#jogamp @ irc.freenode.net - 20150824 05:05:31 (UTC)
20150824 05:05:31 -jogamp- Previous @ http://jogamp.org/log/irc/jogamp_20150823050531.html
20150824 05:05:31 -jogamp- This channel is logged @ http://jogamp.org/log/irc/jogamp_20150824050531.html
20150824 05:09:19 * monsieur_max (~maxime@anon) has joined #jogamp
20150824 06:06:05 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20150824 06:30:22 * elect (~elect@anon) has joined #jogamp
20150824 06:30:34 <elect> hi
20150824 07:16:44 * monsieur_max (~maxime@anon) has joined #jogamp
20150824 07:19:36 * eclesia (~husky@anon) has joined #jogamp
20150824 07:19:41 <eclesia> good morning
20150824 07:20:21 <monsieur_max> hello
20150824 09:14:41 <elect> a11mad11, what's your problem? Are you using one interleaved vbo, right?
20150824 09:49:39 <xranby> monsieur_max: can you elaborate what kind of urls that do not work with the media player?
20150824 09:49:56 <xranby> (11:42:35) monsieur_max: hi there
20150824 09:49:56 <xranby> (11:44:18) monsieur_max: I was trying to use he video player with a custom url ( and all the needed classes to handle a custom scheme )
20150824 09:49:56 <xranby> (11:44:33) monsieur_max: no luck so far, i suppose the native part is not able to call my custom handlers
20150824 09:50:14 <monsieur_max> xranby: i'd like to use a custom url like "mycustomscheme:/myfile.mp4"
20150824 09:50:48 <zubzub> how are you using the custom url?
20150824 09:50:55 <zubzub> to you wrap it in a java url?
20150824 09:50:57 <zubzub> object
20150824 09:50:58 <zubzub> ?
20150824 09:51:03 <monsieur_max> in Java, you can register handlers for such protocols
20150824 09:51:08 <zubzub> idd
20150824 09:51:19 <zubzub> but it depends how your classloaders are used
20150824 09:51:25 <zubzub> eg in osgi it's a nightmare
20150824 09:51:26 <xranby> hmm... can you alter the mediaplayer junit tests to demonstrate this usecase
20150824 09:51:34 <monsieur_max> zubzub: the idea is that i want my files to be accessed in different ways
20150824 09:51:34 <xranby> sounds like an enhancement bug in jogamp
20150824 09:51:58 <xranby> we only support what the junit tests demonstrate
20150824 09:52:09 <zubzub> monsieur_max: do you use some kind of framework?
20150824 09:52:12 <zubzub> or a plain java app?
20150824 09:52:21 <monsieur_max> zubzub: plain
20150824 09:52:41 <zubzub> and is the url passed as a string or as an url object that you create?
20150824 09:52:45 <monsieur_max> the registering works, it's just that the mediaplayer is not able to understand such a scheme
20150824 09:52:57 <zubzub> ok so it's the mediaplayers fault :p
20150824 09:53:03 <monsieur_max> as long as I'm inside my own java code, it's totally fine
20150824 09:53:13 <monsieur_max> zubzub: yes, i was blaming the native part
20150824 09:54:26 <monsieur_max> xranby: i'm afraid there's no real solution there. I was hoping that the streams were opened and read from java
20150824 09:54:44 <zubzub> if they're not than imho that's a design error
20150824 10:04:50 <xranby> zubzub: it is performance related, it is faster to let the video driver open then uri when it unpack the video
20150824 10:06:42 * rmk0_ (~rmk0@anon) has joined #jogamp
20150824 10:08:33 * rmk0 (~rmk0@anon) Quit (Ping timeout: 260 seconds)
20150824 10:09:28 <xranby> for example if you use ffmpex/libav you pass the uri to sp_avformat_open_input
20150824 10:09:39 <xranby> ffmpeg
20150824 10:10:46 <xranby> but i think ffmpeg also suppor using a callback
20150824 10:10:51 <xranby> that asks for more data
20150824 10:11:46 <xranby> thus using a custom url may be possible but more work has to be done supporting it
20150824 10:15:41 <xranby> http://miphol.com/muse/2014/03/custom-io-with-ffmpeg.html
20150824 10:20:11 <xranby> monsieur_max: the above url contain hints how to implement this at the low level of the ffmpeg backend. jogamp needs an enhancement bugreport with a junit test demonstrating your usecase
20150824 10:20:55 <xranby> by having the junit test we can figure out how to model the api to make it possible
20150824 10:37:13 <eclesia> xranby: I made my first animation \o/ .. but it's a bit verbose.. http://pastebin.com/dfYjA6iq
20150824 10:37:44 <xranby> eclesia: \o/
20150824 10:38:37 <xranby> eclesia: you have all parts to create a video editor :D
20150824 10:38:43 <xranby> with overlay effects
20150824 10:40:04 <xranby> the space node will move back and foth along the x-axis suing this code
20150824 10:41:16 <eclesia> xranby: I'll make a builder to make such animations more friendly. but at least with this one you can do anything. rotation,scale,transalate, multiple keyframe ...
20150824 10:53:08 * eclesia (~husky@anon) Quit (Ping timeout: 246 seconds)
20150824 11:03:09 * eclesia (~husky@anon) has joined #jogamp
20150824 11:27:47 <eclesia> quick question, is it possible to get the device from the GLDrawable ? I'm trying to make a GLDrawable with shared memory space. http://pastebin.com/DNdnbEwG
20150824 11:27:57 <eclesia> or is there something faster in the new rc ?
20150824 11:31:47 <eclesia> looks like I can keep both values to null
20150824 11:57:53 <xranby> sgothel: while debugging on the raspberry pi i have discovered a bug in the newt X11 Window driver https://jogamp.org/bugzilla/show_bug.cgi?id=1178#c8
20150824 11:58:13 <zubzub> just one? :p
20150824 11:58:15 <xranby> after a resize it sends a window moved event with wrong possition
20150824 11:58:20 <xranby> zubzub: yes only one
20150824 11:58:24 <xranby> do you know of any other bug?
20150824 11:58:55 <zubzub> no I don't use newt :p
20150824 11:58:56 <xranby> zubzub: if you do know of any bug then add it to the bugzilla
20150824 11:59:10 <zubzub> already have
20150824 11:59:16 <zubzub> the egl platform thing
20150824 11:59:26 <xranby> is that a bug with the newt X11 driver?
