it
May 15 15:04:47 <RST38h> Ah, cool
May 15 15:05:11 <lbt> to get it accepted it needs to be minimal and efficient
May 15 15:05:13 <AStorm> if necessary, do 5 * DECEL_PC/100 / 10
May 15 15:05:27 <lbt> so whilst true physics is nice
May 15 15:05:34 <lbt> approximations are good enough ;)
May 15 15:05:38 <AStorm> it is true physics :)
May 15 15:05:46 <AStorm> that is, almost true physics
May 15 15:05:55 <AStorm> (you're missing the 1/2 factor)
May 15 15:06:12 <lbt> but that's not the problem
May 15 15:06:12 <RST38h> true physics is where it shatters your device when it reaches the end of the list at full speed
May 15 15:06:20 <lbt> and I'm using large values now
May 15 15:06:22 <lbt> to debug
May 15 15:06:30 <AStorm> large values of what?
May 15 15:06:30 <lbt> so I actually see the effect in slow motion
May 15 15:06:40 <lbt> OVERSHOOT_DECEL_PC 50
May 15 15:06:49 <AStorm> mhm
May 15 15:06:54 <lbt> should be closer to 80... rubber wall, not gravity
May 15 15:06:58 <AStorm> then add that 5/10
May 15 15:07:06 <lbt> yeah... so when we tune it that's great
May 15 15:07:12 <lbt> but right now it's broken
May 15 15:07:14 <lbt> :(
May 15 15:07:19 <AStorm> broken, how?
May 15 15:07:33 <lbt> heh,
May 15 15:07:34 <AStorm> (show some print output)
May 15 15:07:45 <jukey> hi , is there anybody who know where to get information about an error of maemoscrobbler?
May 15 15:08:01 <lbt> http://pastebin.com/d28c7b695
May 15 15:08:07 <jukey> the gui shows me: Status: Error submitting (14:56)
May 15 15:08:12 <lbt> how to describe broken animation on irc ?
May 15 15:08:24 <lbt> given my camera went for repair 2 days ago?
May 15 15:08:34 <fiferboy> lbt: Does this make any sense?  http://pastebin.com/m313311cb
May 15 15:08:47 <fiferboy> Line 272
May 15 15:08:50 <AStorm> lbt, something is very wrong... you get constant v
May 15 15:09:14 <lbt> OverShootStabilise is supposed to be constant v
May 15 15:09:19 <AStorm> this shouldn't happen, it should always halve
May 15 15:09:35 <AStorm> no, or it's not physics
May 15 15:09:41 <AStorm> ah, wait
May 15 15:09:43 <lbt> actually it should accelerate and then decelerate
May 15 15:09:45 <AStorm> what does that do?
May 15 15:09:52 <lbt> cos at the turning point v=0
May 15 15:09:54 <AStorm> the Stabilize part
May 15 15:10:06 <lbt> read the state code
May 15 15:10:38 <AStorm> #
May 15 15:10:38 <AStorm>                         if (braking_cycles-- <= 0) {
May 15 15:10:41 <lbt> fiferboy: almost
May 15 15:10:45 <AStorm> bad, use velocity for cutoff
May 15 15:10:49 <AStorm> not cycles
May 15 15:11:10 <lbt> vsb->max is in 'value', not pixels
May 15 15:11:29 <AStorm> e.g. if vel.ry() < something, vel.ry() = 0
May 15 15:11:39 <lbt> fiferboy:   ) - scrollBase.ry()) / scrollFactor.y() - curr.ry())/braking_cycles;
May 15 15:11:43 <AStorm> (actually, abs(vel.ry())
May 15 15:12:05 <lbt> I think we subtract target - old_val
May 15 15:12:23 <lbt> and I did target - (a +b) = target - a +b
May 15 15:13:10 <lbt> AStorm: this code has 2 decelerations and a rebound
May 15 15:13:21 <lbt> friction deceleration on release
May 15 15:13:31 <lbt> rubber deceleration on hitting the limit
May 15 15:13:37 <lbt> and linear rebound
May 15 15:13:45 <AStorm> rubber deceleration?
May 15 15:13:50 <AStorm> wth is that?
May 15 15:13:57 <lbt> imagine hitting a rubber end-stop
May 15 15:14:05 <AStorm> ah, that thing
May 15 15:14:09 <AStorm> you need an oscillation for that
May 15 15:14:23 <AStorm> e.g. sin * 1/2 * k
May 15 15:14:30 <lbt> not pretty in a gui
May 15 15:14:41 <AStorm> yep ;P
May 15 15:14:50 <AStorm> other than that, no need for deceleration - just stop
May 15 15:14:54 <lbt> so we have what I descrobed ... 2 decelerations and a linear rebound
May 15 15:14:57 <AStorm> :)
May 15 15:15:15 <fiferboy> lbt: Are the numbers any closer now?
May 15 15:16:00 <AStorm> ok, again, when is "rubber decel" active
May 15 15:16:10 <AStorm> and rebound happens when you hit an edge?
May 15 15:16:16 <lbt> no
May 15 15:16:25 <lbt> rubber decel at an edge
May 15 15:16:36 <AStorm> uhm, wha?
May 15 15:16:39 <lbt> rebound when rubber decel stops you
May 15 15:16:58 <AStorm> how would that be useful?
May 15 15:17:16 <AStorm> so that it stops always before you hit an edge?
May 15 15:17:27 <AStorm> it's horrible
May 15 15:17:29 <lbt> no, it *starts* to stop
May 15 15:17:33 <lbt> when you hit an edg
May 15 15:17:35 <lbt> e
May 15 15:17:40 <lbt> and then goes past the edge
May 15 15:17:45 <lbt> straining mightily
May 15 15:17:48 <AStorm> ah, so you're simulating an indented surface?
May 15 15:17:49 <lbt> like in a cartoon
May 15 15:18:00 <lbt> a rubber wall
May 15 15:18:10 <lbt> I call my widget Tom
May 15 15:18:32 <AStorm> ok, so proper function is still friction...
May 15 15:18:34 <lbt> then when he has fully stretched out the rubber wall
May 15 15:18:42 <AStorm> but you get additional force added
May 15 15:18:42 <lbt> no, elastic
May 15 15:18:47 <AStorm> same!
May 15 15:18:54 <AStorm> it's still a force
May 15 15:19:04 <lbt> no elastic force proportional to stretch
May 15 15:19:09 <AStorm> (at least, can be modelled as one)
May 15 15:19:22 <AStorm> let me check
May 15 15:19:31 <lbt> bow+arrow
May 15 15:19:54 <fiferboy> lbt: I would describe it more like an elastic band holding something back when you hit the edge
May 15 15:19:55 <AStorm> http://en.wikipedia.org/wiki/Linear_elasticity
May 15 15:20:06 <AStorm> ^
May 15 15:20:13 <fiferboy> It pulls against it (decelerating) and then it starts being pulled back
May 15 15:20:43 <lbt> yes
May 15 15:20:49 <lbt> but mine twangs
May 15 15:20:58 <lbt> imparting energy in a sudden impulse
May 15 15:21:04 <lbt> inside the first timeslice
May 15 15:21:05 <lardman> whoa, still doing elestic modelling
May 15 15:21:11 <AStorm> working  correctly
May 15 15:21:12 <lbt> and having cV after that...
May 15 15:21:21 <AStorm> that's how elasticity works
May 15 15:21:36 <AStorm> all the stored energy is released at once
May 15 15:21:39 <fiferboy> Ah, so different from most kinetic scrolling rebounds I have seen
May 15 15:21:40 <lbt> In addition linear elasticity is only valid for stress states that do not produce yielding.
May 15 15:22:00 <lardman> or heating
May 15 15:22:22 <AStorm> lbt, in other words, hitting a spring is linear
May 15 15:22:26 <AStorm> (more or less)
May 15 15:22:41 <lbt> anyway, I don't care :D
May 15 15:23:16 <lbt> right now I'm getting antimatter hitting non-linear springs and some kind of perpetual motion
May 15 15:23:23 <AStorm> so, your model acts correctly... like a spring :)
May 15 15:23:35 <lardman> do you have a damping term in there too?
May 15 15:23:36 <fiferboy> lbt: I like the sound of that!
May 15 15:23:45 <lardman> obviously not, but do you plan on having one?
May 15 15:23:45 <AStorm> lardman, friction should be the damping term
May 15 15:23:46 <lbt> it's so broken that I don't care about the precise shape of the deceleration curve
May 15 15:23:59 <lbt> I have 2 damping terms
May 15 15:24:02 <AStorm> lbt, you have to apply friction to the rebound too
May 15 15:24:05 <lbt> one for normal slowdown
May 15 15:24:11 <lbt> one for hitting an end stop
May 15 15:24:19 <AStorm> wrong
May 15 15:24:24 <AStorm> friction should apply to both
May 15 15:24:36 <lbt> wrong
May 15 15:24:43 <lbt> one is friction
May 15 15:24:46 <lbt> one is elastic
May 15 15:24:57 <lbt> one results in rebound one doesn't
May 15 15:25:07 <AStorm> "hitting an end stop" should happen before it hits an edge, right?
