#jogamp @ irc.freenode.net - 20130609 17:24:16 (UTC)
20130609 17:24:16 -PircBot- This channel is logged @ http://jogamp.org/log/irc/jogamp_20130609172416.html
20130609 17:24:16 -holmes.freenode.net- *** Looking up your hostname...
20130609 17:24:16 -holmes.freenode.net- *** Checking Ident
20130609 17:24:16 -holmes.freenode.net- *** No Ident response
20130609 17:24:16 -holmes.freenode.net- *** Found your hostname
20130609 17:24:22 * CatOut (~jogamp@anon) has joined #jogamp
20130609 17:24:22 * Topic is 'http://jogamp.org | Hacking 3D Graphics, Multimedia and Processing across Devices'
20130609 17:24:22 * Set by rmk0 on 20130116 23:58:04
20130609 17:25:21 <sgothel> not bad :)
20130609 17:25:37 <rmk0> just realized i have no idea what my unix account password is. i can log in fine with ssh using my public key, but i don't remember ever setting a password
20130609 17:25:51 <sgothel> yes .. you need to for email ..
20130609 17:25:57 <rmk0> yep
20130609 17:26:05 <sgothel> if you like to use our imap / sendmail .. :)
20130609 17:26:13 <sgothel> ALWAYS: use SSL/TLS
20130609 17:26:32 <rmk0> yep... won't be sending mail, but i'd obviously like to receive anything sent to that address
20130609 17:26:38 <sgothel> ALWAYS: I have disabled non secure .. however .. if you find a LEAK, tell me
20130609 17:26:54 <sgothel> it's fwd .. ofc via aliases
20130609 17:27:21 <rmk0> ah yes
20130609 17:27:26 <rmk0> just spotted it
20130609 17:27:34 <rmk0> that's fine then
20130609 17:27:58 <sgothel> but we are a fine mail server w/ sieve and bogofilter :)
20130609 17:28:45 <sgothel> jenkins problem: I change the dependency .. but after boot .. it stays :(
20130609 17:28:47 <sgothel> https://jogamp.org/chuck/view/fwd/depgraph-view/?
20130609 17:35:56 <sgothel> ha
20130609 17:36:19 <sgothel> if you do the dependency on the child itself, i.e. build after other projects .. that one is gone :)
20130609 17:36:30 <sgothel> so you have to do it on the parent itself
20130609 17:36:38 <sgothel> (where it is also stored in the file)
20130609 17:36:56 <sgothel> which is ofc not so convenient .. since you have to edit 2 jobs
20130609 18:22:23 * hharrison (~chatzilla@anon) has joined #jogamp
20130609 18:33:18 <sgothel> thx harvey
20130609 18:33:44 <hharrison> No problem
20130609 18:33:54 <hharrison> I see you've been working on Jenkins today
20130609 18:34:03 <sgothel> yup .. so all good
20130609 18:34:13 <sgothel> we survived a restart as well :)
20130609 18:34:25 <sgothel> the URL/URI is also in .. etc
20130609 18:34:34 <hharrison> You make the systemd jump when changing machines?
20130609 18:34:37 <sgothel> will have to focus on security stuff after dinner now
20130609 18:34:44 <sgothel> nope .. debian7
20130609 18:34:48 <sgothel> or is it ?
20130609 18:34:56 <sgothel> probably :) .. didn't notice
20130609 18:35:16 <sgothel> all debian7 at home .. so if it's systemd .. it is :)
20130609 18:35:34 <hharrison> Hey, if you didn't notice, whichever is fine ;-)
20130609 18:36:13 <rmk0> debian won't move to systemd because of debian k/freebsd, apparently
20130609 18:36:19 <sgothel> so all is documented .. if anybody likes to replicate .. :)
20130609 18:36:21 <sgothel> ah ..
20130609 18:36:39 <rmk0> as my friend says... they don't want to make their user angry
20130609 18:36:55 <sgothel> as long we have a nice upgrade path .. all is great
20130609 18:37:07 <sgothel> and it seems debian has that very well
20130609 18:37:55 <sgothel> was good exercise to actually document the steps .. so next time it's easier to do and plan, i.e. now we know the outline etc
20130609 18:41:11 <sgothel> dinner .. laters
20130609 18:41:16 * rmk0 waves
20130609 20:42:50 * xranby (~xranby@anon) Quit (Ping timeout: 255 seconds)
20130609 20:43:53 * xranby (~xranby@anon) has joined #jogamp
20130609 20:44:06 <sgothel> Hi Xerxes
20130609 20:44:32 <sgothel> (me talking to bots :)
20130609 20:59:22 * hharrison (~chatzilla@anon) Quit (Ping timeout: 246 seconds)
20130609 20:59:30 * [Mike] (~Mike]@anon) has joined #jogamp
20130609 21:35:20 * monsieur_max (~maxime@anon) Quit (Quit: Leaving.)