20150824 11:59:46 <zubzub> not with x11
20150824 12:00:12 <zubzub> it's just a newt egl implementation that's err limited
20150824 12:00:17 <xranby> wayland implementation will be looked at after 2.3.2 and is a new enhancement fro 2.3.3
20150824 12:00:34 <zubzub> the egl thing is relevant for both wayland and gbm
20150824 12:00:44 <xranby> zubzub: yes, both are new enhancements
20150824 12:00:56 <xranby> for jogamp to support
20150824 12:01:13 <zubzub> depends if you want to classify it as a bug or enhancement
20150824 12:01:22 <xranby> new usecase == enhancement
20150824 12:01:29 <zubzub> as you can make a gbm or wayland driver
20150824 12:01:43 <zubzub> but they won't work because you depent on limited egl implementation
20150824 12:02:50 <zubzub> (and it tries to instantiate your egl display as a x11 egl display instead of a gbm one)
20150824 12:02:54 <zubzub> *as it
20150824 12:03:26 <zubzub> so you *could* say it's a bug triggered by a new implementation :p
20150824 12:04:42 <xranby> as i see it it can only be an enhancement, the egl extension that may be used to detect what native window type the egl driver have only recently been published thus when we write the egl code the possibility to query it did not exist
20150824 12:06:25 <xranby> also devices shipped with gbm or wayland have only existed for about 2 years.. starting with the jolla phone
20150824 12:08:06 <zubzub> afaik gbm has been around for some time now
20150824 12:08:31 * gouessej (5ee4b442@anon) has joined #jogamp
20150824 12:08:36 <gouessej> Hi
20150824 12:09:07 <gouessej> elect: As this channel is logged, I have to be very careful...
20150824 12:09:34 <elect> :D
20150824 12:09:55 <gouessej> a11mad11: you have to use glBufferData once for the whole VBO and then you can use glBufferSubData
20150824 12:11:03 <elect> just wanted to say that I find his post totally find and legit
20150824 12:11:31 <elect> if he has some doubt about two libs, it should be free to mention those libs
20150824 12:11:49 <xranby> zubzub: do you have any devices shipped with the gbm driver?
20150824 12:12:05 <gouessej> elect: The remark about the competitor was unnecessary and provocative but I admit that the rest was legit
20150824 12:12:06 <elect> I'd agree with you he'd have typed something totally provocative, but it made sense in his speech
20150824 12:12:35 <elect> I'd also like to add that it was a joke and need to be taken as such
20150824 12:12:51 <zubzub> xranby: every device that exists if you limit yourself to simple dumb gbm buffers
20150824 12:12:56 <zubzub> ;)
20150824 12:13:01 <elect> think about it, probabily he'd have never written that if there were no obsession (from you and other) about naming it
20150824 12:13:14 <xranby> zubzub: i was thinking more like with hardware accelerated drivers
20150824 12:13:17 <zubzub> xranby: you probably mean gbm with kms egl support
20150824 12:13:21 <zubzub> yeah :P)
20150824 12:13:34 <zubzub> every driver with kms support then :p
20150824 12:13:50 <elect> I intervened since I always feel the pressure when I have to mention it
20150824 12:13:55 <elect> and I find this illogic
20150824 12:14:01 <elect> ^^
20150824 12:14:02 <xranby> zubzub: i need example hardware
20150824 12:14:02 <gouessej> elect: it was a joke but anyway, it really gets me out of my nerves. In my humble opinion, there are better places for flamewars.
20150824 12:14:08 <zubzub> which I remember started to happen about 9? years ago for the OS amd driver
20150824 12:14:10 <xranby> so that we have devices to test on
20150824 12:14:21 <elect> but it wasnt a flamewars, that's the point you miss, gouessej
20150824 12:14:39 <zubzub> zubzub: arm or is pc also ok?
20150824 12:14:40 <gouessej> elect: You don't have to mention it, the world can live without it.
20150824 12:14:51 <elect> you are a cool guy and I like you, but something you overreact over stupid things
20150824 12:14:53 <zubzub> xranby: *
20150824 12:15:14 <xranby> zubzub: pc is ok as long as they ship in a working configuration
20150824 12:15:19 <elect> look, I think nobody will ever tell me something if I mention jogl on the lwjgl forum
20150824 12:15:22 <elect> if pertinent
20150824 12:15:35 <elect> otherwise if provocative any response will be totally legit
20150824 12:15:40 <zubzub> xranby: afaik the os ati/amd driver has it, so does the intel driver, not so much for nvidia
20150824 12:15:50 <elect> but if I am comparing the two, I find it totally normal naming them
20150824 12:16:12 <gouessej> elect: The maintainers of this library deserve some respect but I'm free not to see them as examples to follow
20150824 12:16:12 <zubzub> xranby: as for arm you'd have to google it, but there are quite a few. Just about every recent one I believe
20150824 12:17:55 <elect> you are totally free to do that, of course, but it was just an example about freedom
20150824 12:18:28 <gouessej> elect: freedom of expression has a meaning in a certain frame, it's not an absolute right
20150824 12:18:33 <elect> sure
20150824 12:18:57 <elect> but given the context, he had all the logic to mention that
20150824 12:19:29 <gouessej> elect: In the past, I had to refrain some people to use our forum to provide their own forks of Java 3D. It's the same with this other library... in a lot worse
20150824 12:19:40 <elect> not the joke, that was his decision, but I guess interventions like yours give a very very bad image of the community
20150824 12:19:53 <gouessej> elect: He could make this kind of joke in private but not in the forum
20150824 12:21:52 <elect> anyway
20150824 12:21:56 <gouessej> elect: If you talked about some past events concerning JOGL on a particular forum whose moderator is in favour of this other library, he would dump your post
20150824 12:22:12 <gouessej> elect: I don't go that far yet
20150824 12:22:42 <elect> I dont know what you mean
20150824 12:22:44 <elect> exactly
20150824 12:22:45 <sgothel> *keep your cool*
20150824 12:22:56 <elect> *trying*
20150824 12:23:02 <gouessej> the FUD campaign when Sun Microsystems/Oracle left the boat
20150824 12:23:08 <elect> ah
20150824 12:23:12 <elect> I wasnt here
20150824 12:23:25 <sgothel> we are measured by work/results not by 'some folks blubbering around'
20150824 12:23:58 <elect> right
20150824 12:24:14 <gouessej> elect: sgothel kept a better self-control than me at this time
20150824 12:24:26 <elect> :D
20150824 12:24:32 <elect> I can image ^^
20150824 12:24:51 <sgothel> just b/c I don't care what others say .. usually, if I don't know them - only family and close friends can hurt me.
20150824 12:25:00 <sgothel> so .. don't hurt me :)
20150824 12:25:08 <xranby> zubzub: ate least i have a raspberry pi using the mesa3d driver that have GBM / EGL enabled
20150824 12:25:12 <elect> anyway gouessej, I think the decision to leaving jgo was a bad decision
20150824 12:25:19 <elect> we should come back over there
20150824 12:25:22 <xranby> zubzub: thus i will look into it after 2.3.2
20150824 12:25:29 <zubzub> xranby: cool!
