#jogamp @ irc.freenode.net - 20130128 05:06:15 (UTC)
20130128 05:06:15 -CatOut- Previous @ http://jogamp.org/log/irc/jogamp_20130127050601.html
20130128 05:06:15 -CatOut- This channel is logged @ http://jogamp.org/log/irc/jogamp_20130128050615.html
20130128 06:26:55 * DemoscenePassiv (~Lutsche@anon) has joined #jogamp
20130128 07:17:36 * DemoscenePassiv (~Lutsche@anon) Quit (Ping timeout: 252 seconds)
20130128 08:08:31 * EricB1 (~Adium@anon) has joined #jogamp
20130128 09:32:36 * EricB1 (~Adium@anon) Quit (Quit: Leaving.)
20130128 09:47:41 <xranby1> Titanoum: i fixed it <-- did you change anything in jogl to fix it?
20130128 09:57:14 * EricB1 (~Adium@anon) has joined #jogamp
20130128 11:31:36 * EricB1 (~Adium@anon) Quit (Quit: Leaving.)
20130128 11:34:33 * sgothel (~sven@anon) has joined #jogamp
20130128 11:34:36 * ChanServ sets mode +v sgothel
20130128 11:58:48 * EricB1 (~Adium@anon) has joined #jogamp
20130128 12:16:33 <xranby1> http://youtu.be/51-Jfh6ADH0 - Olafur Ragnar Grimsson Iceland president 'Let banks go bankrupt'
20130128 12:17:42 <sgothel> yup .. kudos, and guess what: THEY WERE F** RIGHT - i.e. now in hindsight - guess this is a chapter in history to not allow banks to blackmail you
20130128 12:18:01 <sgothel> I have heard about the Iceland analysis .. I guess last year ?
20130128 12:18:41 <sgothel> (sorry folks .. this politics stuff always gets me :)
20130128 12:20:49 <sgothel> http://www.yesmagazine.org/new-economy/iceland-busts-the-banksters
20130128 12:23:08 <sgothel> .. sadly people are scared of loosing virtual currency / changing that system etc etc. Artificial currency w/o intrinsic value controlled by the non producing people .. fiat currency crap :)
20130128 12:23:54 <sgothel> 10 lines of code for a loaf of bread anyone ? :)
20130128 12:34:12 <xranby1> you can have the bread for a bitcoin
20130128 12:35:11 <sgothel> still have to figure out the bc trading w/o fiat currency conversion .. but I read more about it .. and it seems bc has an intrinsic value properly based on _deflation_
20130128 12:40:58 <xranby1> sgothel: you can try this excange instead http://coinabul.com/index.php/
20130128 12:42:16 <sgothel> thx - holy moly .. I guess w/ those prices, we better keep our BC .. when we start using them :)
20130128 13:33:07 <sgothel> :)
20130128 13:35:20 <sgothel> since we sometimes do chat about other matters than jogamp - DISCLAIMER: The views and conclusions contained in the software and documentation are those of the
20130128 13:35:20 <sgothel> authors and should not be interpreted as representing official policies, either expressed
20130128 13:35:20 <sgothel> or implied, of JogAmp Community.
20130128 13:50:44 <EricB1> @sgothel Hi Sven. I just submitted a pull request, which actually addresses two issues (bugs #677 and #668)
20130128 13:51:09 <EricB1> These are specific to Android
20130128 13:51:24 <EricB1> https://github.com/sgothel/jogl/pull/57
20130128 13:53:07 <xranby1> EricB1: cool, for legal reasons it would be great if you used your real name in the commits
20130128 13:54:13 <EricB1> Ok
20130128 13:54:30 <xranby1> in the author field of the git commit
20130128 14:03:27 <EricB1> I updated my config. Future commits will have my name and e-mail
20130128 14:20:39 <sgothel> Can you re-commit under your real name please?
20130128 14:20:49 <sgothel> .. and thank you Erik .. reading now
20130128 14:22:47 <xranby1> hmm hmm .. btw when i pulled your git a second ago then the author field looks ok!
20130128 14:23:28 <sgothel> will add his to my set of remotes ..
20130128 14:23:43 <xranby1> Author: Eric Brayet <ericb@daysofwonder.com>
20130128 14:23:45 <xranby1> nice
20130128 14:24:15 <xranby1> EricB1: sorry .. i think github simply hide it in their view , it looks great from gitk
20130128 14:24:43 <sgothel> yup - confirmed
20130128 14:25:52 <xranby1> EricB1: so kudos for your pull request
20130128 14:26:04 <xranby1> i will try code review asap
20130128 14:30:55 <xranby1> EricB1: are all files included for the 677 fix?
