#jogamp @ irc.freenode.net - 20140510 05:05:36 (UTC)


20140510 05:05:36 -jogamp- Previous @ http://jogamp.org/log/irc/jogamp_20140509050535.html
20140510 05:05:36 -jogamp- This channel is logged @ http://jogamp.org/log/irc/jogamp_20140510050536.html
20140510 05:51:08 * [Mike] (~Mike]@anon) Quit ()
20140510 06:49:14 * monsieur_max (~maxime@anon) has joined #jogamp
20140510 07:28:37 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20140510 07:29:23 * monsieur_max (~maxime@anon) has joined #jogamp
20140510 10:06:18 * hija (~hija@anon) has joined #jogamp
20140510 10:07:54 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20140510 10:13:11 * monsieur_max (~maxime@anon) has joined #jogamp
20140510 10:37:39 * hija_ (~hija@anon) has joined #jogamp
20140510 10:40:20 * hija (~hija@anon) Quit (Ping timeout: 240 seconds)
20140510 10:40:20 * hija_ is now known as hija
20140510 10:51:37 <sgothel> http://www.cafc.uscourts.gov/images/stories/opinions-orders/13-1021.Opinion.5-7-2014.1.PDF (Oracle's won the appeal re. API 'copyrightable')
20140510 10:53:58 <rmk0> lo
20140510 10:54:03 <rmk0> what are the consequences of this?
20140510 10:54:39 <sgothel> reading ..
20140510 10:55:30 <sgothel> however .. this appeal decision is that those packages can be copyrightable .. or at the very least that the _reasons_ Google gave for non copyrightability are not sufficient
20140510 10:56:13 <sgothel> the f***ing worst case: all interfaces can be copyrightable and license fees to be enforced .. surely .. that won't happen, since otherwise ..
20140510 10:56:28 <sgothel> well, in Europe it can't anyways
20140510 10:57:22 <sgothel> p17 .. rebuttal starts
20140510 10:57:53 * hija (~hija@anon) Quit (Quit: hija)
20140510 10:58:08 <sgothel> ofc .. this is _not_ just about 'Java'
20140510 10:58:38 <rmk0> urhur
20140510 10:59:23 <sgothel> well .. IMHO Google should simply state that Sun didn't sue them and practically gave them consent .. and such decision cannot be changed simply by being bought
20140510 11:00:14 <sgothel> for us .. the copyrightability question is very important .. and since it's out of the bag, they should fight it to the end .. IMHO
20140510 11:02:22 <sgothel> the quote on p18 is pretty much what is valid in Europe (no math, procedure ..)
20140510 11:05:22 <sgothel> another idea surfaced was 'the API derived on work from others' and hence is not solely of Sun/Oracle copyright - yeah, all ridiculous
20140510 11:09:30 <sgothel> 'literal copying of declaring code'
20140510 11:09:59 <sgothel> p27
20140510 11:11:04 <sgothel> (means: Linux copyright holder could sue Google :)
20140510 11:11:26 <sgothel> (header files of bionic etc ..)
20140510 11:15:29 <sgothel> the real pain: p30 'merger doctrine does not apply'
20140510 11:15:51 <sgothel> (.. only one way of doing things bars copyrightability)
20140510 11:16:54 <sgothel> seem they do not respect _compatibility_
20140510 11:17:09 <sgothel> IMHO void
20140510 11:18:27 <rmk0> sounds wonderful
20140510 11:20:13 <sgothel> cynical yes - but truly horrible - i.e. they don't allow to copy the package structure and naming etc (not just verbatim copying) to achieve compatibility
20140510 11:20:36 <sgothel> p30-p33
20140510 11:22:23 <sgothel> p35 'scenes a faire' .. related to compatibility
20140510 11:23:17 <sgothel> rejected since Google failed .. show proper evidence in trial - 'great'
20140510 11:24:19 <sgothel> (not really covering compatibility though)
20140510 11:25:20 <sgothel> p45 'interoperability arguments'
20140510 11:26:09 <sgothel> -> fair use not copyrightability issue (ongoing argument)
20140510 11:27:56 <sgothel> (conclusion @ p50 - 2)
20140510 11:31:43 <sgothel> [p53 - .. - fair use - p59 -> rejected, no fair use]
20140510 11:34:00 <sgothel> p61: still fair use may apply due to interoperability, but no, since harming the market .. etc etc :)
20140510 11:35:25 <sgothel> p60: Google shall cont. w/ 'fair use' defense ..
20140510 11:36:29 <sgothel> Conclusion: Only trust a software vendor located in Europe :)
20140510 11:36:49 <sgothel> (they simply cannot sue you due to these things .. hehe)
20140510 11:37:49 <sgothel> btw .. I tested llvm-vmkit master and the new mcjit (sic) branch
20140510 11:38:04 <sgothel> the latter is incomplete .. could not test
20140510 11:38:35 <sgothel> master branch works sort of, but could not load libGL (having another llvm instance)
20140510 11:38:50 <sgothel> I could not produce an executable w/ master branch, sadly
20140510 11:40:03 <rmk0> the trunk doesn't seem to have had any commits for 5 months
20140510 11:40:09 <sgothel> (vmkit shall allow you to run it via j3, or package bundle the vm to exe etc)
20140510 11:40:34 <sgothel> the mcjit branch is worked on heavily
20140510 11:41:25 <rmk0> what's the "mc" stand for?