20150824 12:25:34 <elect> just as minimal support at least
20150824 12:25:41 <gouessej> elect: I don't want to get banned once again
20150824 12:25:48 <elect> I can do that
20150824 12:25:57 <elect> :D
20150824 12:26:00 <elect> they banned you?
20150824 12:26:29 <gouessej> elect: You can already answer their questions but please don't ask to reopen the JOGL section without our agreement, it should be a collective decision
20150824 12:26:29 <xranby> sgothel: i am chasing a bug that originate from the newt x11 driver today https://jogamp.org/bugzilla/show_bug.cgi?id=1178#c8
20150824 12:26:31 <sgothel> @elect: we don't need to be on all weddings .. and I actually even like 'forums' like LKML most :-/
20150824 12:26:48 <gouessej> elect: I have been temporarily banned several times
20150824 12:27:05 <gouessej> elect: you wrote it, you weren't there during the FUD campaign
20150824 12:27:10 <elect> sure, I wont do anything on jogl without any collective decision
20150824 12:27:16 <elect> that's why I start from here
20150824 12:27:22 <gouessej> elect: then it's difficult for you to understand my decisiosn
20150824 12:27:24 <gouessej> decisions
20150824 12:27:30 <elect> if you dont agree, I wont go further
20150824 12:27:35 <elect> this is implicit
20150824 12:27:40 <sgothel> @Xerxes: great
20150824 12:28:15 <sgothel> bottom line: focus on what makes things great, not just *look* great :)
20150824 12:28:27 <elect> but not being there gives a big edge to lwjgl I think
20150824 12:28:45 <gouessej> elect: You can reply to questions about JOGL on any forum but in my humble opinion, redirecting people on our forum for specific questions is better than participating to those flamewars
20150824 12:28:46 <elect> new users will definitely finish to use it
20150824 12:28:53 <elect> I know
20150824 12:29:00 <gouessej> elect: Some JOGL users are still on JGO
20150824 12:29:05 <gouessej> elect: they reply
20150824 12:29:08 <elect> but still, you can redirect and answer at the same time
20150824 12:29:26 <gouessej> elect: I prefer having a consistent policy
20150824 12:29:52 <elect> I'd too, but it is hurting us
20150824 12:30:01 <gouessej> elect: I try to redirect people to the forum even on StackOverflow
20150824 12:30:13 <sgothel> @Xerxes: its probably just the autoposition, i.e. the underlay's position has not been set yet?
20150824 12:30:20 <elect> look at the activities, on forums and on stackov/gamedev
20150824 12:30:24 <gouessej> elect: We're not totally absent anyway
20150824 12:30:42 <sgothel> Know that the WM finally decides the position of a window, we can only ask/beg for one ..
20150824 12:30:43 <elect> I know, but some people have a different impression
20150824 12:31:13 <sgothel> guys .. in the professional graphics world .. JogAmp is the number one binding for crying out loud!
20150824 12:31:15 <elect> and this is what counts most when a new user has to take a decision between two libs
20150824 12:31:26 <gouessej> elect: We don't look like dead, do we?
20150824 12:31:30 <sgothel> so you don't need to paint the pink elephant :)
20150824 12:31:51 <zubzub> 14:30 <+sgothel> Know that the WM finally decides the position of a window, we can only ask/beg for one ..
20150824 12:31:52 <gouessej> I agree with Sven
20150824 12:31:56 <elect> well, you know better than me, I already read you a couple of times in different places answering that jogl is not dead and so on
20150824 12:31:58 <elect> you tell me
20150824 12:32:02 <zubzub> on wayland you can't even ask! ;)
20150824 12:32:13 <zubzub> except for "move according to mouse pointer plz"
20150824 12:32:33 <sgothel> @zuzub: the WM (-> compositior) I guess you can .. no?
20150824 12:32:37 <zubzub> nope
20150824 12:32:43 <sgothel> i.e. Kwin ..
20150824 12:32:44 <sgothel> :)
20150824 12:32:47 <elect> I never, I repeat never, read about questioning if lwjgl is dead
20150824 12:33:00 <elect> but I did a couple of time about jogl
20150824 12:33:03 <zubzub> you can ask the compositor to move your window through a mouse pointer movement
20150824 12:33:09 <zubzub> but you don't know where how and when
20150824 12:33:28 <sgothel> well .. I hope kwin allows that .. which may run on wayland soon
20150824 12:33:28 <gouessej> elect: What more do you want us to do? It was caused by the FUD campaign
20150824 12:34:04 <zubzub> sgothel: will not happen, it's explictly forbidden in wayland
20150824 12:34:06 <elect> as I said, I'd like to reopen the jogl subsection on jgo
20150824 12:34:16 <zubzub> sgothel: there's compositor space and there's application space
20150824 12:34:18 <elect> invitating the users to come to jogl forum for a better support
20150824 12:34:34 <elect> but if there is any thread, I'll be glad to answer/reply that
20150824 12:34:37 <zubzub> applicaiton only knows about it's own geometry, and it's position (that is in compositor space) is not part of that
20150824 12:35:11 <sgothel> @elect: I can't stop you from doing so - but I personally don't like it. your choice in the end. Q: why distracting yourself - waste time and energy, where you can be productive w/ code?
20150824 12:35:12 <gouessej> elect: it would go completely against the policy that have given us some stability for several years
20150824 12:35:24 <elect> we are civil peoples, I think we can co-live together if everyone keeps calm and civil
20150824 12:35:42 <sgothel> i.e. when I want to discuss Linux Kernel, I go to LKLM ..
20150824 12:35:48 <elect> sgothel, I wont if you dont like it
20150824 12:35:56 <elect> I already said that
20150824 12:36:01 <elect> I am the new one here
20150824 12:36:13 <elect> I just want to share my point of view
20150824 12:36:36 <sgothel> you are a free person - I just wouldn't do it myself, and I personally prefer time spend constructive :)
20150824 12:36:39 <gouessej> elect: I can't bear revisionist people. Isn't there anything that you can do on JGO except reopening this section?
20150824 12:36:39 * packet0 (~packet0@anon) has joined #jogamp
20150824 12:37:02 <sgothel> so me not an evangelist :)
20150824 12:37:07 <elect> :D
20150824 12:37:16 <elect> I dont know, gouessej
20150824 12:37:37 <gouessej> elect: Why not improving the forum first?
20150824 12:37:49 <sgothel> -> wiki :)
20150824 12:37:56 <elect> I tried ^^, I tried also on the wiki
20150824 12:37:56 <sgothel> -> demo code + features ?