20130128 14:31:31 <EricB1> Yes : this is a 1 file fix
20130128 14:31:32 <sgothel> @Erik: Kudos for finding the impl. detail about intercepting the BACK key! Your patch however, is not complete, i.e. IMHO we should not just add an interface for the user, but deliver the intercepted BACK key to our normal key event flow ..
20130128 14:32:02 <EricB1> Well, we had a discussion about this
20130128 14:32:11 <sgothel> if you agree - then it would be nice to simply use private listener w/o interface and feed event to the translator ..
20130128 14:32:14 <sgothel> but?
20130128 14:32:29 <xranby1> i expected to see some implementation of onBackPressed someplace :)
20130128 14:32:34 <xranby1> i only see the interface
20130128 14:32:35 <EricB1> The thing I was worried about was not to add side effects
20130128 14:32:51 <sgothel> well - we are Android agnostic and there is a BACK key in NEWT's KeyEvents already
20130128 14:33:07 <EricB1> Adding just the listnere was a way to modify less things (no change in the wrkflow)
20130128 14:33:19 <EricB1> Ok then
20130128 14:33:53 <EricB1> And Newt's BACK is made to map such devices's Back buttons, right ?
20130128 14:33:55 <sgothel> thx .. well, if we want something public for such case .. it would become difficult to expose it properly
20130128 14:34:14 <sgothel> let's see ..
20130128 14:34:26 <sgothel> i.e. semantics ..
20130128 14:34:32 <sgothel> ESC ?
20130128 14:35:19 <EricB1> *starting the app on my Kindle*
20130128 14:35:21 <sgothel> we have no VK_BACK ..
20130128 14:35:47 <sgothel> VK_ESCAPE
20130128 14:36:05 <EricB1> And there is no Esc button on the softKeyboard.
20130128 14:36:12 <EricB1> Let me check on another device.
20130128 14:36:39 <sgothel> so we could use it, since IMHO .. semantics are same .. but let us/me read more
20130128 14:38:14 <EricB1> *other device's batteries are empty …*
20130128 14:38:15 <sgothel> Currently the only key that is detected is the "back" key (KEY_ESCAPE). http://www.monkeycoder.co.nz/Community/posts.php?topic=2284
20130128 14:38:49 <EricB1> *reading*
20130128 14:39:04 <sgothel> Android KeyEvent.KEYCODE_ESCAPE ..
20130128 14:39:09 <sgothel> so they have it - collision
20130128 14:39:38 <EricB1> yep
20130128 14:40:01 <sgothel> we can match android HOME to our HOME ..
20130128 14:40:19 <EricB1> There is a Home button
20130128 14:40:44 <sgothel> yup .. i.e. also can map their 'HOME' to NEWT 'HOME'
20130128 14:40:56 <EricB1> True
20130128 14:41:11 <sgothel> how about VK_END ?
20130128 14:41:57 <sgothel> they only have 'ENDCALL' ..
20130128 14:42:18 <sgothel> so Android BACK -> NEWT END ? Makes sense ?
20130128 14:42:27 <sgothel> i.e. END this current thingy ?
20130128 14:43:17 <sgothel> ofc .. we could also introduce a new VK token .. if we must, .. not proper semantical match
20130128 14:43:22 <xranby1> please press the end to continue
20130128 14:43:26 <sgothel> :)
20130128 14:43:34 <EricB1> That sounds sensible
20130128 14:43:50 <sgothel> na .. it's more like: press END to END current 'layer' .. (i.e. back to prev. layer)
20130128 14:44:09 <EricB1> That's the way I understand it
20130128 14:44:13 <xranby1> what do VK stand for? virtual key ?
20130128 14:44:20 <sgothel> yup
20130128 14:44:27 <xranby1> ok
20130128 14:45:07 <sgothel> question: what happens if you press android BACK w/ keyboard up ? the keyboard disappears [already in current code] right ?
20130128 14:45:46 <sgothel> I have to check w/ demo .. hmm
20130128 14:45:56 <EricB1> Currently, nothing happens
20130128 14:46:10 <EricB1> The event does not reach our classes
20130128 14:46:19 <EricB1> It's consumed by the softKeyboard
20130128 14:46:37 <EricB1> So the keyboard gets away, but the app does not know
20130128 14:46:43 <sgothel> ah .. ok
20130128 14:47:15 <sgothel> then we may also need a specific event to tell the app that it's gone
20130128 14:47:25 <sgothel> i.e. a simple key event might not be sufficient
20130128 14:47:59 <sgothel> since VK_END is very generic - so .. VK_SOFTKEY_HIDDEN maybe ?
20130128 14:48:44 <sgothel> VK_KEYBOARD_INVISIBLE (to match naming - and reference it in the API doc of GLWindow.setKeyboardVisible(..)