20140510 11:42:10 <sgothel> it's some new 'jit' from llvm ..
20140510 11:42:39 <sgothel> so next test round will go to robovm ..
20140510 11:42:47 <sgothel> then contacting both devs ..
20140510 11:46:16 <sgothel> well .. if shit hits the fan, good that we only use 'core' API functions .. which can be refactored :)
20140510 11:47:49 <sgothel> I guess this case will go for a few more years ..
20140510 11:51:40 <sgothel> worst case: Java API usage beyond OpenJDK w/o Oracle's blessing is at risk
20140510 11:52:15 <sgothel> *usage -> modifications/re-implementations/...*
20140510 11:52:43 <sgothel> [2]: Any API modifications/re-implementations w/o blessing
20140510 11:55:15 <sgothel> - or - go the GPLv2 license way (OpenJDK), even though this is mentioned partially in this decision .. dunno
20140510 11:55:37 <sgothel> GPLv2 should cover situation like: partial re-implementation .. etc
20140510 11:56:08 <sgothel> good: patent waiver seems to be redundant .. i.e. Oracle lost this one in this battle
20140510 11:58:29 <sgothel> Q: Java8's javac annotation processing .. chicken and egg problem ..
20140510 11:58:53 <sgothel> We have a java class, producing java classes based on a c-header file
20140510 11:59:06 <sgothel> produced via that 'anno. proc.'
20140510 11:59:20 <sgothel> said java class imports the produced java classes :)
20140510 11:59:31 <sgothel> javaX <= java7 worked well
20140510 11:59:51 <sgothel> now the not-yet-existing class makes javac fail
20140510 12:00:08 <sgothel> even thought we say '-proc:none' -> no java compilation
20140510 12:00:14 <sgothel> ideas ?
20140510 12:00:25 <sgothel> ^^ Bug 923
20140510 12:00:50 <sgothel> com.jogamp.gluegen.test.junit.structgen.TestStructGen01
20140510 12:00:52 * rmk0 eyes suspiciously
20140510 12:01:15 <sgothel> (me currently cleaning up build and stuff .. incl. java8 compatibility)
20140510 12:01:37 <sgothel> (at least works out of the box w/ openjdk on debian now)
20140510 12:02:15 <sgothel> yes, no openjdk8 on debian yet
20140510 12:05:02 <sgothel> would be ugly if production and usage of generated classes must be split - i.e. this would remove the 'fun' of using the anno.proc. in the 1st place
20140510 12:14:26 <sgothel> CStruct <- can we make that static ? me not an anno.proc. expert ..
20140510 12:36:50 <rmk0> have no experience processing annotations at all outside of using reflection
20140510 12:38:16 <sgothel> using a dummy 'BuildStruct01' class kicking off anno.proc. _not_ using the new type but a dummy boolean works - had to change the anno.proc. code, i.e. favoring the given type name ..
20140510 12:38:18 <sgothel> ugly
20140510 12:45:22 * hija (~hija@anon) has joined #jogamp
20140510 12:57:50 * hija (~hija@anon) Quit (Quit: hija)
20140510 13:06:27 * hija (~hija@anon) has joined #jogamp
20140510 13:24:38 * hija (~hija@anon) Quit (Quit: hija)
20140510 14:29:52 * hija (~hija@anon) has joined #jogamp
20140510 15:09:03 * hija_ (~hija@anon) has joined #jogamp
20140510 15:09:14 * hija (~hija@anon) Quit (Ping timeout: 240 seconds)
20140510 15:09:14 * hija_ is now known as hija
20140510 15:51:16 * rmk0 drops lwjgl support
20140510 16:28:56 * hija_ (~hija@anon) has joined #jogamp
20140510 16:29:26 * hija (~hija@anon) Quit (Ping timeout: 276 seconds)
20140510 16:29:26 * hija_ is now known as hija
20140510 18:49:12 * zaphos (~Matthew@anon) has joined #jogamp
20140510 19:18:40 * zaphos (~Matthew@anon) Quit (Quit: Leaving.)
20140510 20:00:49 -mist- [Global Notice] What are you doing on IRC? Go watch Eurovision! ;)
20140510 21:14:20 -kloeri- [Global Notice] Hi all. I'm going to do some very quick maintainence related to nickserv, chanserv and friends. Services downtime should be very brief. Thanks for using freenode.
20140510 23:19:14 * monsieur_max (~maxime@anon) Quit (Ping timeout: 240 seconds)
20140511 00:56:46 * [Mike] (~Mike]@anon) has joined #jogamp
20140511 01:43:48 <sgothel> CStruct annotation issue solved <http://jogamp.org/git/?p=gluegen.git;a=commit;h=1e53a38eb06aa8ff50660fdf7bd8570b27f27c56>
20140511 01:44:12 <sgothel> best practice now: use the package-info annotation avoiding java8 bug
20140511 01:44:17 * hija_ (~hija@anon) has joined #jogamp
20140511 01:45:14 * hija (~hija@anon) Quit (Ping timeout: 240 seconds)
20140511 01:45:14 * hija_ is now known as hija
20140511 05:05:36 -jogamp- Continue @ http://jogamp.org/log/irc/jogamp_20140511050536.html