20150824 12:37:59 <gouessej> elect: If it's too difficult, start by improving the wiki
20150824 12:38:02 <elect> im doing
20150824 12:38:11 <gouessej> yes
20150824 12:38:20 <elect> the wiki needs a cleaning
20150824 12:38:26 <elect> especially the tutorial page
20150824 12:38:31 <sgothel> wiki: you did / doing: right - KUDOS, will read / review it later on - when there is time, post 2.3.2
20150824 12:38:33 <gouessej> it's a bit messy
20150824 12:38:42 <elect> you should advice legacy tutorials, like opengl wiki does
20150824 12:39:06 <gouessej> The OpenGL wiki is just awesome
20150824 12:39:40 <elect> I started the Api Overview, but I stopped for the texture stuff
20150824 12:40:02 <elect> I also wrote a small Hello Triangle
20150824 12:40:46 <elect> I can write, but before reviewing old part (read: deleting/moving) I wait for a collective decision
20150824 12:41:06 <gouessej> elect: The only drawback of your example is the fact it isn't self contained
20150824 12:41:29 <sgothel> lets do a review after 2.3.2
20150824 12:41:30 <gouessej> elect: but it's ok anyway
20150824 12:41:43 <elect> what you mean?
20150824 12:42:06 <elect> just a jar to execute?
20150824 12:42:51 <sgothel> review your wiki work post 2.3.2, then we all make a decision about the wiki structure of the tutorial
20150824 12:43:18 <elect> I meant the self contained, sgothel, sorry
20150824 12:43:39 <gouessej> elect: Look at my example
20150824 12:44:06 <gouessej> elect: You can copy/paste it somewhere, everything is in a single file. It's just for the newbies
20150824 12:44:23 <elect> ah, you mean because of the external shaders?
20150824 12:44:53 <gouessej> elect: but if putting everything into a single file drives it too much messy, keep it unchanged
20150824 12:48:08 <elect> maybe I will open a thread to discuss the jgo reopening section
20150824 12:48:13 <elect> and see what it will bring
20150824 12:48:57 <gouessej> elect: now? Please don't do it
20150824 12:49:33 <elect> ok
20150824 12:49:48 <elect> I meant on jogl, btw
20150824 12:49:52 <elect> jogamp
20150824 12:50:29 <gouessej> elect: it's a bit strange to talk about another forum on ours. Most concerned people can express their opinions here
20150824 12:50:31 <elect> did you think on jgo?
20150824 12:50:40 <gouessej> honestly, yes
20150824 12:50:42 <gouessej> sorry
20150824 12:50:45 <elect> my fault
20150824 12:50:47 <elect> dont worry
20150824 12:51:27 <elect> I know it is strange, but jgo is a big community too
20150824 12:51:42 <elect> we shouldnt underestimate its influence
20150824 12:52:00 <elect> a minimal presence should be enough
20150824 12:52:26 <gouessej> I'm going to be a bit harsh but the game programming community looks like a big graveyard, tons of projects disappear or never appear
20150824 12:52:44 <gouessej> elect: we are still on JGO
20150824 12:52:45 <elect> people will know that a jogl thread on jgo will be read mostly by java users and some jogl user, on jogamp mostly by jogamp users
20150824 12:52:53 <gouessej> elect: there is a minimal presence on JGO
20150824 12:53:12 <elect> where?
20150824 12:53:18 <elect> not as jogamp community
20150824 12:54:39 <gouessej> elect: The threads are locked, JOGL users on JGO know that they can come on our official forum
20150824 12:55:16 <elect> you see, this is bad, this is what gives the impression of a dead community
20150824 12:55:21 <gouessej> elect: When they don't, some people answer and I don't want to repeat that they should go to our forum
20150824 12:55:39 <gouessej> elect: no, it doesn't, we have our own forum
20150824 12:55:53 <elect> I know, but many have this feelings
20150824 12:56:06 <gouessej> elect: they should look at the commits
20150824 12:56:09 <elect> and some continue doing FUD and repeating that to others too
20150824 12:56:19 <gouessej> elect: just plain facts, look at Sven's commits
20150824 12:56:23 <elect> but a nub doesnt even know what a commits is..
20150824 12:56:33 <gouessej> elect: no, there is no longer any FUD
20150824 12:56:34 <elect> you think as skilled user
20150824 12:57:02 <elect> coz you are
20150824 12:57:07 <elect> but many are not
20150824 12:57:15 <gouessej> elect: If they prefer according any credits to gossips, I can do nothing for them
20150824 12:57:23 <elect> and dont share your same reasoning process
20150824 12:57:49 <gouessej> elect: I haven't noticed any recent FUD campaign against JogAmp
20150824 12:57:56 <gouessej> elect: I'm serious
20150824 12:59:46 <gouessej> elect: Where have you seen such persisting FUD?
20150824 13:00:13 <elect> I read it somewhere
20150824 13:00:29 <elect> when I was looking for tech answers
20150824 13:00:38 <elect> but it wasnt intentional
20150824 13:00:43 <elect> that's the point
20150824 13:01:28 <elect> recently I read it
20150824 13:01:55 <elect> http://gamedev.stackexchange.com/questions/105308/what-kind-of-issues-should-i-watch-out-for-or-avoid-when-using-joglopengl-java
20150824 13:01:59 <gouessej> elect: We can use some bots to detect these craps
20150824 13:02:51 <elect> also on (some dated) stuff like on comments on stackoverflow and some thread on jgo
20150824 13:04:58 <gouessej> elect: Yes but they are dated as you say
20150824 13:05:16 <gouessej> I see nothing wrong in the post above
20150824 13:06:19 <elect> yeah, I couldnt find the one I mean
20150824 13:08:34 <elect> anyway, what about discussin the jgo subsection reopening in a jogamp thread?
20150824 13:08:52 <elect> so I can see what you guys think about it
20150824 13:12:21 <gouessej> Why not talking about it here instead?
20150824 13:12:55 <elect> because it's harder to keep trace of anyone
20150824 13:13:05 <elect> but if you want, ok
20150824 13:13:13 <gouessej> elect: What do you mean exactly? Soory
20150824 13:13:15 <gouessej> Sorry
20150824 13:13:37 <gouessej> the forum is public and this channel is logged
20150824 13:13:43 <elect> that maybe the other dont have time or they arent present
20150824 13:14:01 <elect> with a forum we have all about a thematic condensed in one plac
20150824 13:14:07 <elect> with a *thread*
20150824 13:15:20 <elect> and anyone can also better organizing and expose his thinking
20150824 13:22:41 <gouessej> Ok but only a very few people are active both on JGO and on our JogAmp forum
20150824 13:26:47 * eclesia (~husky@anon) Quit (Ping timeout: 246 seconds)
20150824 13:27:59 <xranby> sgothel: do you think that the rift may work on arm? is there parts of the sdk that are only available for x86?
