JogAmp IRC Chat Logs

#jogamp @ irc.freenode.net - 20130917 06:19:13 (UTC)

20130917 06:19:13 -jogamp- Previous @ http://jogamp.org/log/irc/jogamp_20130917050612.html
20130917 06:19:19 * jogamp (~jogamp@anon) has joined #jogamp
20130917 06:19:19 * Topic is 'http://jogamp.org | Hacking 3D Graphics, Multimedia and Processing across Devices'
20130917 06:19:19 * Set by rmk0 on 20130116 23:58:04
20130917 06:19:49 -NickServ- This nickname is registered. Please choose a different nickname, or identify via /msg NickServ identify <password>.
20130917 06:19:49 -NickServ- You are now identified for jogamp.
20130917 06:52:40 * monsieur_max (~maxime@anon) has joined #jogamp
20130917 12:01:01 <xranby> sgothel: is this JOCL JNI approach right? https://jogamp.org/bugzilla/show_bug.cgi?id=773#c3
20130917 12:02:35 <sgothel> yes
20130917 12:03:33 <xranby> awesome, hoe we can recruit a new JOCL maintainer :)
20130917 12:03:36 <xranby> jope
20130917 12:03:37 <xranby> hope
20130917 12:03:47 <sgothel> :)
20130917 12:23:41 <xranby> feels good to have finnaly written down the notes collected from LA highway discussion/airplane code investigation into a bugreport comment
20130917 12:23:52 <sgothel> :)
20130917 12:25:34 <xranby> i need to re-flash my cellphone, i think it got compromised
20130917 12:25:58 <xranby> i recieved a chain of configuration SMS from the phone number "Collector" thus send from my ISP.
20130917 12:26:10 <xranby> the phone consumed the SMS messages and removed the sms'es
20130917 12:26:22 <xranby> i will refuse to use my phone untill i have reprogammed it
20130917 12:26:25 <sgothel> fascinating
20130917 12:26:40 <monsieur_max> ( and scary )
20130917 12:26:55 <dfj> :/
20130917 12:26:58 <xranby> ISP tech told me that they are unable to help me unless i file a police investigation in order to find the origin on the sms
20130917 12:27:09 <xranby> they do claim they did not send them
20130917 12:27:21 <dfj> that is not 'no worries, that was our update'
20130917 12:27:21 <xranby> which i call bullshit
20130917 12:27:31 <xranby> no i purcased the phone separately
20130917 12:27:38 <xranby> they have no ownership of it
20130917 12:27:47 <dfj> ugh.
20130917 12:27:59 <sgothel> right .. the sms config stuff .. you told us in Annaheim .. wonder how this could be .. err .. restricted, i.e. authorized by user
20130917 12:28:08 * dfj goes with cheap, useless phones.
20130917 12:28:30 <xranby> i will dump the phones userland config into a file and investigate
20130917 12:28:54 <xranby> thanks to /dev/mmcblk* on linux + root :)
20130917 12:29:06 <sgothel> so if th police has no interest in these details, they may already know how it works :)
20130917 12:29:26 <xranby> well i got to have some proof before i start the investigation
20130917 12:29:50 <xranby> if i can get the sms log and possibly the contents of the sms i may be in better possition
20130917 12:29:59 <xranby> that they have been sent
20130917 12:30:34 <dfj> so - in teh expected case, it is the provider pushing a patch, and overriding your setting not to receive it.
20130917 12:30:37 <xranby> i have at least "screenshots" from this "Collector" phone number
20130917 12:30:42 <xranby> of this
20130917 12:30:48 <sgothel> it's all b/c they know that you do hack iphones :)
20130917 12:30:50 <xranby> the phone number is alphanumeric
20130917 12:30:55 <sgothel> haha
20130917 12:31:09 <xranby> i dont hack iphones
20130917 12:31:24 <sgothel> thats what you said .. but they asked you :)
20130917 12:31:25 <xranby> i only hack my nokia that got a do you want to become root gui switch
20130917 12:31:31 <sgothel> kidding .. dude :)
20130917 12:31:59 <sgothel> remember .. 'you jailbreak iphones?'
20130917 12:32:17 <xranby> yes,, haha they did ask me that
20130917 12:32:27 <xranby> before passing the border
20130917 12:33:19 <xranby> time to test android later on the nokia n9
20130917 12:33:38 <xranby> i belive it is possible to dual boot the phone
20130917 12:33:43 <sgothel> http://stackoverflow.com/questions/2404907/mobile-phone-configuration-via-sms
20130917 12:34:09 <sgothel> 3GPP's OTA ..