20130609 23:41:33 <sgothel> Iff it's real .. kudos to Edward Snowden!
20130609 23:43:36 <sgothel> (Even though they assume his biggest fears are reality for a long time now already. The frog has been boiled quite a while and seems to feel pretty comfy .. nobody will give a dime (that is). Or .. in another way, at least here in the EU we already know that all data is being spied upon since WW2 :) .. well
20130609 23:46:36 <sgothel> tl;dr .. so our irc logging simply allows everybody to read such unencrypted blabbering of ours :) Still .. KUDO to Edward and all the others. What information would need to be exposed that the covert wars will end ? I guess none available .. since it's already out for all to read, right.
20130609 23:47:00 <sgothel> maybe I should use my blog for this, but I am too lazy to log in :)
20130609 23:47:27 <sgothel> <http://www.guardian.co.uk/world/2013/jun/09/edward-snowden-nsa-whistleblower-surveillance>
20130610 00:24:49 * hharrison (~chatzilla@anon) has joined #jogamp
20130610 00:35:03 * hharrison (~chatzilla@anon) Quit (Quit: ChatZilla 0.9.90 [Firefox 21.0/20130529083820])
20130610 00:50:54 <sgothel> William Binney <http://www.youtube.com/watch?v=TuET0kpHoyM>, .. etc ..
20130610 02:28:49 * hharrison (~chatzilla@anon) has joined #jogamp
20130610 02:29:14 <hharrison> sgothel: what's the preferred way to add junit tests to the jogl tree?
20130610 02:29:51 <sgothel> put them where they fit ? e.g. I add a 'sec' package in gluegen for security tests etc
20130610 02:30:11 <sgothel> so if you have a bunch new one .. for one cause .. IMHO new package, otherwise ..
20130610 02:30:18 <hharrison> Do you separate the paths for the test classes?
20130610 02:30:22 <sgothel> you see I have grouped the math stuff
20130610 02:30:36 <sgothel> depends ..
20130610 02:30:54 <sgothel> in jogl .. we have diff. package names (i.e. not colliding w/ JOGL) if you mean this
20130610 02:31:15 <sgothel> on gluegen .. we have them colliding .. some use that (package private) .. some not
20130610 02:31:17 <hharrison> Ahh, I see, that's exactly what I was asking
20130610 02:31:41 <sgothel> all what has 'Test' in class name gets tested!
20130610 02:31:51 <sgothel> NEWT .. etc .. all valid IDs for the unit test groups
20130610 02:31:52 <hharrison> Currently I tend to write them in exactly the same package that they test
20130610 02:32:00 <sgothel> pls check build-test.xml !
20130610 02:32:06 <hharrison> But in a separate path in the source tree
20130610 02:32:15 <hharrison> So my tree is like:
20130610 02:32:24 <sgothel> if you write a GLCaps test .. put them in our acore or caps
20130610 02:32:33 <hharrison> src/main/com/foo/Class.java
20130610 02:32:47 <hharrison> src/test/com/foo/TestClass.java
20130610 02:32:50 <sgothel> so pls try to use existing scheme .. if you are talking your stuff .. well
20130610 02:33:02 <hharrison> No no, I'll conform to what's there
20130610 02:33:04 <sgothel> I also try to make it obvious somewhat .. sure
20130610 02:33:15 <sgothel> i.e. in gluegen they match ..
20130610 02:33:37 <sgothel> jogl well .. sort of, even though diff. package domain, package name should speak for itself
20130610 02:34:01 <hharrison> I just ended up with test classes in my release jars too many times, nowadays I just do the maven-std layout above
20130610 02:34:03 <sgothel> so we usually have NOUI, AWT and NEWT in the test names .. and the special *.newt.* package ..
20130610 02:34:18 <sgothel> oh .. we separate product and test builds !