20130128 14:48:44 <xranby1> sgothel: IMHO, we should have NEWT keys that match A) all keys defined by the USB keyboard spec
20130128 14:49:16 <xranby1> and then add B) keys only found on virtual keyboards
20130128 14:49:50 <sgothel> dunno (A) but .. yes - so VK_KEYBOARD_INVISIBLE as a special key event if fine ?
20130128 14:49:58 <sgothel> *is
20130128 14:50:22 <sgothel> this would remove the need for just another listener ..
20130128 14:50:30 <EricB1> Jsut a second : checking what Android defines as mandatory buttons for its standard toolbar
20130128 14:50:35 <sgothel> and user app will be notified
20130128 14:51:10 <sgothel> A:BACK -> N:VK_KEYBOARD_INVISIBLE, A:HOME -> N:HOME, ...
20130128 14:51:54 <sgothel> @Xerxes: VK doesn't mean virtual keyboard, but virtual keycode
20130128 14:52:01 <xranby1> in my work, keyboards usually sends out a really screwed up key code according to some odd specification like the USB spec. operating systems needs a keymap to convert the key number to unicode
20130128 14:52:37 <xranby1> usually the keymap depends on key modifiers sequence
20130128 14:52:44 <xranby1> for the os to know what the key mean
20130128 14:52:52 <sgothel> .. we also have another bug .. (from Ola) properly pointing fingers on fact the we don't use layout independent VKs
20130128 14:53:23 <xranby1> olas problem can be solved by using the layout neutral raw keycodes
20130128 14:53:32 <sgothel> 'just' :)
20130128 14:53:56 <xranby1> yes 'just' that it might be hard to know the raw keycodes when we get the os generated events
20130128 14:54:12 <xranby1> since then we get the os specific keycode
20130128 14:54:16 <sgothel> I 'gladly' got distracted by other tasks .. so this is still a painful one to do and _to_ _test_
20130128 14:54:27 <sgothel> yeah .. then map it to NEWT .. etc .. a PIA
20130128 14:54:48 <xranby1> yup yup
20130128 14:54:48 <sgothel> you are all right though .. it should be done
20130128 14:55:14 <xranby1> the usb spec sends the same codes independednt of the OS
20130128 14:55:27 <xranby1> so it should be possible to convet from os mapping to usb mapping
20130128 14:55:34 <sgothel> https://jogamp.org/bugzilla/show_bug.cgi?id=641
20130128 14:56:01 <xranby1> usb got a unique number for all windows and mac keys etc
20130128 14:56:03 <EricB1> Android doc states that "The action bar houses the device navigation controls Back, Home, and Recents"
20130128 14:56:15 <sgothel> X11 events don't know * about USB mapping
20130128 14:56:17 <EricB1> So there are only standard 3 buttons
20130128 14:56:34 <sgothel> cool
20130128 14:56:39 <sgothel> A:RECENT -> N:??
20130128 14:56:52 <EricB1> Note that Amazon's Kindle already adds its own "Search" button
20130128 14:57:24 <sgothel> well - we can't save the world right now and here - but maybe able to make baby steps :)
20130128 14:57:50 <sgothel> so we agree on above: A:BACK -> N:VK_KEYBOARD_INVISIBLE, A:HOME -> N:HOME, ...
20130128 14:58:07 <EricB1> BTW, the Back buttons lets us within the application, but the others switch to another application
20130128 14:58:32 <sgothel> yup, I know - let me phrase is in bugreport .. and then it would be great to have a revisited patch ofc
20130128 14:58:53 <EricB1> And when the softKeboard is NOT visible, the Back button is handled by the application like any other key event
20130128 14:59:20 <xranby1> sgothel: we simply to to save the world from all standard makers who have already tried to save the world by inventing a new standard to save the world by inventing yet another new standard:)
20130128 14:59:35 <xranby1> try to
20130128 14:59:55 <xranby1> ok i will stop distracting
20130128 15:00:00 <EricB1> :)
20130128 15:00:00 <sgothel> hehe
20130128 15:00:23 <sgothel> Note: A:BACK -> N:VK_KEYBOARD_INVISIBLE - is only true IFF the soft keyboard is up !
20130128 15:00:34 <EricB1> I'll re-write my pull request, and re-propose it
20130128 15:00:43 <sgothel> in any other case it would be: A:BACK -> N:ESC -> since it will end the application
20130128 15:00:45 <sgothel> agreed ?
20130128 15:00:53 <EricB1> @sgothel : Yes, only if the softKeyboard is visible
20130128 15:01:21 <sgothel> ok .. commenting on bug 677 .. one sec ..
20130128 15:02:36 <EricB1> But currently, when the softKeyboard is NOT visible, the back button triggers Activity::onBackPressed()
20130128 15:02:53 <EricB1> (and nothing when it's visible)
20130128 15:03:35 <EricB1> So consuming the Back button might change this regular Activity behavior
20130128 15:03:39 <sgothel> so you mean .. the BACK button is being intercepted _only_ w/ this soft keyboard up anyways?