20150824 13:28:24 <gouessej> xranby: Hasn't it become a Windows only device?
20150824 13:28:42 <xranby> gouessej: jogamp jogl have natives for linux-x86 and linux-x86_64
20150824 13:28:49 <sgothel> the old 0.3.* branch works on any platform (no blob)
20150824 13:29:21 <sgothel> however .. the new 'design' uses a blob, a server .. hmm .. and we don't have the source code
20150824 13:29:48 * eclesia (~husky@anon) has joined #jogamp
20150824 13:29:50 <sgothel> the new design, >= 0.4, supports the positioning and new DK2
20150824 13:30:14 <sgothel> either we take the effort to patch 0.3 .. or use OpenHDM or the like
20150824 13:30:21 <sgothel> OpenHMD
20150824 13:31:00 <sgothel> TBH .. they suck lately, i.e. as Julien stated .. 'Windows only' until further notice .. and no more client renderer
20150824 13:31:23 <sgothel> forcing to use their GL utilization when they have re-written it (currently only D3D)
20150824 13:31:33 <sgothel> it sucks
20150824 13:31:46 <sgothel> creating a walled garden w/o any need
20150824 13:32:03 <sgothel> as if they don't like to sell their hardware :-/
20150824 13:32:51 <gouessej> :s
20150824 13:32:55 <sgothel> so .. patching the 0.3 source tree may work, but for the positional sensor (head tracker) .. no source, but you may double check the old code
20150824 13:33:23 <sgothel> or we ASAP support the free software approach .. OpenHMD .. etc
20150824 13:33:34 <sgothel> since those gladly only focus on the sensors
20150824 13:33:42 <sgothel> (*doing it the right way*)
20150824 13:34:23 <gouessej> Do we need OpenHMD as a whole or only a subset for our purposes?
20150824 13:34:36 <sgothel> 0.5.0.1 <- the last JogAmp support for OculusVR's SDK I am afraid :-/
20150824 13:34:58 <sgothel> I don't know the whole .. i.e. didn't read any details yet
20150824 13:35:20 <sgothel> but I would like to plug-in this project similar to oculusvr-sdk
20150824 13:35:29 <sgothel> (git sub)
20150824 13:35:53 <sgothel> but they are not done yet w/ positional tracking AFAIK
20150824 13:38:20 <gouessej> sgothel: What is it? https://github.com/ValveSoftware/openvr
20150824 13:38:39 <xranby> sgothel: thank you for sharing all the details
20150824 13:38:52 <sgothel> that ,.. looks great :)
20150824 13:39:12 <sgothel> especially the short license
20150824 13:39:46 <sgothel> we have to wait a bit until things are available for public .. then we can add those 'drivers' and test
20150824 13:40:36 <sgothel> I didn't like to jump through hoops .. to either talk to Oculus folks about the driver - nor tried to beg to get a Valve DK (paperwork)
20150824 13:41:42 <sgothel> but when things are available .. we add support, sure
20150824 13:41:53 <gouessej> The source code doesn't seem to be available
20150824 13:42:01 <sgothel> the heavy lifting is already done within our stereo package
20150824 13:42:35 <sgothel> oh really? I see - that is 'legal' on github? duh! :-/
20150824 13:43:15 <gouessej> This article explains how to make something useful with this library: http://www.roadtovr.com/making-valves-openvr-truly-inclusive-for-vr-headsets/
20150824 13:49:28 <elect> sgothel, do you think it has something to do with facebook acquisition?
20150824 13:50:06 <gouessej> https://github.com/phr00t/jmonkeyengine-virtual-reality
20150824 13:50:52 <sgothel> tightening budgets, 'collusion' w/ 'partners', who knows .. :-/
20150824 13:51:19 <elect> I read an interestin intervent of jherico btw
20150824 13:51:22 <elect> let me search for it
20150824 13:52:10 <sgothel> for sure .. the [consumer] market is quite tight anyways today, so it reminds me of https://en.wikipedia.org/wiki/Places_in_The_Hitchhiker's_Guide_to_the_Galaxy#Magrathea (waiting for the economy to recover) .. :)
20150824 13:52:55 <sgothel> jaja .. jherico tried to fight the crap - but they don't listen to the many voices raised ..
20150824 13:53:17 <elect> have you read it too?
20150824 13:53:26 <elect> the one regarding the bug tracking system
20150824 13:53:35 <elect> I cannot find it right now
20150824 13:53:50 <sgothel> sure I walked through their forum a bit to get an idea what the heck is wrong w/ em :-/
20150824 13:54:42 <sgothel> @Julien: OpenVR .. etc .. nice nice
20150824 13:55:18 <sgothel> again .. will pick this up later on, but feel free to add another binding, shouldn't be too hard w/ current existing infrastructure
20150824 13:55:48 <elect> it's really a pity
20150824 13:56:03 <elect> if the co-founder didnt get killed I guess things would have been different
20150824 13:56:09 <elect> he looked to have passion
20150824 13:56:40 <elect> I read it was hurt by a car after a bank rubbery
20150824 13:56:45 <elect> crazy
20150824 13:57:07 <sgothel> didn't know that .. crazy indeed
20150824 14:00:29 <gouessej> sgothel: I have to document ImageType first
20150824 14:05:03 * elect (~elect@anon) Quit (Read error: Connection reset by peer)
20150824 14:46:47 * elect (~elect@anon) has joined #jogamp
20150824 14:55:33 <gouessej> elect: What is the state of progress concerning Discourse?