May 15 15:25:22 <AStorm> or as it hits an edge?
May 15 15:25:30 <lbt> the end stop/edge is a rubber band stretched across the path
May 15 15:25:45 <lbt> and it goes past the limit
May 15 15:25:46 <AStorm> yes
May 15 15:25:59 <lbt> so Y coords:
May 15 15:26:13 <fiferboy> lbt: What are your debug numbers after the latest changes?
May 15 15:26:17 <AStorm> ok, then your rebound velocity should be calculated like this
May 15 15:26:35 <AStorm> velrebound = - vel * springyfactor
May 15 15:26:45 <lbt> keep chatting, I'm going to code some things and start a build
May 15 15:26:45 <AStorm> where that factor is less than 1
May 15 15:27:27 <andrewfblack> RST38h: you around?
May 15 15:27:41 <lardman> bbiab
May 15 15:27:58 <AStorm> lbt, and after the rebound, it should be also affected by friction
May 15 15:28:06 <AStorm> you just get another starting velocity
May 15 15:28:17 <lbt> 0
May 15 15:28:23 <AStorm> not 0
May 15 15:28:30 <AStorm> -vel * 1/springiness
May 15 15:28:56 <lbt> o
May 15 15:29:12 <lbt> I guarantee you it's zero
May 15 15:29:29 <AStorm> it's not if it was scrolling
May 15 15:29:33 <lbt> since it passes from +ve velocity to -ve velocity
May 15 15:30:06 <AStorm> ...
May 15 15:30:22 <AStorm> rebound happens at an edge
May 15 15:30:28 <AStorm> when velocity is >0
May 15 15:30:31 <AStorm> otherwise it stops
May 15 15:31:26 <AStorm> actually, when abs(velocity) > 0
May 15 15:31:53 <AStorm> so you still need to calculate velocity somehow
May 15 15:32:13 <AStorm> this time, not from scrolling, which doesn't happen
May 15 15:32:51 <AStorm> so, you get 2 cases
May 15 15:33:05 <AStorm> one, where point was moving before hitting the edge
May 15 15:33:20 <AStorm> and 2, when your rebound starts at an edge
May 15 15:34:14 <lbt> imagine the scrollbar is a ball on a traintrack. It rolls and eventually stops. That's friction. It rolls and slows down but before it stops it hits a rubber band stretched across the track between two posts. It slows down more quickly but goes past the posts and stretches the rubber band until it stops and the band is stretched. Now the band snaps but gives the ball enough speed to get back to the posts in the same time as it got from the posts to the stopping
May 15 15:34:15 <lbt> point. Then a magic spell stops the ball dead when it is level with the posts.
May 15 15:34:38 <lbt> that's my model that is ;)
May 15 15:34:50 <crashanddie_> a ball on a traintrack would stop in less that 2m
May 15 15:34:56 <crashanddie_> have you never noticed all the bumps?
May 15 15:34:59 <AStorm> yuck
May 15 15:35:07 <lbt> monorail
May 15 15:35:09 <AStorm> this model will suck in use
May 15 15:35:10 <crashanddie_> it's like the wheel of fortune, squared
May 15 15:35:26 <AStorm> so, when do you want to stop at an edge
May 15 15:35:28 <timeless_mbp> hi crashanddie
May 15 15:35:31 <AStorm> 1) when you hit it
May 15 15:35:38 <crashanddie_> hi timeless_mbp
May 15 15:35:39 <timeless_mbp> ok, i think i have a couple of things ironed out
May 15 15:36:03 <timeless_mbp> sadly i failed to hit up everyone i wanted to for this week, but i think overall i've gotten to see quite a few people
May 15 15:36:09 <AStorm> or... 2) when remaining energy of the point is below the band's inertia?
May 15 15:36:11 <lbt> AStorm: http://www.youtube.com/watch?v=G-D-J0Y7ufM  19 seconds in
May 15 15:36:26 <lcuk> lbt, if you did it with a rubber band and the ball hit it with enough power to actually bend the band then the ball will fire back off down the track
May 15 15:36:44 <lbt> that's the 'rebound' phase
May 15 15:36:59 <crashanddie_> lbt: as lcuk just pointed out, there's definitely a few problems with your metaphore
May 15 15:37:04 <lbt> but it doesn't need a full physics engine
May 15 15:37:15 <lcuk> so its a sticky rubber band so when the ball hits it its glued, but oscilates around the natural flex point of the band
May 15 15:37:24 <lcuk> we should add a twoioiinggggg sound effect
May 15 15:37:26 <crashanddie_> that's like brickbreaker
May 15 15:37:27 <lbt> crashanddie_: my metaphor ends in magic.... how can it be wrong
May 15 15:38:13 <AStorm> lbt, note it goes BEYOND the are
May 15 15:38:15 <AStorm> *area
May 15 15:38:18 <fiferboy> lbt: Are your debug numbers any better now?
May 15 15:38:37 <lbt> fiferboy: think so... but not right
May 15 15:38:49 <fiferboy> AStorm: That's what the overshoot is
May 15 15:38:53 <AStorm> so your velocity is being reduced in a linear function of dv
May 15 15:39:04 <AStorm> then goes to zero
May 15 15:39:31 <lbt> fiferboy: you understand... can you explain?
May 15 15:39:32 <AStorm> then springs back again as a linear function of dv, possibly multipled by some "loss factor"
May 15 15:40:05 <AStorm> in other words, dy is a square function of y
May 15 15:40:14 <AStorm> or sqrt of y
May 15 15:40:17 <fiferboy> Basically, it does what the video lbt posted shows.
May 15 15:40:21 * lbt goes for a coffee
May 15 15:40:24 <AStorm> depending on the direction
May 15 15:40:35 <crashanddie_> timeless_mbp: you still on for lunch then?
May 15 15:40:42 <AStorm> (into rubber it's sqrt, out of the rubber it's a square)
May 15 15:40:43 <fiferboy> Shoots past the end, slows down, comes back and settles happily at the list end (or start)
May 15 15:40:44 <timeless_mbp> i believe so
May 15 15:40:46 <AStorm> assuming linear rubber
May 15 15:40:50 <lcuk> is there a maximum amount that an overshoot can be
May 15 15:40:53 <crashanddie_> timeless_mbp: I didn't see you reply, you might want to confirm
May 15 15:40:58 <AStorm> and *all that* is affected by friction still
May 15 15:41:08 <timeless_mbp> although they're going to need to repoke me w/ details real soon
May 15 15:41:26 <fiferboy> lcuk: There is a maximum if you have released your finger.  You can pull it back all you want with your finger on the widget
May 15 15:41:29 <crashanddie_> timeless_mbp: they did,  they're waiting for your reply
May 15 15:41:46 <AStorm> lbt, so, you're missing a cutoff by going to v=0
May 15 15:42:03 <lcuk> upto 1 screen w/h i assume - or can you keep pulling it back release,regrab quickly and pull back more
May 15 15:42:08 <AStorm> you need v <= const instead
May 15 15:42:11 <crashanddie_> timeless_mbp: reply to yesterday's email and propose a time as well
May 15 15:42:13 <timeless_mbp> ok, i need to learn to use Reply To All
May 15 15:42:16 <lcuk> if you do that for a list of 10000 items do you break the speed of light restoring it
May 15 15:42:19 <timeless_mbp> arg, and a time?
May 15 15:42:20 <AStorm> or limit vmax
May 15 15:42:26 <timeless_mbp> hrm
May 15 15:42:28 <fiferboy> lcuk: I don't think you can regrab, but that would be fun
May 15 15:42:33 <timeless_mbp> ok
May 15 15:42:37 <crashanddie_> timeless_mbp: well if it's today, you might want to start pushing things :)
May 15 15:42:37 <lcuk> yeah :D
May 15 15:42:44 <timeless_mbp> yeah yeah
May 15 15:42:50 <AStorm> make sure to get the loss factor in
May 15 15:42:56 <lcuk> i think all list items should be independent, and act like newtons cradle :D
May 15 15:42:56 * timeless_mbp hasn't really woken up yet
May 15 15:43:10 <AStorm> lcuk, that'd be fun
May 15 15:43:12 <crashanddie_> timeless_mbp: oh you fucking twat
May 15 15:43:18 <Jaffa> mgedmin: ping
May 15 15:43:24 <crashanddie_> timeless_mbp: you're a veggie? That's so 90s :D
May 15 15:43:28 <timeless_mbp> i'm not
May 15 15:43:29 <mgedmin> hi, Jaffa
May 15 15:43:33 <lcuk> yeah until a list item at the end fires off and out through the side of the device and takes someones eye out
May 15 15:43:44 <timeless_mbp> I actually had a meat meal last night on Castro street
May 15 15:43:53 <timeless_mbp> it turns out they opened a kosher restaurant there a week ago
May 15 15:43:55 <AStorm> lcuk, no, it'll rebound from the edge
May 15 15:43:57 <AStorm> :D
May 15 15:43:58 <fiferboy> luck: Don't Fremantle devices come with goggles to prevent that?
