#jogamp @ irc.freenode.net - 20131009 05:05:31 (UTC)


20131009 05:05:31 -jogamp- Previous @ http://jogamp.org/log/irc/jogamp_20131008050531.html
20131009 05:05:31 -jogamp- This channel is logged @ http://jogamp.org/log/irc/jogamp_20131009050531.html
20131009 07:10:35 * Eclesia (jsorel@anon) has joined #jogamp
20131009 07:10:40 <Eclesia> hi
20131009 07:31:31 * odin_ (~Odin@anon) Quit (Remote host closed the connection)
20131009 08:01:06 * monsieur_max (~maxime@anon) has joined #jogamp
20131009 08:52:39 * odin_ (~Odin@anon) has joined #jogamp
20131009 10:12:56 <xranby> This years Nobel prize in chemistry is rewarded to rational use of computer models and graphics visualisation effort, to merge the quantum and classical world in chemistry http://www.youtube.com/watch?v=csxKNLm2bhM
20131009 10:13:50 <Eclesia> hi xranby
20131009 10:13:55 <xranby> 2013 #NobelPrize in #Chemistry press release "The computer-your Virgil in the world of atoms" http://www.nobelprize.org/nobel_prizes/chemistry/laureates/2013/
20131009 10:14:27 <xranby> Eclesia: hi, good progress with your animated gif video decoder!
20131009 11:06:03 <sgothel> Hi All .. good morning
20131009 11:06:54 <sgothel> re Bug 821 - I like to remove the !VBO buffer methods of the glDraw.. etc methods for GL2ES2 - core profiles
20131009 11:07:10 <sgothel> while adding such test in the buffer validation as well
20131009 11:07:10 <xranby> i vote yes!
20131009 11:07:36 <xranby> sgothel: have you talked with gouessej about this?
20131009 11:07:44 <sgothel> we could leave it in GLES2 itself .. for correctness - have to see how 'ugly' that is
20131009 11:08:16 <sgothel> Julien likes to have those avail in the compat profiles - which we won't touch ofc
20131009 11:08:24 <sgothel> (his comment - np)
20131009 11:08:38 <sgothel> i.e. GL2, GL3bc .. etc
20131009 11:08:51 <xranby> is it possible to request something like GL2ES2bc
20131009 11:09:02 <sgothel> now :)
20131009 11:09:11 <sgothel> I guess we have enough of those already :)
20131009 11:09:21 <sgothel> *no* I meant .. hehe
20131009 11:09:58 <sgothel> in theory yes .. but I don't like it .. so you or somebody else must convince me (code, tests .. etc)
20131009 11:10:30 <sgothel> most important thing to convince me . no redundancy
20131009 11:10:47 <xranby> how about adding a GL2ES3 ?
20131009 11:10:57 <xranby> sorry i mean
20131009 11:10:59 <xranby> GL3ES2
20131009 11:11:12 <sgothel> opening UML ..
20131009 11:11:46 <sgothel> btw: a common compat profile is GL2 itself already!
20131009 11:11:59 <sgothel> http://jogamp.org/jogl/doc/uml/html/fig128069.png
20131009 11:12:00 <xranby> thus then we have a separation GL3ES2 require core profile and removes the !VBO while GL2ES2 prefers backward compatible profile
20131009 11:12:08 <xranby> hmm but.. GL2 dont work on mobile :/
20131009 11:12:24 <sgothel> I only see 'core vs compat'
20131009 11:12:34 <sgothel> bc -> compat
20131009 11:13:16 <sgothel> so all core would have !VBO removed in interfaces
20131009 11:13:24 <sgothel> maybe except GLES2
20131009 11:14:15 <sgothel> so compat is only: GL2 GL3bc and GL4bc - to be clear
20131009 11:15:31 <sgothel> OSX/CALayer state: best I could do - remaining JSlide* thingy in Applets - me dunno
20131009 11:16:09 <sgothel> reparenting/fullscreen issues fixed as well .. only on X11 the CHILD->TOP w/ browser in applet is initially _behind_ browser - a WM thingy :-/ .. well
20131009 11:21:07 <xranby> sgothel: i think immediate mode should exist someplace in this UML diagram
20131009 11:21:26 <xranby> in the borderland between fixed function and programmable pipeline
20131009 11:21:58 <sgothel> hu?
20131009 11:22:04 <sgothel> it's ffp
20131009 11:22:15 <xranby> i will draw a picture and.. give me a minute
20131009 11:22:57 <sgothel> im = ffp - es1 .. to be correct :)
20131009 11:23:18 <xranby> right
20131009 11:23:40 <xranby> and inside im we may find the GL2ES2bc
20131009 11:23:43 <sgothel> im -> don't use it .. ever :)
20131009 11:24:06 <sgothel> what is that purpose ? it's in GL2
20131009 11:24:18 <xranby> GL2ES2 <- GL2ES2bc <- GLES2 for mobile compatibility
20131009 11:24:23 <sgothel> the common denominator of im
20131009 11:24:30 <sgothel> hu?