20150824 14:56:09 <elect> as far as I know, the only problem is javascript
20150824 14:56:49 <elect> I'd switch anyway to Discourse and then if we find a better alternative we switch again
20150824 14:57:27 <elect> I know it can be deadly, but I dont think there is any other alternative to nabble with the javascript constrain
20150824 14:59:12 <elect> as I wrote on the bug tracking, in my opinion having a javascript forum such as Discourse is much better than the current situation, but this is my 2 cents
20150824 14:59:32 <elect> I'd not like to result too insistent
20150824 15:04:05 <gouessej> elect: Just make a test with noscript enabled. If it works, the end users will be free to disable Javascript at the cost of a degradation of the display
20150824 15:04:39 <elect> I asked if that'd have been possible on the discourse forum
20150824 15:05:09 <elect> but they replied if you aim to disable javascript there is no point to use discourse in the first place, gouessej
20150824 15:05:22 <elect> or something similar
20150824 15:05:47 <xranby> sgothel: http://www.x.org/archive/X11R7.5/doc/man/man3/XConfigureEvent.3.html when we recive this event then x and y is set to 1 and 30
20150824 15:05:52 <xranby> "The x and y members are set to the coordinates relative to the parent window's origin and indicate the position of the upper-left outside corner of the window. The width and height members are set to the inside size of the window, not including the border. The border_width member is set to the width of the window's border, in pixels. "
20150824 15:06:18 <gouessej> elect: It's up to some end users to disable it, it would be a pragmatic solution
20150824 15:06:54 <elect> uhm, it sounds logic
20150824 15:06:54 <gouessej> elect: It would change nothing for most end users
20150824 15:07:09 <elect> I'll investigate on this fact again then
20150824 15:07:16 <elect> and I'll update
20150824 15:07:41 <gouessej> elect: You could make the new forum with Discourse and some of us could still disable Javascript
20150824 15:07:52 <xranby> sgothel: inside the ConfigureNotify
20150824 15:07:53 <gouessej> elect: you can test on the official JMonkeyEngine forum
20150824 15:08:07 <gouessej> elect: it uses Discourse
20150824 15:08:35 <xranby> sgothel: line 204 of src/newt/native/X11Event.c
20150824 15:09:11 <elect> I saw, gouessej
20150824 15:12:26 <xranby> sgothel: thus the bug is that we use the relative x and y instead of the absolute x and y
20150824 15:12:27 * gouessej (5ee4b442@anon) Quit (Ping timeout: 246 seconds)
20150824 15:12:41 * xranby trying to figure out how to find the absolute x and y
20150824 15:12:56 <xranby> (xwininfo can figure it out)
20150824 15:13:44 <zubzub> xranby: you basically have to translate your coordinates to those of the root window
20150824 15:14:21 <zubzub> https://tronche.com/gui/x/xlib/window-information/XTranslateCoordinates.html
20150824 15:14:26 <zubzub> xranby: ^
20150824 15:14:32 <zubzub> get the root window
20150824 15:14:38 <zubzub> get your window
20150824 15:14:42 <zubzub> etc
20150824 15:17:33 <zubzub> I wonder how long xranby will be googling before he reads my solution :p
20150824 15:17:49 <xranby> i am reading source etc
20150824 15:18:13 <zubzub> of xwininfo?
20150824 15:18:16 <xranby> at least i am happy that i have found the bug origin (i think)
20150824 15:18:38 <xranby> no i am looking if we have a reference to the rootwindow handy
20150824 15:18:45 <xranby> inside the event loop
20150824 15:19:02 <zubzub> probably not I would think, unless it's used to do some init
20150824 15:19:54 * zubzub goes home
20150824 15:20:15 <xranby> Display * dpy = (Display *) (intptr_t) jdisplay;
20150824 15:20:15 <xranby> Screen * scrn = ScreenOfDisplay(dpy, (int)screen_index);
20150824 15:20:15 <xranby> Window root = XRootWindowOfScreen(scrn);
20150824 15:20:52 <zubzub> yup that's how you can get it
20150824 15:21:12 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20150824 15:22:10 <zubzub> https://tronche.com/gui/x/xlib/window-information/XGetWindowAttributes.html
20150824 15:22:12 <zubzub> also possible
20150824 15:22:44 <zubzub> https://tronche.com/gui/x/xlib/window-information/XQueryTree.html
20150824 15:22:46 <zubzub> same
20150824 15:23:27 * zubzub goes home
20150824 15:23:30 <zubzub> gl with your bug :)
20150824 15:24:36 * monsieur_max (~maxime@anon) has joined #jogamp
20150824 15:26:45 * gouessej (5ee4b442@anon) has joined #jogamp
20150824 15:26:54 <gouessej> elect: "Powered by Discourse, best viewed with JavaScript enabled" :)
20150824 15:26:59 <gouessej> elect: It works
20150824 15:27:12 <gouessej> elect: The fallback is acceptable
20150824 15:27:13 <xranby> zubzub: thank you
20150824 15:29:40 * elect_ (~elect@anon) has joined #jogamp
20150824 15:35:03 <gouessej> elect_: https://jogamp.org/bugzilla/show_bug.cgi?id=1095#c26
20150824 15:35:37 <elect_> awesome, mercì gouessej
20150824 15:35:46 <gouessej> elect_: You're welcome
20150824 15:36:12 <gouessej> elect_: I think that we can find a solution to preserve the old posts in read only state in Discourse
20150824 15:36:37 <elect_> I hope so
20150824 15:36:44 <gouessej> elect_: I know how to do it
20150824 15:36:51 <elect_> otherwise we can always keep jogamp as read only
20150824 15:36:59 <elect_> oh nice, that's perfect, gouessej
20150824 15:37:31 <gouessej> elect_: We can generate some URL rewriting rules and/or some redirections for Apache so that everything will be in a single forum
20150824 15:38:17 <gouessej> elect_: We just have to compare the structures of the URLs in Nabble and in Discourse to write such rules
20150824 15:47:18 <gouessej> elect_: I understand the structures of the URLs created by both except the permalinks in Discourse
20150824 15:47:57 <gouessej> elect_: now I get it
20150824 15:50:03 <gouessej> elect_: f for each folder, tp or td for each topic, p for each post, the name of the topic is build by replacing the illegal characters by '-', Discourse just uses '/' instead of f/tp/td/p/#
20150824 15:51:19 <gouessej> elect_: We could even allow discussions started with Nabble to go on with Discourse :)
20150824 15:51:59 <elect_> !
20150824 15:53:10 <gouessej> elect_: Please ensure that the mailing list feature works as expected and I can take care of the rewriting rules
20150824 15:53:19 <elect_> yep, I tested it
20150824 15:53:29 <elect_> I also created a jogamp gmail account
20150824 15:53:29 <gouessej> elect_: Cool :)
20150824 15:53:36 <gouessej> elect_: gmail :(
20150824 15:53:43 <elect_> lol
20150824 15:54:00 <gouessej> elect_: We can host a small mail server instead
20150824 15:54:29 <gouessej> elect_: Actually, it is already done
20150824 15:54:34 <elect_> sure
20150824 15:54:45 <elect_> but I needed something fast for my setup
20150824 15:54:51 <gouessej> elect_: look at the "mediastream" email address
20150824 15:55:49 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20150824 15:55:52 <gouessej> elect_: I'll write the rewriting rules and you'll have to put them into the root of your server
20150824 15:58:33 <elect_> I shutdown my server
20150824 15:58:55 <elect_> it costed me money
20150824 15:59:02 * badshah400 (~badshah40@anon) Quit (Quit: badshah400)
20150824 15:59:06 <elect_> and I didnt need for anything else
20150824 16:00:34 <elect_> what would you like to do?
20150824 16:00:51 <elect_> put on it again and import all nabble threads?
20150824 16:02:01 <gouessej> I can run a local server but not now
20150824 16:02:08 <gouessej> elect_: Bye.