May 15 15:44:09 <timeless_mbp> but I'm not going to go there two days in a row
May 15 15:44:13 <lcuk> AStorm, are you sure the materials at the end are structurally sound
May 15 15:44:26 <AStorm> they are
May 15 15:44:30 <lcuk> nokia testing labs need to build a device to test this theory
May 15 15:44:36 <fiferboy> Some of my list items have a lot of mass
May 15 15:44:40 <lcuk> like the pocket shuffler thing
May 15 15:44:45 <lcuk> heh
May 15 15:45:07 <AStorm> lbt, so, it's fake rubber band
May 15 15:45:12 <lcuk> we should use the accel unit promised for fremantle to allow a list to be shook up
May 15 15:45:14 <lcuk> to test it
May 15 15:45:18 <AStorm> the item's velocity is set to some constant upon hitting an edge
May 15 15:45:27 <AStorm> that's how it looks
May 15 15:45:32 <fiferboy> lcuk:  That would be an awesome way to suffle a playlist
May 15 15:45:36 <lcuk> yeah
May 15 15:45:45 <AStorm> otherwise, it'd overshoot A LOT
May 15 15:45:52 <lbt> no, it starts to decelerate on hitting an edge
May 15 15:46:05 <AStorm> or the factor is huge
May 15 15:46:10 <lbt> but faster than friction
May 15 15:46:10 <lcuk> fiferboy, ive been playing with lists quite a lot as you know, have you seen and felt the new framework stuff - its got an almost infinite kinetic caledar thats smooth as silk
May 15 15:46:16 <Jaffa> mgedmin: your cp's failing on nfs - you using nfs-user-server?
May 15 15:46:20 <AStorm> then it shouldn't overshoot at all at small v
May 15 15:46:25 <lbt> correct
May 15 15:46:32 <AStorm> check it
May 15 15:46:38 <fiferboy> lcuk:  You mean the Qt Declarative UI videos?
May 15 15:46:38 <mgedmin> Jaffa: no, I don't think so
May 15 15:46:44 <AStorm> btw, the correct factor is based on the width
May 15 15:46:45 <AStorm> :)
May 15 15:46:51 <lcuk> fiferboy, theres a playtest out in the wild and keesj is tracking the source
May 15 15:46:57 <AStorm> (in case of dy, or height in case of dx)
May 15 15:47:00 <mgedmin> Jaffa: nfs-kernel-server 1:1.1.2-6lenny1
May 15 15:47:01 <lcuk> its on a git server somewhere
May 15 15:47:12 * mgedmin needs blog comments damnit!
May 15 15:47:16 <Jaffa> mgedmin: Kernel server? I've had loads of problems from hardy onwards with the user server - but it's the only one which gives me the uid mapping I need :(
May 15 15:47:29 <mgedmin> yeah, that sounds familiar
May 15 15:47:31 <lcuk> fiferboy, much more is in place since the pretty zooming vid :)
May 15 15:47:32 <crashanddie_> how is java dev going on the n8x0?
May 15 15:47:42 <mgedmin> I tried to use it for the same reasons and had to stop for some other reasons
May 15 15:47:46 * AStorm needs that python-processing... fairly quickly
May 15 15:47:47 <timeless_mbp> crashanddie_: who does that?
May 15 15:47:51 <fiferboy> lcuk:  Which video are you referring to?
May 15 15:47:56 <AStorm> and jeremiah is still out
May 15 15:47:58 <AStorm> :/
May 15 15:48:01 <Jaffa> mgedmin: don't go into too much detail now ;-)
May 15 15:48:03 <lcuk> http://www.youtube.com/watch?v=iMXp0Dg_UaY this vid
May 15 15:48:47 <lcuk> http://talk.maemo.org/showpost.php?p=285896&postcount=260 this post where i mention about the playtest and how to run it etc
May 15 15:49:22 <lcuk> find out how different mechanisms work on actual hardware and what works etc :)
May 15 15:50:26 <AStorm> lbt, indeed, this is fake spring
May 15 15:50:26 <AStorm> see how it takes some time when it reaches max overshoot
May 15 15:50:28 <AStorm> normal elastic would either break or act at once
May 15 15:50:36 <AStorm> if it breaks, velocity is still non-negative
May 15 15:50:38 <AStorm> if it acts at once, v == 0 at some point
May 15 15:50:50 <crashanddie_> timeless_mbp: https://wiki.evolvis.org/jalimo/index.php/Maemo
May 15 15:51:15 <AStorm> so, they probably set the velocity to constant with - upon hitting some max_overshoot
May 15 15:51:32 <AStorm> (and it still gets friction applied when returning)
May 15 15:53:59 <timeless_mbp> crashanddie_: could you ask those clowns what they have against getting a proper certificate?
May 15 15:54:16 <timeless_mbp> Chinoook
May 15 15:54:22 <timeless_mbp> 3 o's?
May 15 15:54:24 <timeless_mbp> gah
May 15 15:55:19 <AStorm> lbt, ok, let me write you the correct equation
May 15 15:55:34 <lbt> :)
May 15 15:57:27 <AStorm> v *= 1/2 * slowness - (hit_edge ? springfactor * 1/overshoot : 0)
May 15 15:58:00 <AStorm> argh
May 15 15:58:10 <AStorm> no, it is correct :)
May 15 15:58:50 <AStorm> now, when v reaches zero and you're overshot, you apply
May 15 15:59:05 <AStorm> v = - springfactor * 1/overshoot
May 15 15:59:57 <lardman> lol, still going :)
May 15 16:00:50 <lcuk> yeah lardman the mechanics of underpant elastic and the forces needed to give a satisfying wedgy are complex
May 15 16:01:29 <fiferboy> Well, I'm going to go get some fresh air
May 15 16:01:34 <lcuk> im surprised you havent got involved and modelled it for them in [*insert random mathematical prog here*]
May 15 16:01:41 <lardman> I missed the start
May 15 16:01:46 <lardman> was too busy doing some other modelling
May 15 16:01:48 <fiferboy> Hopefully all the friction/spring/elastic/magic will be worked out before I get back
May 15 16:01:51 <lcuk> spring back
May 15 16:02:05 <AStorm> or rather, no
May 15 16:02:05 <AStorm> no, that is correct
May 15 16:02:05 <AStorm> then apply friction to the resulting constant
May 15 16:02:06 <lardman> spring back will a well damped frictional halt
May 15 16:02:10 <AStorm> obviously, this is invalid, but seems to be what's shown in the movie
May 15 16:02:22 <AStorm> should be v = 1/2 * speedfactor * (v - springfactor * 1/overshoot)
May 15 16:02:24 <AStorm> and I'm lagged :P
May 15 16:02:29 <lcuk> cya later fiferboy_away
May 15 16:03:05 * AStorm needs pyprocessing BADLY!
May 15 16:03:14 <AStorm> maybe someone can do it quicker than jeremiah?
May 15 16:03:35 <lardman> can I ask what you're having troubles with?
May 15 16:03:41 <biiter_> did someone found any reasonable use for the camera on n810 ??
May 15 16:03:43 <lardman> some sort of kinetic scrolling thing?
May 15 16:03:58 <lcuk> its the end of the kinetics they are trying to work out
May 15 16:03:59 <lardman> biiter_: barcode scanning
May 15 16:04:00 <AStorm> biiter_, yes, automatic rotation and barcode scanning
May 15 16:04:16 <lardman> biiter_: I like the sound of starmap matching
May 15 16:04:36 <lcuk> use it for video calling
May 15 16:04:46 <lardman> yes, that one too
May 15 16:04:48 <lcuk> and for taking pics in liqbase
May 15 16:04:59 <AStorm> pics of yourself, you mean
May 15 16:05:03 <lardman> did anyone get around to making the gesture recognition app?
May 15 16:05:09 <AStorm> hard to aim otherwise
May 15 16:05:10 <lcuk> AStorm, or your friends :)
May 15 16:05:26 <lbt> lardman http://pastebin.com/d77515911
May 15 16:05:36 <lardman> needs face recognition so you can point it randomly and it will boing when it's pointing at someone
May 15 16:05:41 <biiter_> ok :) thx
May 15 16:05:48 <AStorm> lardman, neat idea!