20130917 12:35:16 <xranby> right.. so if the representative of my ISP say they did not send these sms they are lying
20130917 12:35:46 <sgothel> waiiittt ... hacking and faking to be a base station is quite 'common' .. sadly
20130917 12:36:01 <xranby> sure.. that is a possibility as well
20130917 12:36:13 <xranby> i keep it in mind while looking for proof
20130917 12:37:21 <sgothel> plus the phone crypto is very weak usually .. some simple xor stuff .. and so forth, guess Bruce Schneier was writing about it
20130917 12:37:36 <xranby> also true
20130917 12:37:55 <xranby> i have requested my isp to send the call logs next month
20130917 12:38:12 <xranby> that should include sms recieved as well
20130917 12:38:41 <sgothel> 'good thinking' hacking phones now, where folks start to store their bank transaction number there .. etc etc .. hmm. I prefer the old paper one-time-pin numbers.
20130917 12:39:20 <dfj> iirc - if yer crypto is really excellent, xor-ing with a stream is acceptable - but from mobile providers, this is teh furthest thing from the case.
20130917 12:41:14 * xranby is inspecting the phone rootfs filesystem in "airplane mode"
20130917 12:41:23 <xranby> with sim ejected
20130917 12:42:55 <dfj> g'lucks.
20130917 12:43:28 * dfj stubles off to attempt reviving last years code - currently in 'balck screen' bugged state. :/
20130917 12:43:29 <xranby> generating images of the phones /root and user partition
20130917 12:43:49 <xranby> dd if=/dev/mmcblk0p3 of=user bs=1024k
20130917 12:44:24 <xranby> that will create a 1:1 copy of the user partition
20130917 12:44:48 <xranby> written to the user data that i can extract from the phone using a usb connection
20130917 12:45:31 <xranby> really nice that they have dd in the phones root shell
20130917 12:46:25 <xranby> without having the aegis software protection kick in and killing my process
20130917 12:46:47 <xranby> aegis is nokias security manager running on the phone designed to prevent snooping inside some files
20130917 12:48:20 <dfj> them beasts have any analog of a jtag header?
20130917 12:48:35 <xranby> it surely have if i remove the phone screen
20130917 12:48:45 <dfj> cool.
20130917 12:49:26 <dfj> I've found them sprayed around on the test pads, rather than exposed for after-test access.
20130917 12:56:37 <xranby> http://youtu.be/9dxsTGpGRhs?t=2m40s this is how my phone looks inside
20130917 12:56:51 <xranby> i have unfortunally not found any good guide to the board layout
20130917 12:57:05 <xranby> it looks to be a small connector white in the middle
20130917 13:07:20 <dfj> those flat ribbons and connectors are awesome - if nly they weren't so bloody expensive.
20130917 13:10:52 <xranby> http://youtu.be/i9u7r4j0atU?t=7m42s - i think i see the jtag! next by the micro usb socket and the earpiece, needs pressure pin contacts
20130917 13:11:08 <xranby> 2x3 pins
20130917 13:12:13 <xranby> there is also 1x3 pins at different place of the main board
20130917 13:12:41 <xranby> and 1x5 pins next to the memory flash chip
20130917 13:12:54 <dfj> well, there is your last resort. :)
20130917 13:14:00 <xranby> .. some people are calm enough to do these tricky disassembly recovery projects
20130917 13:14:03 <dfj> here's hoping the main firmware is poorly locked down and you can pull it from flash - the flash chip, presumably, is doable, unless encrypted.
20130917 13:14:27 <dfj> s|calm|angry|
20130917 13:14:33 <xranby> haha
20130917 13:14:50 <xranby> jtag-rage
20130917 13:17:04 <dfj> modding a working device is rarely enough to motivate me... fixing one that I define to be 'broke' in some way...
20130917 13:49:43 <xranby> i got proof! after extracting all strings from the root filesystem strings root > strings-root i am able to detect traces of these "Collector" messages. in plaintext xml
20130917 13:51:02 <xranby> <entry name="Collector" mtime="1379407964" type="string"> <- including unix timestamp :)
20130917 13:51:32 <xranby> that is Tue, 17 Sep 2013 08:52:44 GMT
20130917 13:51:55 <xranby> thats the time of the consumed message
20130917 13:53:21 <xranby> i have similar xml fragments for other sms on my phone and there name is replaced by sms numbers
20130917 13:57:23 <dfj> ! grats, progress. :)
20130917 13:59:36 <xranby> the timestamp matches exactly where i want it, i got the sms'es between10.40 to 11.05 GMT+2
20130917 14:00:27 <dfj> do you have the contents?