20150824 16:02:13 * gouessej (5ee4b442@anon) Quit (Quit: Page closed)
20150824 16:03:05 <elect_> wow, that was fast ^^
20150824 16:03:39 * eclesia (~husky@anon) has left #jogamp
20150824 16:13:34 <elect_> removing oracle java btw
20150824 16:23:39 * packet0 (~packet0@anon) Quit ()
20150824 16:32:26 * elect_ (~elect@anon) Quit (Remote host closed the connection)
20150824 16:41:07 * elect_ (~elect@anon) has joined #jogamp
20150824 16:41:32 <elect_> I cant get, I have openjdk now
20150824 16:41:33 <elect_> elect@elect-desktop:~/Downloads$ java -version
20150824 16:41:33 <elect_> openjdk version "1.8.0_45-internal"
20150824 16:41:33 <elect_> OpenJDK Runtime Environment (build 1.8.0_45-internal-b14)
20150824 16:41:33 <elect_> OpenJDK 64-Bit Server VM (build 25.45-b02, mixed mode)
20150824 16:42:18 <elect_> but if I try to reinstall netbeans it says "no compatible jdk found"
20150824 16:47:03 <elect_> it seems netbeans and openjdk dont work well together
20150824 16:52:23 <elect_> ok fixed
20150824 17:02:22 <elect_> Im using openjdk 8, is it compatible for building jogl or it is mandatory to use the 7?
20150824 17:09:52 <sgothel> you shall validate against a java7 setup (you can use java8 for that), see make/scripts/make...*.sh (i.e. for linux amd64)
20150824 17:13:37 <elect_> I have 15.04 x64
20150824 17:13:45 <elect_> make.gluegen.all.linux-x86_64-clang.sh
20150824 17:13:46 <elect_> make.gluegen.all.linux-x86_64.sh
20150824 17:13:54 <elect_> shall I run the clang or not?
20150824 17:14:10 <elect_> I have no idea what that means >.>
20150824 17:16:36 <elect_> wait, I do have java 7, but it's using 8 as default one
20150824 17:16:37 <elect_> elect@elect-desktop:~/NetBeansProjects/jogamp/gluegen/make/scripts$ sudo update-alternatives --config java
20150824 17:16:37 <elect_> [sudo] password for elect:
20150824 17:16:37 <elect_> There are 2 choices for the alternative java (providing /usr/bin/java).
20150824 17:16:37 <elect_> Selection Path Priority Status
20150824 17:16:37 <elect_> ------------------------------------------------------------
20150824 17:16:39 <elect_> * 0 /usr/lib/jvm/java-8-openjdk-amd64/jre/bin/java 1081 auto mode
20150824 17:16:41 <elect_> 1 /usr/lib/jvm/java-7-openjdk-amd64/jre/bin/java 1071 manual mode
20150824 17:16:43 <elect_> 2 /usr/lib/jvm/java-8-openjdk-amd64/jre/bin/java 1081 manual mode
20150824 17:34:42 <sgothel> see: gluegen/make/scripts/make.gluegen.all.linux-x86_64.sh
20150824 17:34:54 <sgothel> export SOURCE_LEVEL=1.6
20150824 17:34:54 <sgothel> export TARGET_LEVEL=1.6
20150824 17:34:54 <sgothel> export TARGET_RT_JAR=/opt-share/jre1.6.0_30/lib/rt.jar
20150824 17:37:07 <elect_> why the exports?
20150824 17:40:25 <sgothel> gluegen does interpret those env-vars
20150824 17:41:00 <sgothel> read: http://jogamp.org/git/?p=gluegen.git;a=blob;f=make/jogamp-env.xml;h=3e997629644d83597d629bd497779d630dea6ecb;hb=HEAD
20150824 18:32:22 * monsieur_max (~maxime@anon) has joined #jogamp
20150824 18:43:07 <a11mad11> elect : I want to put different type of data (float int short byte) in the same vbo using the stride (v1,c1,n1,v2,c2,n2,...) but the glbufferdata doesn't have the stride argument
20150824 19:15:03 * badshah400 (~badshah40@anon) has joined #jogamp
20150824 19:32:30 * badshah400 (~badshah40@anon) Quit (Quit: badshah400)
20150824 19:36:25 * xranby_ (~familjen@anon) has joined #jogamp
20150824 20:13:01 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20150824 20:23:23 <elect_> a11mad11, you need the stride only for glVertexAttribPointer, glBufferData need only the size (in bytes) and the buffer itself, it doesnt care of nothing else
20150824 20:23:35 <elect_> thanks sgothel, I'll continue tomorrow
20150824 20:23:42 <elect_> gute Nacht
20150824 20:23:58 <sgothel> good night
20150824 20:25:19 <a11mad11> but the data that a place in the buffer need to be like that (v1,c1,n1,v2,c2,n2)?
20150824 20:25:46 <a11mad11> i already allocate the size of the buffer with glbufferdata
20150824 20:26:48 <a11mad11> now i want to place different type of data like (vertex=float;color=byte;normal=float)
20150824 20:27:22 <sgothel> @a11mad11: are you referring to a JOGL [util] class?
20150824 20:28:05 * elect_ (~elect@anon) Quit (Ping timeout: 252 seconds)
20150824 20:28:10 <sgothel> please check GearsES2 .. it shows you how we use stride and interleaved ..
20150824 20:28:15 <a11mad11> which class, I use java.nio.buffer and GL2 of jogl
20150824 20:29:25 <xranby_> hi and hello night hackers
20150824 20:30:39 <sgothel> hi ho
20150824 20:30:46 <sgothel> @a11: yup ..
20150824 20:31:07 <sgothel> so your stride / interleave gaps etc .. must be within the allocation
20150824 20:31:27 <xranby_> Bool NewtWindows_updateInsets(Display *dpy, JavaWindow * w, int *left, int *right, int *top, int *bottom); <- definition
20150824 20:31:38 <xranby_> but we have some places this function is called with
20150824 20:31:40 <xranby_> NewtWindows_updateInsets(env, w->jwindow, dpy, evt.xany.window, &left, &right, &top, &bottom);
20150824 20:31:56 <xranby_> can someone explain why the compiler did not complain?
20150824 20:32:06 <sgothel> *headache*
20150824 20:32:24 <sgothel> checking .. fixing .. one sec
20150824 20:33:15 <sgothel> X11Event.c <- not used
20150824 20:33:44 <a11mad11> +sgothel what is the allocation
20150824 20:33:56 <sgothel> NewtWindows_updateInsets(dpy, jw, ...) <- always used, hence .. all good
20150824 20:34:17 <xranby_> sgothel: oh that explains why my "fixes" to X11Event.c has no effect :P
20150824 20:34:22 <sgothel> either your mmap'ed GL buffer or the glBufferData ..