May 15 16:06:01 <AStorm> we have opencv on the tablet?
May 15 16:06:30 <lardman> there's a Garage project isn't there?
May 15 16:06:37 <AStorm> IDK
May 15 16:06:50 <lardman> there was talk of getting it working, I thought just some basic image correlation code ought to do the job though
May 15 16:07:03 <AStorm> standard Haar recognizer
May 15 16:07:05 <lcuk> bbiab tracy poorly again
May 15 16:07:18 <AStorm> it will take like, 30 lines of code to find that face
May 15 16:07:27 <AStorm> or those faces
May 15 16:07:27 <lardman> lcuk: more wine!
May 15 16:08:00 <lardman> AStorm: from scratch in C?!
May 15 16:08:01 <lardman> :)
May 15 16:08:08 <AStorm> lardman, C++, with OpenCV
May 15 16:08:25 <lardman> ah cool
May 15 16:08:40 <AStorm> of course, given training material
May 15 16:09:05 <AStorm> but there are ready-made bases of pics to get a good classifier
May 15 16:09:07 <derf> Don't the Haar approaches use like 200,000 features?
May 15 16:09:16 <AStorm> derf, yes, more
May 15 16:09:19 <AStorm> but not that many examples
May 15 16:09:24 <derf> You seriously think that'll run on a tablet?
May 15 16:09:32 <AStorm> you can get good results with like 9000 positive + 9000 negative
May 15 16:09:33 <derf> At anything approaching video frame rates?
May 15 16:09:37 <AStorm> sure yeah
May 15 16:09:42 <derf> You're crazy.
May 15 16:09:49 <AStorm> assuming camera doesn't eat CPU
May 15 16:09:58 <derf> Hahahaha.
May 15 16:09:59 <AStorm> 10 FPS
May 15 16:10:25 <lardman> lbt: what is VSCALE? Some sort of scaling factor to slow the speed?
May 15 16:10:27 <AStorm> it takes like, 10ms to find a face on a single core of Turion X2 Ultra 2200 MHz
May 15 16:10:36 <lcuk> how do they do facial recognition on the canon cameras?
May 15 16:10:37 <AStorm> lardman, just 100
May 15 16:10:50 <lcuk> or is that just facial feature recognition
May 15 16:10:53 <AStorm> lcuk, probably the same, or morphological detection
May 15 16:10:53 <lardman> ok, just a scaling factor
May 15 16:11:10 <lcuk> i put sketch searching into liqbase last night
May 15 16:11:20 <ShadowJK> the facial recognition on canon cameras is funny. It thinks a beer can is a face, but fails to notice people :-)
May 15 16:11:25 <lcuk> i made an api to draw a sketch and it shows all the sketches which have these features
May 15 16:11:38 <derf> ShadowJK: Needs more AdaBoost.
May 15 16:11:52 <lcuk> lol shadow
May 15 16:15:06 <AStorm> ok, so, again
May 15 16:15:17 <AStorm> you only need 1 FPS for it to be useful for pics
May 15 16:15:44 <AStorm> better to have more, of course
May 15 16:15:56 <AStorm> 5 FPS is certainly possible
May 15 16:19:43 <crashanddie_> ShadowJK: is it an empty can or full can?
May 15 16:22:38 <lardman> yuck, Haar is a learning thing
May 15 16:22:49 <lardman> not a plain-Jane algorithmic solution :)
May 15 16:23:00 <derf> You're asking for a classifier.
May 15 16:23:05 <derf> Of course it's a learning thing.
May 15 16:23:11 <lardman> yeah, not my field
May 15 16:23:21 <derf> It's (one of) mine.
May 15 16:23:36 <lardman> how many tuning params are there then?
May 15 16:24:17 <ShadowJK> crashanddie, half-full
May 15 16:24:21 <derf> Presumably, the number of boosting stages, the size of the sample images, and the number of sample images.
May 15 16:24:51 <derf> I'm sure one could add more, but I don't think any more are necessary.
May 15 16:26:45 <lardman> glad to hear it ;)
May 15 16:28:43 <AStorm> derf, and mine too :)
May 15 16:29:45 <derf> AStorm: So why haven't you written the 30 lines of code yet?
May 15 16:31:16 <lardman> where's that chap who wanted an STFT on the DSP I wonder?
May 15 16:31:40 <AStorm> hmm
May 15 16:31:41 <AStorm> in which repo can I get gcc for the tablet?
May 15 16:31:50 <lardman> poke lcuk
May 15 16:31:57 <AStorm> derf, yes, but not compiled for the tablet
May 15 16:32:01 <AStorm> and especially not packaged
May 15 16:32:02 <lardman> probably the sdk one
May 15 16:32:12 <lbt> AStorm: Mer .... standard.
May 15 16:32:13 <AStorm> lardman, for diablo? link/url please
May 15 16:32:22 <AStorm> lbt, ... Maemo.
May 15 16:32:31 <AStorm> for Mer, it's in the default
May 15 16:32:34 <lbt> ah, there's you're problem right there.
May 15 16:32:36 <lardman> don't have one off hand, sorry
May 15 16:32:57 <AStorm> lbt, but target isn't a non-existent release-quality Mer
May 15 16:33:04 <AStorm> ;P
May 15 16:33:12 <AStorm> that will happen, but much later
May 15 16:33:22 <lbt> it's an approximation to release-quality...
May 15 16:33:49 <AStorm> uhm, no.
May 15 16:34:32 <AStorm> just as linear function is an approximation of an exponential one ;P
May 15 16:39:44 <lcuk> yeah lardman
May 15 16:40:17 <lardman> wine?
May 15 16:40:26 <lardman> or the location of gcc for on-tablet dev>
May 15 16:40:27 <lardman> ?
May 15 16:40:27 <Jaffa> Nearly.
May 15 16:40:38 <lcuk> deb http://repository.maemo.org/ diablo/sdk free non-free
May 15 16:40:44 <lardman> Jaffa: how's things?
May 15 16:40:52 <lardman> Jaffa: I have a glass on the go, wonderful :)
May 15 16:40:59 <Jaffa> lardman: first day back at work today - working from home sorting email
May 15 16:41:00 <lcuk> same repository scratchbox uses
May 15 16:41:09 <lardman> lcuk: thanks on AStorm's behalf
May 15 16:41:09 <AStorm> lcuk, does it contain armel (non-cross) compiler?
May 15 16:41:20 <Jaffa> Got a glass of a Normandy aperitif before curry & polishing off last night's bottle of red
May 15 16:41:31 <lcuk> yes astorm i run everything directly on tablet after using standard apt-get stuffs
May 15 16:41:33 <lardman> mmm, curry
May 15 16:41:34 <AStorm> because AFAICS Scratchbox still uses one
May 15 16:41:36 <AStorm> mhm :)
May 15 16:41:40 <AStorm> checking
May 15 16:41:50 <AStorm> whoops, not enough disk space I guess
May 15 16:41:57 <AStorm> time to drop that OpenTyrian ;P
May 15 16:42:01 <lcuk> plenty if you dont waste it
May 15 16:42:10 * lardman needs to push through the slightly dizzy faze and get onto some coding
May 15 16:42:11 <lcuk> tho i ran out the other night
May 15 16:42:24 <AStorm> lcuk, on 256 MB flash? :P
May 15 16:42:27 <lcuk> more wine
May 15 16:42:30 <lardman> +1
May 15 16:42:41 <lcuk> AStorm, yeah theres enough to grab lots of things
May 15 16:42:45 <AStorm> vim is fairly huge for example
May 15 16:42:54 <lcuk> i wouldnt know lol
May 15 16:43:19 <lardman> use vi you coward! ;)
May 15 16:43:25 <AStorm> abiword is large too
May 15 16:43:39 <crashanddie_> timeless_mbp: ping, Yves replied to your email
May 15 16:43:41 <AStorm> lardman, sorry, need proper multifile editor
May 15 16:43:58 <lardman> AStorm: I know, was only taking the mick
May 15 16:46:12 <timeless_mbp> ok, i'm getting dressed and hunting for my cell phone
May 15 16:48:58 * timeless_mbp ponders
May 15 16:49:24 <AStorm> gcc = No Matches Found... argh
May 15 16:49:34 <AStorm> just added the SDK repo
May 15 16:50:31 <timeless_mbp> crashanddie_: ok, this might be complicated
May 15 16:51:05 <lcuk> AStorm,
May 15 16:51:08 <lcuk> Nokia-N810-23-14:~# apt-cache showpkg gcc
May 15 16:51:08 <lcuk> Package: gcc
May 15 16:51:08 <lcuk> Versions:
May 15 16:51:08 <lcuk> 4:3.4.4-7osso2 (/var/lib/apt/lists/repository.maemo.org_dists_diablo_sdk_free_binary-armel_Packages) (/var/lib/dpkg/status)
May 15 16:51:08 <lcuk>  
May 15 16:51:22 * lardman wanders off to watch some TV
May 15 16:51:25 <lbt> fiferboy_away: done
May 15 16:51:44 <AStorm> lcuk, check again the repo, hmmh
May 15 16:51:46 * timeless_mbp can't see any way to get there by mass transit w/o needing an hour
May 15 16:51:51 <AStorm> maybe it didn't dl correctly or sth
May 15 16:52:34 <AStorm> ok, darned redpill ;)
May 15 16:59:27 <fiferboy> lbt: You got it?