20130610 02:34:26 <sgothel> so 1st we build product
20130610 02:34:54 <sgothel> then the tests are being build separately (own source folder, using product jars .. and producing own JAR files)
20130610 02:35:05 <hharrison> I know, they just got into my jars due to oversight before, now with separate paths I never make that mistake...this is not so much a jogl thing
20130610 02:35:08 <sgothel> and NEWT tests shall be build w/o AWT .. etc
20130610 02:35:30 <sgothel> yup .. I see .. more general
20130610 02:35:49 <sgothel> w/ Android stuff as well, all separated
20130610 02:35:55 <sgothel> and SWT
20130610 02:36:03 <hharrison> At first the maven layout really annoyed, but I've come around to really liking it
20130610 02:36:31 <sgothel> well .. *thx to Mark* .. Maven is just in the end of our pipe right now :)
20130610 02:36:34 <hharrison> Anyways, off to commit a few unit tests
20130610 02:36:39 <sgothel> nice nice
20130610 02:36:53 <sgothel> me trying to produce SecurityManager enabled security tests ..
20130610 02:36:57 <sgothel> you have experience ?
20130610 02:37:08 <sgothel> i.e. try to load native library .. etc
20130610 02:37:22 <sgothel> so need to simulate that some JARs are signed ..
20130610 02:37:24 <hharrison> Never really had to fight that battle, sorry
20130610 02:37:27 <sgothel> and the test JARs not
20130610 02:37:38 <sgothel> ok .. will cont on that then
20130610 02:37:53 <sgothel> since we cannot sign them for real for a simple test :)
20130610 02:40:04 <hharrison> Quick question, I have some patches for the classes in com/jogamp/gluegen/jgram to have proper generic annotation
20130610 02:40:33 <hharrison> It's just to quiet the build for me, do you care to get them?
20130610 02:40:37 <sgothel> will we make the beauty contest for our code generating non exposed stuff ? :)
20130610 02:40:48 <sgothel> ofc
20130610 02:41:24 <hharrison> I seems like lots of that is code imported from antlr 2.7.7, just not sure it is worth deviating if it truly is a generated file
20130610 02:41:30 <sgothel> since you have hard earned credibility .. it will work and won't break our neck :)
20130610 02:41:45 <sgothel> your choice
20130610 02:42:04 <sgothel> I have thought about updating antlr as well .. but you know :)
20130610 02:42:08 <hharrison> Oh yeah, these kind of changes it's super easy to verify as you can literally diff the bytecode before/after (thanks type erasure)
20130610 02:42:24 <sgothel> in jogl I have these diff scripts as well
20130610 02:42:39 <sgothel> jogl/make/scripts/cmpOld2New.sh
20130610 02:42:50 <sgothel> comparing the generated GL OO stuff
20130610 02:43:06 <sgothel> especially if changing the GL inheritance .. methods, enums etc
20130610 02:43:25 <sgothel> a zero diff is always a good thing then :)
20130610 02:44:14 <hharrison> OK, I'll split them into small chunks you can pull at your leisure this week
20130610 02:44:47 <hharrison> fedora upgrade has broken my findbugs install, still figuring that out
20130610 02:45:08 <hharrison> grumble, grumble
20130610 02:48:21 <sgothel> the antlr cleanup ? nope - pls make it one patch .. and I will test it only once then :)
20130610 02:48:33 <hharrison> If you wish
20130610 02:48:47 <hharrison> I had it done by logical type
20130610 02:48:54 <hharrison> remove unneeded casts
20130610 02:48:59 <sgothel> what was your server problem btw ? anything here for me to learn or 'just' a hw failure ?
20130610 02:49:05 <hharrison> add missing annoatations to hashmap/lists
20130610 02:49:16 <hharrison> Not a server, a laptop
20130610 02:49:21 <sgothel> well
20130610 02:49:57 <hharrison> Turned out to be a thermal shutdown, temp driver was returnign garbage, machine thought it was 105C
20130610 02:50:47 <hharrison> Finally found the shutdown logs with the reason it kept shutting off
20130610 02:51:10 <sgothel> hohoho
20130610 02:51:17 <sgothel> which CPU ?
20130610 02:51:41 <sgothel> IMHO the new ones (at least AMD) shut off some cores automatically .. and run on low frequencies
20130610 02:51:53 <hharrison> Intel i5....sandybridge
20130610 02:51:58 <sgothel> if > 90C
20130610 02:52:16 <sgothel> hmm
20130610 02:52:37 <hharrison> Aye, this one does that, but I had an old cpufreq daemon left over that really thought it was important to do something
20130610 02:52:47 <sgothel> :)
20130610 02:53:11 <hharrison> Purely user config error ;-)
20130610 05:05:16 -CatOut- Continue @ http://jogamp.org/log/irc/jogamp_20130610050516.html