20130917 14:01:23 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20130917 14:02:38 <xranby> i need to check in the file with an hex editor before the strings extraction
20130917 14:04:43 <xranby> i can see that i recived 12 sms total
20130917 14:04:53 <xranby> by having 12 timetamped entrys
20130917 14:10:43 <dfj> here's hoping the phone reconstructs a patch from them sms, then checks the signature at the end before applying it.
20130917 14:10:58 * dfj livesn in a fantasyland with elves, fairies, unicorns, and the like.
20130917 14:16:36 * monsieur_max (~maxime@anon) has joined #jogamp
20130917 14:18:23 <xranby> the file i found with the timestamps looks like a configuration file for the sms gtk application used on the nokia n9. unfortunally the file do not contain the sms contents , cold case
20130917 14:18:56 <xranby> i dont know if it is possible to extract deleted sms from a sim card
20130917 14:19:25 <xranby> all i know is the timestamps when i recived the sms messages
20130917 14:19:35 <xranby> at least better than nothing
20130917 14:20:42 <xranby> i think i need some tool http://www.dekart.com/howto/howto_sim_reader/how_to_recover_a_deleted_sms/
20130917 14:21:15 <sgothel> if you could access the sim card's flash .. recover filesystem .. and not storing anything until then ..
20130917 14:21:59 <xranby> i have copys of the phone flash.. i have the simcard ejected next to the phone
20130917 14:22:30 <xranby> i found the timestamps of the deleted sms inside the 1:1 root filesystem copy
20130917 14:23:06 <xranby> right now i would like to have a sim card reader
20130917 14:23:23 <xranby> and look at the deleted sms slots
20130917 14:32:01 * xranby unbox a siemens gsm devboard from 2008
20130917 14:32:15 <xranby> funny what you have lying in the closet
20130917 14:33:11 <xranby> last time i used this devboard was when i tred to add GSM sms message support to a blood ananlyser
20130917 14:34:56 <monsieur_max> haha what a hacker :)
20130917 14:35:14 <monsieur_max> "this sms confirm you that you have human blood"
20130917 14:54:12 <sgothel> http://semiaccurate.com/2013/09/05/vivante-has-a-high-precision-mobile-gpu/ <- Vivante story from Charlie ..
20130917 15:01:52 <xranby> sgothel: btw.. penberg one of the gnu classpath hackers now work on a solution to provide java running inside a secure by design sandbox
20130917 15:02:07 <xranby> https://github.com/cloudius-systems http://osv.io/ https://t.co/I0RLjaIT9T OSv lets java run secure by design inside a virtual machine hypervisor
20130917 15:02:17 <xranby> you can then make a secure by design icedtea-web
20130917 15:02:30 <sgothel> sounds nice indeed
20130917 15:07:52 <sgothel> hmm .. a whole hypervisor .. I was more thinking like simple uid/process separation :)
20130917 15:12:04 <sgothel> http://youilabs.com/blog/mobile-gpu-floating-point-accuracy-variances/
20130917 15:23:08 <xranby> https://gist.github.com/xranby/6186418edecad0467a27#file-openjdk-17-sept-2013-log-L24
20130917 15:24:49 <xranby> i will update the lgo if mr replies
20130917 15:24:51 <xranby> the log
20130917 15:58:32 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20130917 16:34:38 * monsieur_max (~maxime@anon) has joined #jogamp
20130917 17:11:55 <xranby> http://www.open-zfs.org/wiki/Announcement - OpenZFS launch announcement, September 17th, 2013 "Today we announce OpenZFS: the truly open source successor to the ZFS project. "
20130917 17:24:53 <rmk0> ♥
20130917 18:02:01 <sgothel> so seems like the 'official' community of what has been done in recent years .. w/ same folks - great they created a 'place'
20130917 18:19:45 <sgothel> @Xerxes: "OSv lacks support for multiple address spaces – which boots its performance – and instead of mandating separation between applications, the kernel uses the JVM to block accesses to kernel memory. These two methods free up resources for app performance" from http://www.theregister.co.uk/2013/09/17/cloudius_systems_osv_cloud_software/ - sound very much pretty _nasty_ to me!
20130917 18:20:32 <xranby> one jvm in each hypervisor hmm
20130917 18:20:39 <xranby> i have not seen that
20130917 18:21:36 <xranby> ok i have now done my part, jamvm update and testing for the next icedtea 2.4.2 release sheduled to go out tomorrow
20130917 18:21:44 <sgothel> I would be happy to see process/memory separation working well for a dedicated jvm-uid
20130917 18:21:58 <sgothel> kudos
20130917 18:32:35 <xranby> btw, i have recieved green light by aph who now run the openjdk 6 project to merge mjw's patch that remove the restricted crypto hooks now that oracle no longer runs the project
20130917 18:33:42 * void256 (~chatzilla@anon) has joined #jogamp
20130917 18:36:47 <xranby> https://gist.github.com/xranby/75da4098f1fbe535d080
20130917 18:38:02 <xranby> if we build OpenJDK 8 in Ji Gong we probably want to have a similar patch in place
20130917 18:47:59 <xranby> void256: cool shader rendered lines!