20131009 11:24:33 <xranby> if you use GL2 you run into a not a GL2 exception
20131009 11:24:47 <sgothel> there is no IM w/o GL2
20131009 11:25:29 <xranby> hu? i can use the im with gles2 on all arm chips i have around me
20131009 11:25:41 <sgothel> ok .. define IM pls
20131009 11:25:50 <xranby> sure.. let me draw this graph...
20131009 11:25:57 <sgothel> I mean in GL funcs ..
20131009 11:26:07 <sgothel> glVertex3f(..) ..
20131009 11:26:18 <sgothel> begin/end
20131009 11:26:39 <sgothel> that is only avail in GL2 and higher compat
20131009 11:27:09 <xranby> thats fixed function
20131009 11:27:22 <xranby> thats only supported by gles1 etc...
20131009 11:27:57 <sgothel> no - fixed function means fixed rendering features like glLIght .. etc
20131009 11:28:04 <sgothel> or SMOOTH .. etc
20131009 11:28:12 <sgothel> i.e. fixed build-in shaders
20131009 11:28:32 <sgothel> IM means you can assemble a geometry (VBO) .. immediately
20131009 11:28:43 <sgothel> AFAIK ..
20131009 11:28:43 <xranby> ok so what do people call the !VBO glPointerAttrib functions that you can pass Buffer directly?
20131009 11:29:12 <sgothel> the VBO buffer offset you mean ?
20131009 11:29:34 <sgothel> the 'long' value is the offset to current bound VBO of desired VBO-type
20131009 11:29:58 <xranby> the two fuctions that do *not* take an offset and where you instead passes a buffer
20131009 11:30:20 <xranby> the one you want to remove
20131009 11:30:48 <sgothel> ah you mean the name / type - just the non VBO variant .. no special 'name' .. just a memory management thingy
20131009 11:31:00 <sgothel> i.e. !VBO means automatic memory management
20131009 11:31:10 <sgothel> VBO - manual memory management
20131009 11:31:31 <xranby> ok so the GL2ES2bc support automatic memory management
20131009 11:31:32 <sgothel> the !VBO would actually do the same .. sort of .. (alloc GPU buffer .. copy .. etc)
20131009 11:32:05 <sgothel> but I dislike adding an interface for it .. since it will only prevail in ES1, ES2 and GL2
20131009 11:32:23 <sgothel> however .. we could do a sub-interface .. like GLPointerIf ..
20131009 11:32:41 <sgothel> GLPointerFunc
20131009 11:32:48 <sgothel> but IMHO overkill
20131009 11:33:22 <sgothel> plus the set of functions varies .. GL3, GL4 ..
20131009 11:33:55 <sgothel> I only added GLPointerFunc interface a long time ago .. for our FFP impl.
20131009 11:34:30 <sgothel> now we have added tons of new buffer operations in GL3 and GL4 .. where you would need to create subinterfaces etc
20131009 11:38:58 <sgothel> @Mark: Do you approve the removal of !VBO on core interfaces GL2ES2+ and impl. GLES3, GL3, .. ?
20131009 11:39:27 <rmk0> 'lo
20131009 11:39:31 <rmk0> yeah, i'm fine with that
20131009 11:39:34 <sgothel> @Mark: So by now .. I am afraid the t-shirts were stolen/lost if you haven't received them by now - damn it
20131009 11:39:43 <rmk0> erk, yes, still not received them
20131009 11:39:47 <rmk0> bleh!
20131009 11:40:14 <sgothel> wonder if they would have treaten them better if insured .. grrr
20131009 11:41:21 <rmk0> "CONTENTS ARE ELECTRONICALLY TAGGED"
20131009 11:42:52 <sgothel> the moment where I could slap myself .. alot .. I will ask the provider here to backtrack .. if possible, and if all fails .. well, you get my shirts :)
20131009 11:43:13 <rmk0> hehe, don't put yourself to too much trouble
20131009 11:43:28 <rmk0> there'll be more shirts
20131009 11:44:16 <sgothel> my father just had the same experience .. lost a package, same thing, sentimental value .. work lost
20131009 11:45:13 <sgothel> we were kidding .. couldn't they put it back to the 'mail-circle' when realizing it's of no monetary value ? :)
20131009 11:46:16 <sgothel> yeah .. everybody can print their own jogamp shirt now - files are public at http://jogamp.org/doc/siggraph2013/ :)
20131009 11:57:50 <xranby> sgothel: http://labb.zafena.se/jogamp/UML/fig128069-GL2ES2bc.png
20131009 11:57:59 <xranby> so this was what i had in mind
20131009 11:59:17 <xranby> hmm
20131009 12:00:15 <sgothel> GL2ES1 :)
20131009 12:00:34 <sgothel> not quite complete (GLES2 would be missing .. well .. and other GL3/GL4 ..)