May 15 17:01:47 <lbt> yes
May 15 17:02:46 <lbt> fiferboy... http://pastebin.com/d5cf63d55
May 15 17:03:06 <lbt> want the .so?
May 15 17:03:49 <fiferboy> lbt: I sure do!
May 15 17:11:26 <AStorm> darn
May 15 17:11:31 <AStorm> python2.5-dev is huge :/
May 15 17:11:39 <AStorm> to heck ;)
May 15 17:11:53 <Stskeeps> anyone who happen to know what it means when NetworkManager is greying out APs? :P
May 15 17:11:57 <Stskeeps>  / SSIDs
May 15 17:13:00 <thopiekar> Stskeeps: greying out == leaving out?
May 15 17:13:26 <Stskeeps> no, it shows them but they're grey
May 15 17:13:52 <thopiekar> I've never seen that :P
May 15 17:21:35 <AStorm> hidden networks?
May 15 17:24:06 <AStorm> neat, now I need libc6-dev and stuff ;P
May 15 17:24:34 <AStorm> weird that python2.5-dev doesn't dep on these
May 15 17:24:38 <AStorm> it should
May 15 17:28:03 <AStorm> now I'm a proud owner of pyprocessing :>
May 15 17:36:54 <AStorm> and now multiprocessing too :)
May 15 17:37:50 <RST38h> mulpyprocessing?
May 15 17:38:07 <AStorm> ;) almost
May 15 17:38:21 <AStorm> multiprocessing module backported from python 2.6.1.1
May 15 17:38:46 <AStorm> (same as in 2.6.2)
May 15 17:39:06 <AStorm> doesn't need any weird javascript it seems - a bug in debian package it seems :)
May 15 17:39:17 <GAN800> RST38h, the 5800 is pretty nice but damn the browser sucks.
May 15 17:39:28 <AStorm> and it seems that debian maintainers don't test their packages either
May 15 17:40:22 <AStorm> when jeremiah shows back, tell him that - the exact same settings will work for python-multiprocessing as for python-processing :)
May 15 17:51:17 <mikkov_> anybody have a clue how big is the flash where rootfs resides is going to be in rx-51?
May 15 17:52:37 <konttori_> sure, but cannot tell
May 15 17:52:42 <konttori_> how so?
May 15 17:54:13 <Stskeeps> mikkov_: theres going to be a large immc source code claims.. no idea on nand
May 15 17:54:44 <RST38h> mikkov: there is just two possibilities
May 15 17:55:09 <RST38h> mikkov: it is either 256MB or 512MB
May 15 17:55:48 <ShadowJK> not 128? :)
May 15 17:55:58 <RST38h> He asked about flash
May 15 17:56:04 <ShadowJK> yes
May 15 17:56:14 <RST38h> Samsung makes 128 SDRAM + 256 flash and 256 SDRAM + 512 flash
May 15 17:57:08 <ShadowJK> I'm more curious if it's going to be jffs2 or one of the newer things :)
May 15 17:57:37 <RST38h> Shadow: Not very important to me
May 15 17:57:48 <lcuk> do we know it the rx-51 has a data crystal reader?
May 15 17:57:51 <RST38h> I suppose whatever Nokia chooses it will be something sensible
May 15 17:58:11 <RST38h> lcuk: Yes, and you lubricate it with blood
May 15 17:58:43 <ShadowJK> some would argue jffs2 isn't sensible even for n8x0 :)
May 15 17:58:52 <ShadowJK> but the replacements weren't really ready yet
May 15 17:58:55 <RST38h> it works
May 15 17:59:03 <fiferboy> lbt: :O
May 15 17:59:14 <RST38h> I definitely want real file system on the internal mmc though
May 15 17:59:27 <RST38h> One that lets me install software
May 15 17:59:34 <lbt> fiferboy: how does it look...
May 15 17:59:37 <fiferboy> lbt: I have a meeting right now, but wow!
May 15 17:59:42 <lbt> it's 'exagerated'
May 15 17:59:51 <fiferboy> I like the new VMAX too
May 15 18:00:05 <fiferboy> Wow, very nice.  I'll give you more feedback later
May 15 18:00:11 <lbt> can you do video?
May 15 18:00:16 <lbt> my camera died :((
May 15 18:00:29 <fiferboy_away> My camera is dead too!  Battery cacked
May 15 18:00:39 <fiferboy_away> Gotta go, talk later
May 15 18:02:10 <crashanddie_> lbt: camera on n810 died some time ago as well, opened it up, and the data cable was disconnected
May 15 18:02:17 <crashanddie_> lbt: plugged it back in, and works fine
May 15 18:02:38 <lbt> I meant my canon ;)
May 15 18:02:44 <lbt> I want to video the Qt bounce
May 15 18:02:50 <lbt> I suppose a mirror....
May 15 18:03:01 <lcuk> heh
May 15 18:03:21 <lbt> vnc2avi ?
May 15 18:03:30 <ShadowJK> RST38h, even with real filesystem on it you'd have to mount it in the rootfs somewhere?
May 15 18:03:32 <lcuk> 2slow
May 15 18:03:50 <lbt> food... back in an hour
May 15 18:04:34 <RST38h> Shadow: Guess so
May 15 18:04:55 <ShadowJK> so like one step less than using loop?
May 15 18:05:14 <ShadowJK> or like, reformat the mmc yourself, heh
May 15 18:05:25 <RST38h> Shadow: I dunno what you mean by using loop, but the current OS does not let me run stuff from mmcs
May 15 18:06:33 <mikkov_> konttori_: I was just thinking if big data files (games) should still be installed to mmc
May 15 18:07:36 <ShadowJK> oh, it mounts it noexec :)
May 15 18:08:27 <RST38h> Shadow: there does not seem to be any security reason to mount internal mmc as noexec, so I guess there is some other reason
May 15 18:08:27 <konttori_> mikkov_: whatever the size of one nand, you would want to install big data files on emmc.
May 15 18:08:51 <konttori_> anything of tens or hundreds of megabytes should go to emmc.
May 15 18:09:18 <mikkov_> ok, then it's pretty small as in N8x0 :)
May 15 18:09:26 <RST38h> mikkov: there is
May 15 18:09:33 <RST38h> small and limited
May 15 18:09:39 <RST38h> 512 is small, 256 is limited
May 15 18:10:01 <mikkov_> /media/mmc2 is still the mount point?
May 15 18:10:01 <konttori_> Having exec code on a partition that gets mouted to PC is not a good idea.
May 15 18:10:31 <konttori_> mikkov_: for what?
May 15 18:10:44 <lcuk> you say that like pc is a security risk lol
May 15 18:11:01 <Stskeeps> mikkov_: ke-recv src is open
May 15 18:11:09 <mikkov_> konttori_: for emmc where I want to install my big files
May 15 18:11:21 <konttori_> yeah.
May 15 18:11:39 <RST38h> konttori: if you mean windows virii, tablet won't run 'em :)
May 15 18:12:31 <ShadowJK> RST38h, quick, make up theory about why he said that, maybe there's wine+qemu-userspace
May 15 18:13:47 <RST38h> Shadow: it is possible to write a custom pc virus that will copy maemo-specific payload when the NIT is mounted to PC
May 15 18:14:00 <RST38h> Shadow: but I personally thing it is a scenario not worthy of consideration
May 15 18:14:05 <lcuk> or a scripted virus that uses python or something..
May 15 18:14:50 <lcuk> surely though, its the way it is because a decision was made in the past
May 15 18:15:06 <ShadowJK> jokes aside, noexec on removable media is just common sense and good practice :)
May 15 18:15:11 <lcuk> if anyone thinks that decision  needs changing surely an official bug can be made
May 15 18:15:11 <RST38h> One reason is that if you are really sneaky you will give victim a data file causing NIT app to overflow a buffer and execute the payload
May 15 18:15:22 <RST38h> Shadow: Removable yes, dev/mmc2 no.
May 15 18:16:09 <lcuk> but from the system standpoint the mountpoint mmc2 used to be external and removable didnt it
May 15 18:16:43 <mikkov_> mmc1 is external
May 15 18:16:59 <ShadowJK> I think if you boot N810 without external card, it swaps them around or something..