20130917 18:48:20 <xranby> have you checked if the paper you based the code on are patent free?
20130917 18:49:26 <void256> hi xranby ... good question. I've not checked. how do you check that?
20130917 18:50:51 <xranby> sgothel: dow do one know is this work http://www.cg.tuwien.ac.at/research/publications/2012/ROESSLER-2012-OGLES/ is patent free?
20130917 18:51:02 <xranby> how
20130917 18:51:10 <void256> line rendering was only a small part of his work tho and it's not like a whole fully fledged algorithm that I took from it
20130917 18:51:24 <void256> basically just some initial ideas and the rest was my work
20130917 18:51:45 <xranby> sgothel: void256 used it to do: http://imgur.com/h4pb93x
20130917 18:52:37 <void256> I now have miter line joins working which I worked out all by myself :)
20130917 18:53:17 <void256> next is bevel and round line joins
20130917 18:53:33 <void256> which I'm struggeling with a bit
20130917 18:54:25 <xranby> void256: have you seen the work done inside GLG2D, maybe you and brandon are rolving the same challanges?
20130917 18:54:38 <xranby> http://brandonborkholder.github.io/glg2d/
20130917 18:56:05 <xranby> void256: all in all i like the phase and entusiasm you have while putting the nifty gui code in shape
20130917 18:56:55 <void256> I like Nifty GUI 2.0 a lot ... unfortunately it's not much yet :)
20130917 18:57:09 <void256> haven't seen GLG2D yet but will look at it
20130917 18:57:44 <void256> at first glance it's doing too much of what I have in mind
20130917 18:57:56 <xranby> brandon is a maven fan as well :)
20130917 19:00:36 <void256> figuring things out by myself is half the fun too =)
20130917 19:06:08 <sgothel> Hi Jens .. I will look into it sure. Q: Using graph is out of the question for lines and curves ?
20130917 19:06:39 <sgothel> @Xerxes: Sure .. please add the patch-o-matic to Ji Gong git repo
20130917 19:06:55 <sgothel> .. sure I will look at the impl. if it rings a bell .. (lines)
20130917 19:08:52 <sgothel> well, if it is a 'proper' thesis it should not reuse patent ideas :) however, if this is 'just' for masters .. nobody checks novelty ..
20130917 19:10:30 <xranby> you may talk to the cacao jvm engineers who have graduated from wien of their university maters tesis practises
20130917 19:10:58 <sgothel> hu?
20130917 19:12:02 <xranby> the CACAO JVM project is originating from the same university.. http://mips.complang.tuwien.ac.at/hg/cacao/ the developers are found in #classpath & #openjdk
20130917 19:12:11 <sgothel> well .. our beloved distance function in action .. I would say :)
20130917 19:12:26 <xranby> http://www.cg.tuwien.ac.at/research/publications/2012/ROESSLER-2012-OGLES/
20130917 19:12:43 <sgothel> and how is this related Xerxes ?
20130917 19:12:57 <xranby> only my brain can tell i am afraid :)
20130917 19:13:16 <xranby> my idea was
20130917 19:13:42 <xranby> that people who have graduated from the university could tell if the university allow master tesis to use patented ideas
20130917 19:14:14 <sgothel> I would say .. only Loop, Blinn had the nerves to patent curved AA GPU rendering (probably due to MS involved) .. so this should be Ok, since it doesn't involve quadratics or .. err .. cubic curves
20130917 19:14:36 <sgothel> what makes my mouth watery now is - we could use this for graph for straight line rendering!
20130917 19:14:46 <sgothel> since this part still uses MSAA !
20130917 19:14:56 <sgothel> @Jens: Can we collab on this ?
20130917 19:15:19 <sgothel> seems to close a weakness of our graph implementation
20130917 19:15:59 <xranby> void256: ^_^
20130917 19:16:07 <sgothel> @Mark: You still don't have the t-shirt ? Jens (void) got it already .. hve he has it already AFAIK.