20131009 12:00:44 <sgothel> it's a dying path anyways ..
20131009 12:02:11 <sgothel> _however_ cute UML patching :)
20131009 12:17:12 * Eclesia doesn't think VBO should be part of the diagram. my 2 cents
20131009 12:23:44 * masterzen (~masterzen@anon) has joined #jogamp
20131009 12:24:40 <masterzen> sgothel: ping, regarding our issue with galaxy tab 3
20131009 12:39:01 <sgothel> Hi Brice(?) - Whats that Galaxy thingy ?
20131009 12:39:16 <sgothel> I may have missed that issue ?
20131009 12:50:23 <sgothel> masterzen: ?
20131009 12:51:06 <masterzen> sgothel: ericb should have sent you an email on an issue we have with jogl on the galaxy tab 3 10.1 (actually it's more a gluegen issue)
20131009 12:51:22 <masterzen> sgothel: BTW, hi :)
20131009 12:52:25 <masterzen> the galaxy tab 3 10.1 is an atom based tablet, running arm program, but when we launch our game on it, we have an exception that prevent the launch
20131009 12:52:52 * masterzen is asking ericb for the various details :)
20131009 12:53:29 <sgothel> email might be lost, pls resend .. sorry
20131009 12:54:01 <masterzen> we'll open a bugzilla ticket, that will be easier to track
20131009 12:54:23 <masterzen> we also have another bug to report on android too
20131009 12:54:29 <sgothel> intel atom running arm code ? interesting
20131009 12:54:37 <masterzen> yeah, we just discovered this
20131009 12:55:04 <masterzen> apparently that confuses the part of the code that tries to infer the size of ints, and so on
20131009 12:55:18 <sgothel> I can imaging .. :)
20131009 12:55:23 <masterzen> :)
20131009 12:55:23 <sgothel> *imagine*
20131009 12:55:43 <sgothel> hmm .. no native intel atom mode ?
20131009 12:55:55 * ebrayet (~Adium@anon) has joined #jogamp
20131009 12:55:58 <sgothel> maybe send me such a device .. curious
20131009 12:57:09 <sgothel> sure we should have an android intel build .. soon I guess, since we shall expect such devices on the market
20131009 12:57:54 <xranby> sgothel: thank you for the suggestions: http://labb.zafena.se/jogamp/UML/fig128069-GLbc-GL2ES2bc.png
20131009 12:58:31 <xranby> Eclesia: renamed !VBO -> GLbc
20131009 12:59:09 <masterzen> we have another app that is natively built for arm, and it runs on this tablet, so go figure
20131009 12:59:17 <sgothel> @Xerxes: GL2ES2bc is wrong, no relation to GL2ES2 - otherwise GL2/ES3 is broken
20131009 13:00:33 <sgothel> @Brice - hehe, yes will 'go figure' with details on bug report - maybe I need the device .. or we coop closely (you do patch and validation w/ my help)
20131009 13:01:04 <masterzen> sgothel: yep, no issue. We're going to open the tickets ASAP.
20131009 15:28:39 * ebrayet (~Adium@anon) Quit (Ping timeout: 252 seconds)
20131009 15:55:50 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20131009 16:02:47 * Eclesia (jsorel@anon) Quit (Quit: Leaving.)
20131009 16:42:40 * monsieur_max (~maxime@anon) has joined #jogamp
20131009 16:54:56 * kermyt (~kermyt@anon) Quit (Ping timeout: 265 seconds)
20131009 17:01:42 * kermyt (~kermyt@anon) has joined #jogamp
20131009 18:08:34 * [Mike] (~Mike]@anon) has joined #jogamp
20131009 19:12:24 * Eclesia (~eclesia@anon) has joined #jogamp
20131009 19:12:30 <Eclesia> hi
20131009 20:28:51 <Eclesia> do I forget something ? VBO, VAO, FBO, PBO, RBO, UBO, IBO
20131009 21:12:00 * Eclesia (~eclesia@anon) Quit (Quit: Leaving.)
20131009 21:19:53 * ebrayet (~Adium@anon) has joined #jogamp
20131009 21:24:27 <sgothel> https://jogamp.org/bugzilla/show_bug.cgi?id=852
20131009 21:26:39 * monsieur_max (~maxime@anon) has left #jogamp
20131009 22:29:25 * ebrayet (~Adium@anon) Quit (Quit: Leaving.)
20131009 22:32:14 * ebrayet (~Adium@anon) has joined #jogamp
20131009 22:49:54 * ebrayet (~Adium@anon) Quit (Quit: Leaving.)
20131010 01:12:34 * ruzu (~ruzu@anon) has joined #jogamp
20131010 04:45:03 * [Mike] (~Mike]@anon) Quit ()
20131010 05:05:32 -jogamp- Continue @ http://jogamp.org/log/irc/jogamp_20131010050532.html