May 15 18:17:10 <lcuk> on n800 mmc1 and mmc2 are both external
May 15 18:17:34 <mikkov_> lcuk: but mmc2 is under battery cover
May 15 18:17:39 <lcuk> so?
May 15 18:17:47 <lcuk> its removable
May 15 18:17:54 <mikkov_> so pretty much internal to me :)
May 15 18:17:57 <lcuk> and either way any of these can be reformatted to something better
May 15 18:18:17 <RST38h> mikkov: do not offend n800 users :)
May 15 18:18:18 <lcuk> i have my mmc1 which was fat as ext2 and i almost did same with mmc2
May 15 18:18:40 <RST38h> they consider it external and will kill you for telling them it is not really necessary to have two cards
May 15 18:19:13 <lcuk> simplified ui :) http://www.lolpix.com/pictures/18/Funny_Pictures_548.htm
May 15 18:19:20 <ShadowJK> please don't bunch us N800+N810 owners together with the N800-only owners
May 15 18:19:21 <ShadowJK> ;)
May 15 18:19:53 <mikkov_> UI calls it also internal
May 15 18:20:43 <lcuk> and its non executable unless  you change something
May 15 18:20:47 <lcuk> but it can be changed
May 15 18:30:43 <RST38h> "It should be noted that no ethically-trained software engineer would ever consent to write a DestroyBaghdad procedure. Basic professional ethics would instead require him to write a DestroyCity procedure, to which Baghdad could be given as a parameter."
May 15 18:31:40 <RST38h> (and then discover the wonders of forall() statement =))
May 15 18:32:22 <Jaffa> :)
May 15 18:57:31 <tds5016> hey all. Is there an easy way to get the mac address with python?
May 15 19:03:23 <fiferboy> lbt: You back?
May 15 19:09:56 <fiferboy> lcuk: You can regrab lbt's overshoot when scrolling horizontally
May 15 19:10:11 <lcuk> ?
May 15 19:10:16 <lcuk> i dont use overshoot lol
May 15 19:10:23 <lcuk> ahhhh
May 15 19:10:26 <lcuk> i see what you mean
May 15 19:10:34 <lcuk> cool then!
May 15 19:11:18 <fiferboy> But it doesn't come back, so you can keep sending it off to the left
May 15 19:11:33 <lcuk> :D excellent
May 15 19:11:49 <lcuk> someone in ML today asked about grabbing a horizontal movable frame which had a text box on it, but the textbox ate the event
May 15 19:11:55 <fiferboy> For the record, he just hasn't implemented the rubber band effect horizontally
May 15 19:12:02 <lcuk> you have to wonder why people want to mix their metaphors
May 15 19:17:12 <lcuk> fiferboy, can you re-energize your scrolling whilst its going, like the little impulses you give to a roundabout to keep it at speed
May 15 19:17:56 <fiferboy> I'll pretend I know what a roundabout is for the purposes of your similie
May 15 19:19:16 <lcuk> http://www.flickr.com/photos/ilike/10662072/
May 15 19:19:33 <lcuk> kids stand on it, mates push them round, friends puke
May 15 19:19:42 <lcuk> much hilarity :)
May 15 19:21:40 <fiferboy> Ah, I don't beleive I have ever heard it called that, but I can't recall what we used to call them.
May 15 19:21:52 <fiferboy> I haven't seen one at a playground here in many year.
May 15 19:22:03 <fiferboy> Too dangerous.  Kids might have fun and fall down
May 15 19:23:46 <lcuk> lol
May 15 19:36:26 <RST38h> fiferboy: What is this country where they banished roundabouts?
May 15 19:36:49 <fiferboy> Canada.  At least, I haven't seen any here in Ontario since I was a kid.
May 15 19:36:50 <Stskeeps> insert random european country
May 15 19:37:04 * Stskeeps doesn't believe in cotton-coating kids
May 15 19:37:04 <t_s_o> http://www.linuxdevices.com/news/NS9676037801.html <- how long until running on maemo? ;)
May 15 19:37:10 <lardman|tv> RST38h: roundabouts are something children shouldn't play on
May 15 19:37:22 <fiferboy> They also don't generally have wooden wingsets and climbers any more.
May 15 19:37:34 <fiferboy> /s/wingsets/swingsets
May 15 19:38:08 <lardman|tv> lcuk: do they not have roundabouts in the North?
May 15 19:38:22 <Stskeeps> lardman|tv: saw the smartq7 unboxing pics earlier?
May 15 19:38:35 <lardman|tv> Stskeeps: new ones? Saw something the other day
May 15 19:38:35 <lcuk> no, they removed then after the chavs were using em with motorbike power
May 15 19:38:59 <lardman|tv> lcuk: I tend to go around them with 45 degrees bank angle... in a car
May 15 19:39:02 <Stskeeps> "lardman|afk: http://www.jiongtang.com/blog/html/digital-product/smartq7-unboxing.html btw"
May 15 19:39:06 <Stskeeps> (earlier today)
May 15 19:39:18 <lardman|tv> thanks
May 15 19:39:19 <lcuk> lol lardman|tv different kind of roundabout
May 15 19:39:39 <RST38h> fiferboy: Oh, the horror
May 15 19:39:45 <lardman|tv> the other child's toy has another names doesn't it?
May 15 19:39:53 <RST38h> lardman: roundabouts are a useful tool of natural selection
May 15 19:40:07 <lardman|tv> RST38h: indeed :)
May 15 19:40:15 * RST38h suggests renaming roundabout "a Darwin's Gate"
May 15 19:40:17 <fiferboy> lardman|tv: Sometimes they are called Merry-go-rounds, according to wikipedia
May 15 19:40:20 <lardman|tv> they need bog spikes in the middle rather than grass
May 15 19:40:25 <lardman|tv> s/bog/big
May 15 19:40:34 <t_s_o> ye freaking mcalendarsync runaway...
May 15 19:40:43 <lardman|tv> fiferboy: a merry-g-round is larger and powered
May 15 19:41:08 <fiferboy> That's what I thought too, but apparently some people call the little ones that too
May 15 19:41:12 <lardman|tv> and often has horses and the like that go up and down
May 15 19:41:15 <RST38h> lardman: See, Canadians abolish roundabouts and the next thing they have to do is place warning stickers on everything, down to toothbrushes
May 15 19:41:30 <lardman|tv> :)
May 15 19:41:35 <fiferboy> RST38h: In English and French
May 15 19:41:41 <RST38h> fiferboy: Oh.
May 15 19:41:44 * RST38h almost forgot
May 15 19:42:00 <suihkulokki> wtf is a chav?
May 15 19:42:03 <lardman|tv> lcuk: aren;t they called spinny-pushy-thingy-ma-bobs?
May 15 19:42:19 <lardman|tv> http://en.wikipedia.org/wiki/Chav
May 15 19:42:25 <lcuk> "you comin to the park to play on the swings and slide and roundabout" ..
May 15 19:42:40 <lardman|tv> as I said, the North... ;)
May 15 19:42:41 <fiferboy> http://en.wikipedia.org/wiki/Roundabout_(play)
May 15 19:42:52 <fiferboy> I love the picture of the homemade roundabout in that article!
May 15 19:42:58 <tds5016> anyone know of an easy way to get the mac address with python?
May 15 19:43:04 <lardman|tv> yeah, well I bet it was an American who made that entry
May 15 19:43:12 <RST38h> lardman: both the large and small kinds are called "carousel" in Russian (fancy French word for roundabout I guess)
May 15 19:43:35 <lardman|tv> not that I have anything against you guys, except that you don't know the names of children's playthings in England
May 15 19:43:48 <lcuk> http://hobsons-bay-leader.whereilive.com.au/news/story/security-guards-watch-over-point-cook-playground/
May 15 19:43:54 <fiferboy> Funny that the wikipedia article doesn't mention any roundabout-related deaths.
May 15 19:43:57 * RST38h has never been to England, so he has got an excuse
May 15 19:44:04 <lcuk> look at photo3 and you see why its needed to be guarded, but WHY did they allow it in the first place
May 15 19:44:07 <lardman|tv> RST38h: carousel would be more like a merry-go-round in my book, but perhaps without the horses
May 15 19:44:10 <RST38h> fiferboy: Maybe there weren't any?
May 15 19:44:13 <Stskeeps> fiferboy: because if you go fast enough you fly into orbit!
May 15 19:44:22 <lcuk> drill into the ground
May 15 19:44:28 <fiferboy> Escape velocity!
May 15 19:44:34 <RST38h> No bodies.
May 15 19:44:42 <lcuk> i really like the "roundabout playpump" idea
May 15 19:44:48 <lcuk> thats clever
May 15 19:44:52 * RST38h proud of making such a practical remark
May 15 19:44:53 <fiferboy> But what if there was a rubber band stretched above the roundabout and at least two other forces acting on the body?