20130917 19:17:03 <sgothel> so when I go ahead to cleanup our graph stuff .. I will for sure consider these referenced papers [10][50][18]
20130917 19:17:23 <rmk0> sgothel: no sign of it yet, no
20130917 19:17:24 <sgothel> have to ask Qun to get a copy of the originals :)
20130917 19:17:34 <rmk0> mail isn't exactly super fast here
20130917 19:17:35 <sgothel> @Mark: We sent them as a 'letter' ..
20130917 19:17:39 <sgothel> hmm
20130917 19:17:45 <sgothel> a little box ..
20130917 19:18:05 <rmk0> hm... how long's it been now?
20130917 19:18:19 <sgothel> let me check whether I have the receipt ..
20130917 19:22:11 <sgothel> 2 weeks I guess
20130917 19:23:33 <rmk0> right
20130917 19:23:39 <rmk0> yeah, could still be in the system
20130917 19:23:42 <void256> sgothel: hi sven. collab for sure. just let me finish and clean up the line shaders in the next couple of days maybe weeks. should be simple to reuse
20130917 19:23:51 <void256> :)
20130917 19:23:52 <rmk0> you may've read that the government's trying to destroy the public mail system here in order to privatise it
20130917 19:24:27 <rmk0> two weeks from germany sent by ordinary mail probably isn't surprising
20130917 19:24:32 <sgothel> that would be awesome Jens .. then we may get a chance to tackle http://jogamp.org/wiki/index.php/SW_Tracking_Report_Objectives_for_the_release_2.1.0#Graph_Cleanup_and_UI together ..
20130917 19:24:58 <sgothel> oh .. ok Mark, fine then :) well, I haven't insured those precious items :)
20130917 19:25:04 <rmk0> hehe
20130917 19:26:06 <sgothel> @Jens, if we are lucky .. we can integrate the shader in our curve algo .. simply by dropping the 'bezier' ctrl points
20130917 19:26:41 <sgothel> .. all looks very familiar
20130917 19:27:31 <sgothel> seems to be a great writeup that thesis, i.e. good to dive into shader/gl programming
20130917 19:27:50 <void256> yeah it was pretty nice
20130917 19:27:59 <void256> they don't use edge functions tho
20130917 19:28:42 <sgothel> I would say as a goal for us: migrate the lines into graph .. and try to use graph for nifty as well (which includes curves)
20130917 19:29:09 <sgothel> have you tested SVG as a cheap import format for UI layouts .. ?
20130917 19:29:23 <void256> well, yes, but can someone use graph without jogamp? :)
20130917 19:29:27 <sgothel> I wonder why they didn't use JOGL :)
20130917 19:29:40 <void256> or without JOGL to be precise
20130917 19:29:43 <sgothel> is there a world outside of JOGL ? :)
20130917 19:29:53 <void256> yeah, I guess so ;)
20130917 19:30:01 <sgothel> in theory sure, we could extract it
20130917 19:30:14 <void256> I would be very interested then
20130917 19:30:21 <sgothel> hard dependencies do not really exist to JOGL ..
20130917 19:30:35 <sgothel> sure .. the rendering uses JOGL commands and so forth ..
20130917 19:31:14 <void256> I don't want to be dependent on JOGL too much
20130917 19:31:35 <sgothel> right .. you like to keep compatible w/ LWJGL .. hmm, curious if that really matters so much for the long term. but hey, I stop right now and here :)
20130917 19:31:37 <void256> I still view it as one rendering backend
20130917 19:31:47 <void256> hehe ;)
20130917 19:32:20 <sgothel> the only really inconvenience would be the used math utils .. since I don't like duplication
20130917 19:32:33 <sgothel> but for sure .. we could expose interfaces ..
20130917 19:33:01 <sgothel> and somebody would need to maintain .. so IMHO best after aforementioned cleanup
20130917 19:33:02 <void256> the text rendering looks attractive for sure =)
20130917 19:33:10 <void256> and curves of course
20130917 19:33:50 <sgothel> the truetype font stuff .. can be 'ejected' or simply used-as is
20130917 19:33:58 <void256> but for Nifty and where it is at the moment I really want to keep things simple and see where it will go
20130917 19:33:59 <sgothel> (rebundled)
20130917 19:34:21 <void256> can it be used with lwjgl?
20130917 19:34:33 <sgothel> sure .. to follow the Graph UI goals .. I might need your expertise .. and code
20130917 19:34:56 <sgothel> all can be used w/ LWJGL sure .. we just need to see how to rebundle etc ..
20130917 19:35:09 <sgothel> ttf font is standalone
20130917 19:35:19 <sgothel> and most of graph is GL agnostic
20130917 19:35:35 <sgothel> hence we even have an opengl subpackage
20130917 19:36:15 <void256> so you're saying I can download a ttf font jar somewhere and use it as is? o_O
20130917 19:36:29 <sgothel> you would need to repackage it ..