20150824 20:34:51 <sgothel> @xerxes: we shall remove unmaintained code .. I guess, i.e. X11Event.c
20150824 20:35:12 <sgothel> since its in X11Display.c .. sorry for the confusion
20150824 20:36:12 <xranby_> awesome, i will continue hacking then
20150824 20:36:25 <sgothel> 'awesome' ? :)
20150824 20:36:44 <a11mad11> i need to put my data with the stride (v1,c1,v2,c2) in the buffer using glbufferdata?
20150824 20:37:13 <sgothel> for example .. this would be an interleaved example
20150824 20:37:14 <xranby_> sgothel: you have restored my sanity
20150824 20:37:18 <sgothel> ah :)
20150824 20:37:19 <a11mad11> but my vertex is float data and my color is byte
20150824 20:37:41 <sgothel> well .. then not :)
20150824 20:37:51 <sgothel> interleaved data must be of same type ofc
20150824 20:38:13 <sgothel> or .. err .. wait .. maybe not .. using an underlying ByteBuffer
20150824 20:38:13 <a11mad11> ok so i need to use different vbo
20150824 20:38:19 <sgothel> so pls check GearsES2 ..
20150824 20:38:37 <a11mad11> ok i am checking
20150824 20:41:46 <sgothel> GearsObject uses same type .. but it is not required by the spec AFAIK, would need to check ..
20150824 20:41:55 <sgothel> i.e. you simply pass an offset to the bound buffer
20150824 20:42:24 <sgothel> GearsObject* uses our GLArrayDataServer utility ..
20150824 20:42:47 <sgothel> (which only allows same type stuff)
20150824 20:44:27 <a11mad11> is more simple using different vbo for each type
20150824 20:44:30 <a11mad11> of data
20150824 20:44:49 <xranby_> sgothel: in X11Display.c case ConfigureNotify then evt.xconfigure.x and evt.xconfigure.y is in relative coordinates to the parent window
20150824 20:45:14 <xranby_> i will try hack these to be absolute according to the root window
20150824 20:45:16 <sgothel> if child .. probably - was this a statement or Q?
20150824 20:45:22 <sgothel> ah
20150824 20:45:33 <sgothel> we have a getLocationOnScreen already ..
20150824 20:46:41 <sgothel> Java_jogamp_nativewindow_x11_X11Lib_GetRelativeLocation0
20150824 20:46:42 <xranby_> sgothel: yes.. but that method is designed to be called from java
20150824 20:46:52 <sgothel> well ..
20150824 20:46:55 <xranby_> i will add something similar here..
20150824 20:47:08 <sgothel> XTranslateCoordinates directly then :)
20150824 20:47:48 <sgothel> sizePosMaxInsetsChangedID <- does expect the relative coords in case of being a child window!
20150824 20:48:01 <sgothel> -> NEWT Window semantics that is!
20150824 20:48:17 <sgothel> otherwise you will break it all ..
20150824 20:49:15 <xranby_> .. or do you mean that i should call this jni method from the sizePosMaxInsetsChanged java method?
20150824 20:49:52 <xranby_> i will take a look ofc... trying to make things work
20150824 20:49:56 <sgothel> 1st: NEWT Window always maintains/uses the relative coordinates (if TOP-LEVEL it's absolute ..)
20150824 20:50:36 <sgothel> if you need the absolute position .. usually you call the getLocationOnScreen method
20150824 20:51:26 <sgothel> however, if you need to speed things up a bit .. sure, you could modify the X11 WindowDriver a bit to always maintain the absolute position .. but IMHO thats a bit weird .. would need to know reason/use-case
20150824 20:52:41 <sgothel> IMHO just use the java method to propagate the abs position .. done
20150824 20:53:27 * xranby_ looks inside /src/newt/classes/jogamp/newt/WindowImpl.java: protected void sizePosMaxInsetsChanged
20150824 20:54:44 <xranby_> the issue i observe is that when i resize my window i get many newt events with possition 0,0 on my x86_64 machine a 1,31 on my raspberry pi
20150824 20:55:41 <xranby_> do we have a robot that can move and resize newt windows hmm...
20150824 20:55:41 <sgothel> you have fixed position before? i.e. setPosition(x, y)
20150824 20:55:57 * xranby_ think i better create a junit test to show sgothel
20150824 20:56:20 <sgothel> resize only programmatically .. yet
20150824 20:56:43 <sgothel> its quite hard to get the 'resize corner/handle' .. via robot
20150824 20:56:53 <sgothel> bottom right corner .. usually
20150824 20:57:38 <sgothel> however ConfigureNotify's position data is not the problem here (relative coords delivered and expected)
20150824 20:58:06 <sgothel> would be more the WM settings .. and/or our way to 'ask for the position'
20150824 20:58:28 <sgothel> NewtWindow_setVisiblePosSize
20150824 20:58:55 <sgothel> (wrong OS :)
20150824 20:59:37 <sgothel> NewtWindows_setPosSize
20150824 21:00:11 <sgothel> uses XConfigureWindow ..
20150824 21:00:36 <sgothel> but needs either valid size or valid position .. hmm :)
20150824 21:00:56 <sgothel> (the latter reads like a bug)
20150824 21:24:46 <xranby_> it always consider the possiton passed as valid
20150824 21:49:53 <xranby_> so it turn out that... WindowImpl getX() did not work as I expected... it returns relative instead of on screen absolute coordinates.. hence I have to use getLocationOnScreen
20150824 22:00:01 <xranby_> sgothel: i have fixed my underlay code it used the wrong assumption
20150824 22:01:37 <xranby_> one think is kind of funny... newt windows do not consume events unless they have focus?
20150824 22:10:17 <xranby_> the lession of the day... always edit code in your ide with javadoc enabled
20150824 22:10:41 <xranby_> and you will find out why your understanding of things are wrong faster
20150824 22:31:30 <sgothel> sure .. always good to know the semantics .. ahem :-]
20150824 22:32:51 <sgothel> so .. it seems like 'black September' hast started early on the markets already .. well well, 'Slartibartfast' goes to bed :)
20150824 22:34:30 <xranby_> not everyone can afford a fjord
20150824 22:35:29 <sgothel> oh .. he was just a contractor .. for mice
20150824 22:54:27 <xranby_> sgothel: good night... one cornercase fixed
20150824 22:54:56 <xranby_> https://github.com/xranby/jogl/commit/5db1ba769b6a4b8f8e7148728bb1a82149c822ee
20150824 22:59:01 * xranby_ (~familjen@anon) Quit (Quit: Leaving.)
20150825 05:05:31 -jogamp- Continue @ http://jogamp.org/log/irc/jogamp_20150825050531.html