May 15 19:45:07 <lardman|tv> perhaps it is a merry-gop-round
May 15 19:45:13 <lardman|tv> aargh, can't type
May 15 19:45:15 <RST38h> fiferboy: children would either remove or tear it.
May 15 19:45:23 <RST38h> fiferboy: next question.
May 15 19:45:42 * lardman|tv goes back to the TV and wine, still not reached the perfect programming saturation point
May 15 19:45:47 <fiferboy> How many skinned knees were suffered at the hands of the roundabouts?
May 15 19:46:02 <RST38h> fiferboy: In fact, given enough time, they will pull the whole thing out of the ground too
May 15 19:46:11 <lcuk> not that many, but banged heads and puke cleanup was high
May 15 19:46:41 * RST38h does not get the "was" part. These things are still in every second courtyard around here
May 15 19:47:17 <lcuk> anyway fiferboy now we all know what a roundabout is, can you give your kinetic list a little burst of energy every now and then to keep the list upto speed
May 15 19:47:35 <fiferboy> lcuk:  Oh, is that what you were getting at?  Yes.
May 15 19:47:36 <lcuk> like you can with a roundabout.. :P
May 15 19:47:38 <fiferboy> You can.
May 15 19:47:56 <lcuk> ok good lol
May 15 19:48:00 <fiferboy> lcuk:  I would have gone with a tetherball methaphor.
May 15 19:48:05 <lcuk> i just saw the code to reset motion when you start a stroke
May 15 19:48:28 <RST38h> Will any improvement in Tear kinetic scrolling come out of this BTW?
May 15 19:48:35 <lcuk> 3 hours to explain "tetherball"?
May 15 19:48:57 <fiferboy> lcuk: Well, I guess if you get down to it, when you touch the screen while it is scrolling the widget will start to follow your finger again
May 15 19:49:13 <lcuk> RST38h, i would hope with all this discussion from all parties (no matter which framework) we will have a workable set of nice listviews in next device
May 15 19:49:15 <fiferboy> But for all practical purposes, if you are fast enough to start flicking it will accelerate again.
May 15 19:49:47 <RST38h> lcuk: Only if Nokia accepts the code.
May 15 19:49:49 <lcuk> yeah i like to "strum" at the screen to keep it scrolling
May 15 19:50:01 <lcuk> its relaxing and i can just change direction again with inverted stroke
May 15 19:50:09 <fiferboy> RST38h: lbt's work won't directly affect Tear, but if the developers want to look at the code and implement something similar they are free to.
May 15 19:50:15 <lcuk> RST38h, this is just discussion for everyone :)
May 15 19:50:21 <lbt> hi
May 15 19:50:30 <fiferboy> lbt: Hi, send you an email
May 15 19:50:32 <lcuk> getting it into the minds of people is a good thing (TM)
May 15 19:51:06 <lcuk> hi lbt \o
May 15 19:51:24 <lcuk> RST38h, tear is webkit on gtk isnt it
May 15 19:51:44 <RST38h> yea
May 15 19:52:04 <RST38h> fiferboy: Somebody has to tell Bundyo then
May 15 19:52:07 <lbt> oops .. just looking at the backlog
May 15 19:52:20 <fiferboy> lbt: It is mostly about old playground equipment
May 15 19:52:27 * RST38h has got a practical question
May 15 19:52:48 <lbt> so it seems
May 15 19:52:59 <RST38h> I am running a program taking the whole system memory (32MB) and need to invoke amixer from it
May 15 19:53:08 <lcuk> lbt, which was your favorite from the playground
May 15 19:53:22 <lbt> swings
May 15 19:53:23 <RST38h> There is a little bit of memory left for amixer, but Linux misses spawn*() functions
May 15 19:53:29 <lbt> you could jump off
May 15 19:53:34 <fiferboy> lbt: We still have those here1
May 15 19:53:37 <lbt> or try and go over the top
May 15 19:53:38 <lbt> cool
May 15 19:53:42 <fiferboy> lbt: Oh, but you aren't allowed to jump off...
May 15 19:53:48 <lcuk> indeed
May 15 19:53:49 <lbt> safety-bars?
May 15 19:53:53 <lbt> sigh
May 15 19:54:00 <Stskeeps> RST38h: dlopen it? :P
May 15 19:54:04 <RST38h> And to do exec() I need to do fork() first and fork() fails because it is out of memory
May 15 19:54:06 <fiferboy> No, I think you are just discouraged from it.
May 15 19:54:11 <lcuk> lbt, we used to do this twisting thing and end up standing horizontally
May 15 19:54:11 <RST38h> Stskeeps: an executable???
May 15 19:54:13 <lbt> oh, sod that then!
May 15 19:54:19 <lcuk> RST38h, no more memory == free some first
May 15 19:54:24 <RST38h> lcuk: Can't
May 15 19:54:34 <Stskeeps> RST38h: it can be PIC/-shared and executable at same time, yeah
May 15 19:54:37 <lcuk> if its full then you have to find a way
May 15 19:54:45 <Stskeeps> RST38h: maemo-launcher uses this trick extensively
May 15 19:54:57 <lcuk> you could try syncing with a wristwatch and using its memory as expansion
May 15 19:55:46 <RST38h> Sts: hmmm...I doubt it will work =(
May 15 19:56:03 <Stskeeps> RST38h: why not call alsa directly then? :P
May 15 19:58:00 <RST38h> Sts: If I knew how to control ALSA DAC volume, I would
May 15 19:58:34 <lcuk> RST38h, i thought we had more than 32m anyway
May 15 19:58:41 <RST38h> But ALSA has got super messy APIs
May 15 19:58:43 <Stskeeps> RST38h: http://alsa.opensrc.org/index.php/HowTo_access_a_mixer_control
May 15 19:58:56 <RST38h> Sts: !!!!
May 15 19:59:16 <lbt> fiferboy: thanks for the comments...I suspect it's mainl because of how I cheat the value()
May 15 19:59:27 <RST38h> Sts: Except that some of these functions seem to crash on this Linux =)
May 15 19:59:38 <Stskeeps> RST38h: how can you expect amixer to work then? :P
May 15 19:59:46 <fiferboy> lbt: Do you think a setOvershoot function would solve some of the problems?
May 15 20:00:30 <lbt> mmm sadly I think each scrollwidget needs to implement outOfBounds scrolling
May 15 20:00:46 <RST38h> Sts: amixer works somehow, probably using some other calls (there are lots to choose from)
May 15 20:00:52 <lbt> so QScrollArea and freinds
May 15 20:01:19 <fiferboy> lbt: Why can't it be in qabstractslider?
May 15 20:01:49 <Stskeeps> RST38h: do i dare to ask what kind of machine you're doing it on? :P
May 15 20:02:01 <lbt> well, I could just put a function there... but it wouldn't fix the other issues
May 15 20:02:51 <lbt> ok, lets do setOvershoot() and work from there
May 15 20:03:21 <fiferboy> If you don't update the srollbar position in setOvershoot, just the value, that may help
May 15 20:03:36 <RST38h> Stskeeps: STMP3770 SoC from SigmaTel
May 15 20:03:59 <RST38h> Stskeeps: It actually has 64MB RAM installed but I can't make the second half work unfortunately
May 15 20:04:27 <lbt> will try that... hmm, setValueOvershoot() I think
May 15 20:04:49 <RST38h> Stskeeps: Running EA Linux that is kinda half-baked and not supported by SigmaTel's new owner
May 15 20:09:49 * Stskeeps should really try out this OMAP2/maemo-dsp pulseaudio thing
May 15 20:16:05 <lbt> Stskeeps: or fix Numpy Physics in Mer
May 15 20:16:11 <lardman|tv> Stskeeps: you got a link for that?
May 15 20:16:48 <Stskeeps> lardman|tv: http://gitorious.org/~elmarco/pulseaudio/pa-work/commit/d92f79b8efe591a4dc184ed803859bcdf7173b46
May 15 20:17:33 <lardman|tv> cool
May 15 20:17:38 <lardman|tv> sound in Mer then
May 15 20:18:07 <Stskeeps> yeah
May 15 20:19:00 <lardman|tv> and some more clues as to the existing DSP tasks
May 15 20:20:33 <guerby> hi, is there anything new on the rumoured next nokia internet tablet since http://thenokiablog.com/2008/09/20/next-nokia-internet-tablet-some-specs-revealed/
May 15 20:20:35 <guerby> ?