20130917 19:36:36 <sgothel> i.e. extract it from jogl-core.jar
20130917 19:36:37 <void256> aye
20130917 19:36:48 <sgothel> we can also add it as an atomic ..
20130917 19:36:57 <sgothel> for your convenience
20130917 19:37:14 <void256> :>
20130917 19:37:27 <sgothel> same w/ graph .. up until a point .. (jogl / gl stuff)
20130917 19:37:36 <void256> of course
20130917 19:37:52 <sgothel> we have 2 math packages .. jogl/math and graph/math .. that might be a problem
20130917 19:38:15 <sgothel> so .. just tedious work - but license allows it ofc
20130917 19:38:54 <void256> yeah
20130917 19:39:17 <void256> probably easier to use it as is ;)
20130917 19:39:44 <void256> or find someone who would extract and maintain it as a separate project
20130917 19:40:48 <void256> like the opengl ttf font project with adapters to jogl/lwjgl
20130917 19:40:54 <void256> :)
20130917 19:41:59 <sgothel> sure, that is a viewpoint I could understand as a non jogamp guy :) for jogamp'
20130917 19:42:12 <sgothel> cause .. it's for sure an exceptional feature :)
20130917 19:43:50 <void256> actually I started something like that for bitmap fonts already
20130917 19:44:13 <void256> https://github.com/void256/jglfont
20130917 19:44:14 <sgothel> yeah, we have that right, but who likes bitmapped fonts
20130917 19:44:55 <void256> ;)
20130917 20:50:59 * void256_ (~chatzilla@anon) has joined #jogamp
20130917 20:53:02 * void256 (~chatzilla@anon) Quit (Ping timeout: 264 seconds)
20130917 20:53:10 * void256_ is now known as void256
20130917 21:42:40 * void256 (~chatzilla@anon) Quit (Remote host closed the connection)
20130917 22:32:54 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20130918 00:42:19 * hharrison (~chatzilla@anon) has joined #jogamp
20130918 00:43:11 <hharrison> sgothel: I always have a soft spot for math libs, and would consider doing a bit of graph/math and jogl/math common work
20130918 00:43:39 <hharrison> If you find time to put any thoughts down, I'd be very interested
20130918 00:43:55 <sgothel> will do while working on the graph/UI things ..
20130918 00:44:06 <sgothel> thank you .. yeah, much appreciated
20130918 00:44:33 <sgothel> but before .. also need to do the osx 7u40 stuff :)
20130918 00:44:37 <hharrison> Do you have a survey off the top of your head where the diff. math libs are currently?...package names would be fine
20130918 00:45:14 <hharrison> Yeah, waiting on a debug dump from a few testers to try and correlate the java3d and jogl state, try to figure out where it is going wrong
20130918 00:45:58 <sgothel> com.jogamp.opengl.math .. hmm right .. cleaned it up last time
20130918 00:46:40 <sgothel> com.jogamp.graph.geom <- this one is special to graph
20130918 00:47:01 <sgothel> com.jogamp.opengl.math.** it is
20130918 00:47:15 <sgothel> so pretty much isolated now
20130918 00:47:23 <hharrison> THoughts on the final destination? com.jogamp.math perhaps?
20130918 00:47:41 <sgothel> good question ..
20130918 00:47:49 <sgothel> i.e. do we desire to do so ?
20130918 00:47:58 <hharrison> com.jogamp.vecmath ;-)
20130918 00:48:17 <sgothel> so gluegen's runtime is com.jogamp.common.**
20130918 00:48:20 <sgothel> hehe
20130918 00:48:50 <hharrison> com.jogamp.common.math really isn't that bad I think
20130918 00:48:54 <sgothel> iff we do so, it's solely for things .. hmm .. reusing parts of our stuff (like graph)
20130918 00:49:14 <hharrison> Sorry, have to run
20130918 00:49:18 * hharrison (~chatzilla@anon) Quit (Quit: ChatZilla 0.9.90.1 [Firefox 24.0/20130909203154])
20130918 00:49:20 <sgothel> ay
20130918 03:59:10 * sgothel (~sven@anon) Quit (Quit: Leaving.)
20130918 07:01:55 * monsieur_max (~maxime@anon) has joined #jogamp
20130918 07:47:23 <xranby> OK lets make a jogamp demo of this: https://www.simonsfoundation.org/quanta/20130917-a-jewel-at-the-heart-of-quantum-physics/
20130918 07:47:38 <xranby> A geometry model is found for quantum physics interactions!
