#jogamp @ irc.freenode.net - 20140708 05:05:30 (UTC)
20140708 05:05:30 -jogamp- Previous @ http://jogamp.org/log/irc/jogamp_20140707050530.html
20140708 05:05:30 -jogamp- This channel is logged @ http://jogamp.org/log/irc/jogamp_20140708050530.html
20140708 06:53:15 * monsieur_max (~maxime@anon) has joined #jogamp
20140708 07:23:40 * jvanek (jvanek@anon) has joined #jogamp
20140708 10:55:14 * zzuegg (~zzuegg@anon) has joined #jogamp
20140708 15:12:04 * kermyt (~kermyt@anon) Quit (Ping timeout: 240 seconds)
20140708 15:14:26 * kermyt (~kermyt@anon) has joined #jogamp
20140708 15:59:08 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20140708 16:24:38 * kermyt (~kermyt@anon) Quit (Ping timeout: 240 seconds)
20140708 16:28:31 * kermyt (~kermyt@anon) has joined #jogamp
20140708 16:39:50 * monsieur_max (~maxime@anon) has joined #jogamp
20140708 16:58:50 * jvanek (jvanek@anon) Quit (Quit: Leaving)
20140708 17:16:46 * [Mike] (~Mike]@anon) has joined #jogamp
20140708 17:23:13 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20140708 17:48:48 <sgothel> @Harvey/All: Walking through JOGL Findbugs issues .. a long path :)
20140708 17:48:58 <hharrison> Yep
20140708 17:49:08 <hharrison> I nibble away here and there
20140708 17:49:17 <sgothel> KUDOS to Harvey kicking me in that direction - already found a few pearls :)
20140708 17:49:40 <hharrison> I find it easiest to go by 'category' of bug, gets my mental pattern-matchers working
20140708 17:49:55 <sgothel> yup - I would like to annotate cases .. to force Findbugs to drop a few issue, I find OK ..
20140708 17:50:06 <hharrison> To be honest gluegen/jogl is really not that bad for such a large project, even before my previous commits
20140708 17:50:08 <sgothel> so I found the annotation thingy .. does it work well ?
20140708 17:50:25 <hharrison> Which annotations?
20140708 17:50:39 <hharrison> the ones that landed in Java8, or the findbugs specific ones?
20140708 17:50:41 <sgothel> http://findbugs.sourceforge.net/manual/annotations.html
20140708 17:51:06 <sgothel> like string comparison w/ == due to .intern() usage
20140708 17:51:07 <hharrison> You might want to check what landed in Java8, some of these were pulled in there....
20140708 17:51:18 <sgothel> so I like to say .. hey, we do use intern ..
20140708 17:51:20 <hharrison> and the names need a bit of tweaking, lemme go find it
20140708 17:51:40 <sgothel> java8 is n/a for JogAmp yet .. will take a while .. :)
20140708 17:52:14 <sgothel> (even though, ofc, jogamp is available to java8 :)
20140708 17:52:39 <hharrison> You can drop a jsr into the build which lets you use them for older releases
20140708 17:53:05 <hharrison> backwards and forwards compatible, and then findbugs and static checkers can use the same annotations
20140708 17:53:10 <sgothel> so it's really something needed for bytecode/compile-time .. hmm
20140708 17:53:31 <hharrison> Nope, just compile time checking, don't think it affects the bytecode
20140708 17:53:39 <sgothel> right, findbugs is bytecode level ..
20140708 17:53:48 <sgothel> so it must affect bytecode .. hmm
20140708 17:54:14 <rmk0> annotations like @Nonnull are retained at runtime
20140708 17:54:18 <sgothel> well .. keeping things alive .. adding old fashion comments comment for now ..
20140708 17:54:19 <hharrison> My googling is failing
20140708 17:54:44 <sgothel> later on, it might be nice to suppress some
20140708 17:54:51 <sgothel> maybe a special filter ..
20140708 18:31:32 <hharrison> You can set up diffs BTW
20140708 18:31:52 <hharrison> So generate it with everything turned on, then rerun over time looking for just the new warnings
20140708 18:32:00 <hharrison> Then you never get any 'worse'
20140708 18:32:15 <hharrison> and over time you chip away at the pile, that's what I've been mostly doing
20140708 19:36:42 * [Mike] (~Mike]@anon) Quit (Ping timeout: 240 seconds)
20140708 20:35:36 * [Mike] (~Mike]@anon) has joined #jogamp
20140708 20:52:16 * monsieur_max (~maxime@anon) has joined #jogamp
20140708 21:25:13 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20140708 22:04:03 <sgothel> @Harvey: First bunch of 'findbugs' review changes pushed .. oh well
20140708 22:24:20 <hharrison> gluegen, jogl? both?
20140708 22:24:40 <sgothel> jogl - gluegen was cleaned up by you already 'eh ? :)
20140708 22:25:21 <sgothel> low hanging fruits in jogl are eaten .. maybe there is room for long discussions now :)
20140708 22:25:54 <sgothel> who wants to pick the fruits for joal + jocl ?
20140708 22:26:04 <hharrison> commit 73d2a436f92e360126167d9ce09b3d9637204921
20140708 22:26:23 <hharrison> I looked at that for like 20 minutes a long time ago trying to decide if the fallthrough was intentional or not
20140708 22:26:52 <sgothel> it was like: dunno what to do .. now I grew a backbone and let's fail hard :)
20140708 22:27:05 <sgothel> same w/ hashCode() .. -> exception
20140708 22:28:11 <sgothel> b5910f18f0b82a8a1f6f6252dc19971d5e487f39
20140708 22:29:43 <sgothel> real fixed bugs: db25b1ba6575210741e485838d0882a1590125e6
20140708 22:30:04 <sgothel> 73d2a436f92e360126167d9ce09b3d9637204921
20140708 22:31:03 <sgothel> 9f0d303c2c325a5580054fe92f384c9a4076b065
20140708 22:31:19 <sgothel> (skipping the tests ..)
20140708 22:31:48 <sgothel> and triggering thinking about the many other 'warnings' .. but IMHO most are ok
20140708 22:32:12 <sgothel> I tuned to 'maximum detection' and enabled all pattern
20140708 22:32:54 <sgothel> but however hard I tried, I could get FB to show me that 'wrapper code' thingy for inner/outer class access :)
20140708 22:33:03 <sgothel> I changed some cases anyway
20140708 22:33:52 <sgothel> performance wise .. maybe we gained a very tiny bit :)
20140708 22:34:23 <sgothel> good exercise!
20140708 22:43:55 <hharrison> Looks like that detector might not be enabled in findbugs 3, too many false-positives
20140708 22:44:36 <hharrison> findbugs very much embraces only trying to give you ham to deal with
20140708 23:01:17 <sgothel> good night .. & laters
20140709 04:05:25 * [Mike] (~Mike]@anon) Quit ()
20140709 05:05:30 -jogamp- Continue @ http://jogamp.org/log/irc/jogamp_20140709050530.html