May 15 20:20:47 <guerby> I haven't followed nokia forums lately
May 15 20:21:05 <lardman|tv> old news I guess, nothing new since the alpha SDK was released afaik
May 15 20:22:16 <fiferboy> lbt: I'm heading home now.  If you need something tested send it my way
May 15 20:22:16 <guerby> lardman|tv, thx
May 15 20:22:36 <lbt> fiferboy: willdo  - ta
May 15 20:22:49 <guerby> lardman|tv, I assume it should be out any time now with nokia pricing n810 a 249 euros
May 15 20:23:05 <guerby> and the sdk being out
May 15 20:23:30 <fiferboy> Have a good weekend everyone
May 15 20:24:56 <lardman|tv> guerby: we live in hope!
May 15 20:28:47 <biiter_> plz.. where do i turnoff that bar showing at the bottom indicating keyboard input case?
May 15 20:44:45 <biiter_> got it
May 15 20:45:20 * RST38h would like to find ALSA people and kill them slowly
May 15 20:45:50 <RST38h> Except that would probably require scaling an ivory tower first
May 15 20:47:29 <jaska> :D
May 15 21:31:50 <oli> is there any n800 stuff for reading qrcodes?
May 15 21:41:56 * RST38h fixed the volume problem
May 15 21:42:21 <RST38h> [at the cost of severe mental degradation]
May 15 21:43:12 <derf> oli: Yes.
May 15 21:43:30 <derf> But right now that "stuff" is just a test program from maemo-barcode.
May 15 21:43:52 <lcuk> gsoc project for it i think
May 15 21:44:02 <derf> Really? Awesome.
May 15 21:44:52 <fireun> mmm, mental degradation
May 15 21:52:32 <lardman|tv> oli: derf's qrcode stuff in maemo-barcode
May 15 21:52:46 <lardman|tv> ah, too late
May 15 21:52:55 * lardman|tv should read the scroll between
May 15 21:55:03 <RST38h> <sleep>
May 15 22:49:42 <timeless_mbp> crashanddie: all went well
May 15 22:49:49 <timeless_mbp> thanks for helping make arrangements
May 15 23:27:00 <GeneralAntilles> Did I miss anything fun or exciting today?
May 15 23:27:31 <lcuk> a 3 hour discussion on roundabouts
May 15 23:42:59 <crashanddie> timeless, ping me when you're around
May 16 00:52:25 <Marques> it's hard port a app to the maemo?
May 16 00:54:45 <Marques> it is hard to port a program to the maemo?
May 16 00:55:17 <`0660> i think setting up the sdk is the biggest hurdle :)
May 16 00:56:45 <Marques> but how i setup the, i need a linux on the desktop
May 16 00:57:15 <Marques> how i setup the sdk?Needs a Linux Distro on the desktop?
May 16 00:57:19 <Marques> a LiveCD works
May 16 00:57:22 <Marques> ?
May 16 00:57:47 <`0660> you could use vmware image
May 16 01:00:07 <Marques> yes but the port needs know the programming language right, like the python, C#, and the others?(sorry my bad english, i am brazillian)
May 16 01:01:09 <ShadowJK> yes
May 16 01:03:31 <Marques> oh... so i don't go port any program, i don't know any programing language...
May 16 01:05:59 <Marques> it is much difficulty learn a programing language?
May 16 01:07:18 <lcuk> the language is the easy part, it helps if you explain what the app is currently written in and for what language
May 16 01:07:27 <lcuk> some ports from existing linux apps are easy
May 16 01:07:49 <lcuk> for other languages and operating systems you are in for much more work and its not trivial
May 16 01:09:31 <Marques> and users of maemo, the KDE port is a window manager only or it change the maemo to other OS?(I am noob on the linux)
May 16 01:13:04 <lcuk> that doesnt make much sense,   what is the app you are trying to port, what os does it currently work on and what language does it use, and if possible which widget toolkit is it using..
May 16 01:17:01 <Marques> no i don't want port any program, i am only curious, and like know if KDE is only a windows manager and runs the programs of maemo or KDE is OS and the programs of Maemo don't work on the KDE
May 16 01:22:34 <lcuk> maemo uses a different window manager, but the core gtk code works with extra specific maemoisms
May 16 01:22:57 <lcuk> so a gtk program that can work in kde or gnome can be adjusted to work in maemo
May 16 01:23:02 <lcuk> and vice versa
May 16 01:24:06 <Marques> but it is automatic or easy?Because i see that the KDE have the OpenOffice, konqueror, GIMP
May 16 01:24:20 <Marques> and the ubuntu anybody have tested?
May 16 01:25:46 <pupnik> yes
May 16 01:26:04 <pupnik> ask only yes/no questions Marques
May 16 01:26:10 <pupnik> makes things easier for me
May 16 01:26:23 <lcuk> Marques, a large complex program does not run well in a constrained screen :)
May 16 01:26:39 <lcuk> gnite tho
May 16 01:26:42 <lcuk> nite pupnik
May 16 01:27:36 <Marques> but the maemo have a openoffice alternative?
May 16 01:40:22 <soap> vi
May 16 01:45:03 <GeneralAntilles> Abiword
May 16 01:45:06 <Guest77195> .nick jumentous
May 16 01:50:16 <jumentou> can someone tell me the basics of what parts of maemo are open source and what is closed? essentially can i use maemo on top of an omap based device as i dont need any communication stack
May 16 01:52:05 <ShadowJK> it's not like maemo has much of a communication stack either...
May 16 01:52:18 <jumentou> i know that hildon obviously linux etc are all open
May 16 01:52:28 <ShadowJK> You probably want to check out Mer
May 16 01:52:38 <GAN800> ~mer
May 16 01:52:39 <infobot> somebody said mer was http://wiki.maemo.org/Mer
May 16 01:53:06 <jumentou> was thinking of going with ubuntu mid and hildon, but it could be more difficult if it becomes arm based
May 16 02:18:14 <chx> After years of eyeing the tablets, when buy.com priced it at 199 i could not resist any more. I have my own BT GPS receiver, so that's not a problem. Is there a public transit planner for the maps application it ships with or do I need google maps (which require online connction) for that
May 16 02:19:18 <ShadowJK> i woudlnt expect much from the bundled wayfinder gps app
May 16 02:20:00 <Macer> anybody here have a ps3 with linux on it?
May 16 02:24:58 <GAN800> jumentou, MID is probably moving to Mer upstream anyway
May 16 02:26:15 <jumentou> thats interesting, i haven't seen a lot of development recently on MID edition though i'd heard it heading more towards moblin
May 16 02:27:11 <pupnik> do those voltage muscle stimulators work
May 16 02:27:16 <pupnik> my MID could use some toning
May 16 02:27:31 <GAN800> It was based on Moblin, now it's probably moving to Mer
May 16 02:28:02 <GAN800> Canonical is being stupid and shortsighted about the MID range I think.
May 16 02:28:32 <GAN800> pupnik, newsflash: sweat is very conductive.
May 16 02:29:20 <jumentou> they seem to be more focused on their UNR, any idea of Mer's stability etc, i'm looking at it now but just what you know
May 16 02:29:37 <GAN800> jumentou,
May 16 02:29:57 <GAN800> jumentou, it's alpha stage
May 16 02:30:13 <GAN800> Devel is very unstable at this point
May 16 02:30:34 <GAN800> Its upstream is only at a beta release, so. . . .
May 16 02:30:46 <jumentou> yea assumed so
May 16 02:31:52 <jumentou> as far as i see though using MID edition really rules me out of using an OMAP chip
May 16 02:33:06 <GAN800> Meh, it's only vaguely Atom targeted. But, really, Mer is your best bet in the mobile Linux game at this point.
May 16 02:35:38 <jumentou> its only vaguely atom targetted but the repositorys and prebuilt are all x86 targetted, i could possibly build it all myself but its a lot of work
May 16 02:41:25 <chx> ShadowJK: i see. are there other apps? paid is fine.
May 16 02:41:40 <chx> also. is it possible to get a VGA out for projectors? :)
May 16 02:41:58 <ShadowJK> heh, some people have used usb-vga adapters
May 16 02:42:18 <ShadowJK> (combined with f-f usb adapter and powered usb hub I'd think)
May 16 02:42:41 <ShadowJK> the bundled wayfinder is actually not free to use
May 16 02:42:54 <ShadowJK> well, only navigation costs
May 16 02:43:30 <chx> LOL http://xsisusb.garage.maemo.org/seriously?
May 16 02:43:42 <chx> erm , space before the last word :p
May 16 02:43:54 <chx> thats quite something from a 200 bucks device.
May 16 02:44:46 <ShadowJK> maemo-mapper isnt that great withoyt internet connectivity
May 16 02:45:08 <ShadowJK> I've seen navit mentioned in the forums but I don't know anything about it
May 16 02:52:16 <chx> more questions... just tell me if i run out of questions quota. did anyone try to bolt a fingerprint reader ? i tried to google
**** ENDING LOGGING AT Sat May 16 02:59:57 2009