20130918 07:48:44 <xranby> it would be sweet to get Dominik involved
20130918 07:49:39 <xranby> it may also fit into the common.math package :)
20130918 07:51:13 <xranby> “You can easily do, on paper, computations that were infeasible even with a computer before.”
20130918 07:52:23 <xranby> Rami take a look at this!
20130918 08:01:01 * jk4 (~jk4@anon) Quit (Ping timeout: 263 seconds)
20130918 08:01:25 * jk4 (~jk4@anon) has joined #jogamp
20130918 08:01:37 * jk4 (~jk4@anon) Quit (Changing host)
20130918 08:01:37 * jk4 (~jk4@anon) has joined #jogamp
20130918 10:44:46 <xranby> Jaroslav Trnka: The Amplituhedron: http://www.staff.science.uu.nl/~tonge105/igst13/Trnka.pdf
20130918 10:45:16 <xranby> "This rich structure is also completely new to the mathematicians."
20130918 10:57:02 <xranby> Nima Arkani-Hamed: The Amplituhedron: http://susy2013.ictp.it/video/05_Friday/2013_08_30_Arkani-Hamed_4-3.html
20130918 11:19:34 <xranby> fast forward 20min 13 sec into Nima Arkani-Hamed's talk and then the Amplituhedron is explained in a way that you can understand!
20130918 11:34:04 <xranby> 27min 21sec "you give me the external momentum of the gluons, that just tells me where to draw these points, i draw this amplituhedron (triangulate), i compute the volume and thats the amplitude. Thats hundered of pages of feynman diagrams and its not just that this answer is simple, we known for a while now that the answer is simple, the remarkable new thing we know, we know a different question the volume is the answer to."
20130918 13:03:13 * sgothel (~sven@anon) has joined #jogamp
20130918 13:03:13 * ChanServ sets mode +v sgothel
20130918 13:18:55 * xranby (~xranby@anon) Quit (Quit: Leaving.)
20130918 13:34:41 * xranby (~xranby@anon) has joined #jogamp
20130918 13:43:49 <xranby> sgothel: a new mathematical structure is found , that work as a model for quantum mecanical interactions, i fyou read in the backlogs :)
20130918 13:44:12 <xranby> The Amplituhedron
20130918 13:45:53 <sgothel> oh ho .. will read later, thx :)
20130918 14:32:36 * xranby (~xranby@anon) Quit (Quit: Leaving.)
20130918 14:33:29 * xranby (~xranby@anon) has joined #jogamp
20130918 15:08:03 * xranby (~xranby@anon) Quit (Ping timeout: 245 seconds)
20130918 15:09:40 * xranby (~xranby@anon) has joined #jogamp
20130918 15:09:50 * xranby (~xranby@anon) has left #jogamp
20130918 15:58:33 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20130918 16:53:58 * monsieur_max (~maxime@anon) has joined #jogamp
20130918 19:47:33 * void256 (~chatzilla@anon) has joined #jogamp
20130918 20:06:23 * jk4 (~jk4@anon) Quit (Ping timeout: 245 seconds)
20130918 20:12:45 * jk4 (~jk4@anon) has joined #jogamp
20130918 21:31:48 * xranby-64 (~familjen@anon) has joined #jogamp
20130918 21:32:40 * xranby-64 is now known as xranby64
20130918 21:36:02 <xranby64> The post about the Amplituhedronen included a lot of concepts, mostly known to LHC particle physicists, such as "Scattering Amplitudes" and "Positive Grassmannian" on many occasions. Here is a talk where these concepts are explained using illustrations and math and how they relate to "on shell" classical physics. https://cast.itunes.uni-muenchen.de/vod/clips/ifNaivs8Pf/flash.html
20130918 21:38:06 <xranby64> for a 3D graphics hacker, the concepts of Grasmanian, actually allows you to do vertex cloud -> faces so this all is quite good to learn as cross dicipline
20130918 21:38:44 <xranby64> grassmannian
20130918 21:43:26 <xranby64> and allows you to plot nice 3d patterns
20130918 21:45:24 <xranby64> like this one: https://lh6.googleusercontent.com/-BHYc2VJV5cc/UNveePZmlWI/AAAAAAAAGAA/Qaeiupw2TBY/s1600/nima-permutation-grassmannian-final-picture.PNG this is a grassmanian permutation plotted in 2D, representing a #D structure
20130918 21:48:14 <xranby64> ... and the paper behind "Scattering Amplitudesand" the "Positive Grassmannian" is here: http://arxiv.org/pdf/1212.5605v1.pdf
20130918 21:50:54 <xranby64> how this all relates to 3D graphics is touched in chapter 5
20130918 21:55:36 <xranby64> ... and then pysicists wondered where all these black and white dots came from and finally found the answer in the new Amplituhedronen geometry... https://www.simonsfoundation.org/quanta/20130917-a-jewel-at-the-heart-of-quantum-physics/ which looks cool!
20130918 21:56:19 <xranby64> the dots tells you do to left or right turn inside these diagrams to create permutations
20130918 21:56:25 <xranby64> if i understood it ..
20130918 22:01:42 <sgothel> holy moly .. and I fiddle w/ AWT 2d transformation and printing :)
20130918 22:02:45 <xranby64> oh and now physicist are on fire because by using this new Amplituhedron they can calculate the "Scattering Amplitude" by simply placing points based on gluon data and calculating the volume of this Amplituhedron
20130918 22:04:07 <xranby64> thus reducing 500 pages of feynman calculations to something that can be teached in highschool
20130918 22:04:29 <xranby64> "A sketch of the amplituhedron representing an 8-gluon particle interaction. Using Feynman diagrams, the same calculation would take roughly 500 pages of algebra." --> https://www.simonsfoundation.org/wp-content/uploads/2013/09/amplituhedron-drawing_web-271x300.jpg
20130918 22:04:44 <xranby64> now fits on a napkin
20130918 22:04:50 <sgothel> you are telling me .. that this could solve 'point clouds' (like scanned w/ some laser ..) .. i.e. creating triangles ? :)
20130918 22:05:15 <sgothel> and .. btw .. what kind of drugs do you take when you grok all this ? gimme some of those pls :)
20130918 22:05:20 * monsieur_max (~maxime@anon) has left #jogamp
20130918 22:06:22 <sgothel> nice napkin :) probably only takes years to understand for me :)
20130918 22:06:59 <xranby64> yes i got that impression see page 33 http://wwwth.mpp.mpg.de/members/strings/strings2012/strings_files/program/Talks/Tuesday/Arkani-Hamed.pdf
20130918 22:08:26 <xranby64> so they have invented methods to traverse point clouds and locate faces
20130918 22:08:59 <xranby64> they have of course only tested this on their physics quantum mecanical graphs
20130918 22:09:23 <xranby64> point cloud where the lines are transition between quantum physics particles
20130918 22:09:41 <xranby64> and the vertex points are the elemental particles
20130918 22:10:34 <sgothel> nice to see arvix and other _public_ media is being used to publish it
20130918 22:11:04 <xranby64> yes i am really happy this nice new research is available online instantly
20130918 22:11:38 <xranby64> the drug i take is simply try to understand the words he say inside the presentations by watching more presentations
20130918 22:12:00 <sgothel> hehe .. which eventually will blow your mind ..
20130918 22:15:23 <xranby64> hopefully the mind will be forced to fuse ideas in order to fit it all inside the head :D
20130918 22:16:14 <xranby64> and thus the result of the packed mind become new insight and knowledge.... or simply awesome new dreams
20130918 22:19:14 <sgothel> hmm .. here the language itself is already blocking me off .. semantics of their variables etc
20130918 22:30:58 <sgothel> The Amplituhedron: http://susy2013.ictp.it/video/05_Friday/2013_08_30_Arkani-Hamed_4-3.html @ 20min 13s .. yes, now we are talking :)
20130918 22:50:33 <void256> >_<
20130918 22:51:39 <void256> nuts
20130918 22:51:42 <xranby64> sgothel: yes.. sounds like a fragment shader :D
20130918 22:51:58 <xranby64> but he is talking in physics terms
20130918 22:52:15 <xranby64> and calculation of weight instead of color
20130918 22:54:05 <sgothel> guess time for my midnight dinner .. will check back later .. or good night - thx for sharing Xerxes, always nice to enlighten my old brain
20130918 22:54:47 <xranby64> sgothel: thank you for taking a look! enjoy your meal
20130918 23:01:11 <xranby64> physics talk for loops can be related to projection in computer graphics
20130918 23:01:33 <xranby64> the loop allows hiding particles
20130918 23:01:46 <xranby64> thus not taking some particles into the calculation
20130918 23:01:54 <xranby64> thus a projection
20130918 23:03:01 * xranby64 maybe should create a wordlist for use by 3d graphics engineers who want to talk to LCH particle physicists
20130918 23:03:22 <xranby64> LHC
20130918 23:10:59 * void256 (~chatzilla@anon) Quit (Quit: ChatZilla 0.9.90.1 [Firefox 23.0.1/20130814063812])
20130919 00:16:54 * xranby64 (~familjen@anon) Quit (Ping timeout: 264 seconds)