20130128 15:03:46 <EricB1> Correct
20130128 15:03:53 <EricB1> Hence my fix
20130128 15:04:16 <EricB1> (sorry i did not make this clar from the start)
20130128 15:04:24 <EricB1> s/clar/clear
20130128 15:04:45 <sgothel> well .. than thats fine as well, i.e. we ignore the other BACK button semantics not handled by dispatchKeyEventPreIme(..)
20130128 15:05:07 <sgothel> i.e. dispatchKeyEventPreIme() works only w/ soft keyboard up (right?)
20130128 15:05:26 <xranby1> USB spec hid usage table section 5.7 miscellaneous simulation devices - Magic Carpet Simulation Device ... i want one :D http://www.usb.org/developers/devclass_docs/Hut1_12v2.pdf
20130128 15:06:15 <sgothel> no: "a
20130128 15:06:15 <sgothel> * typical example would be handling the BACK key to update the application's
20130128 15:06:15 <sgothel> * UI instead of allowing the IME to see it and close itself."
20130128 15:06:40 <sgothel> so I guess - we don't want to do that .. or do we ?
20130128 15:06:59 <sgothel> this would actually map better to windowClosing abstraction for non keyboard events
20130128 15:07:41 <sgothel> -> NEWT's WindowClosingProtocol
20130128 15:08:18 <sgothel> so: if virtual-keyboard -> VK_KEYBOARD_INVISIBLE, else use WindowClosingProtocol
20130128 15:08:50 <EricB1> Not sure about WindowClosing
20130128 15:08:52 <sgothel> the latter already allows a user to ignore the closing .. etc and receive such event via the WindowListener
20130128 15:09:21 <EricB1> Ah Ok
20130128 15:09:38 <EricB1> *checking code*
20130128 15:09:54 <xranby1> page 53 lists the USB keyboard codes
20130128 15:09:59 <xranby1> Usage ID
20130128 15:10:38 <xranby1> so when you press 'a' usb sends hex 04
20130128 15:10:42 <sgothel> WindowImpl.java line 1569
20130128 15:10:49 <sgothel> line 2569
20130128 15:11:02 <sgothel> // send synced destroy notifications
20130128 15:11:02 <sgothel> enqueueWindowEvent(true, WindowEvent.EVENT_WINDOW_DESTROY_NOTIFY);
20130128 15:11:02 <sgothel> if(handleDestroyNotify && WindowClosingMode.DISPOSE_ON_CLOSE == getDefaultCloseOperation()) {
20130128 15:11:02 <sgothel> destroy();
20130128 15:11:02 <sgothel> }
20130128 15:11:29 <EricB1> @sgothel But we might not want to close the app when the user presses Back
20130128 15:11:43 <EricB1> For instance : we have a menu view, and the a settings view
20130128 15:11:44 <sgothel> yes, hence the WindowProtocol
20130128 15:12:10 <sgothel> i.e. you already can say that closing shall do nothing
20130128 15:12:18 <sgothel> but the notify event comes through
20130128 15:12:23 <EricB1> Ah. I don't know how this works. I'll have to read before I can argue :)
20130128 15:12:25 <sgothel> this matches the Android spec
20130128 15:12:54 <sgothel> I have to add this detail to the WindowListener.windowDestroyNotify(..)
20130128 15:15:22 <EricB1> So I should try to use WindowClosingProtocol rather than the listener mechanism.
20130128 15:15:41 <EricB1> I'm going to re-write my fix.
20130128 15:16:20 <sgothel> http://jogamp.org/git/?p=jogl.git;a=commit;h=2f63a43fd6ff9964251c43e248c51bc821f3ecbd
20130128 15:17:32 <sgothel> I would say: if virt-kb is up -> VK_KEYBOARD_INVISIBLE - if not - call NEWT's WindowImpl.windowDestroyNotify(false) !
20130128 15:17:47 <sgothel> The latter will respect WindowCloseProtocol .. and all is fine :)
20130128 15:20:45 <sgothel> https://jogamp.org/bugzilla/show_bug.cgi?id=677#c2
20130128 15:22:58 <sgothel> https://jogamp.org/bugzilla/show_bug.cgi?id=677#c3
20130128 15:24:05 <sgothel> now it sounds more clear - great, I guess we have assimilated this feature as well in an agnostic way
20130128 15:24:53 <EricB1> Last point : what value for VK_KEYBOARD_INVISIBLE ?
20130128 15:25:23 <sgothel> something new .. actual value of your artistic choice IMHO :)
20130128 15:30:25 <masterzen> let it be 0xDEADBEEF then :)
20130128 15:31:18 <xranby1> the beginning of a class file.. why not
20130128 15:31:31 <xranby1> err
20130128 15:31:31 <sgothel> a good choice - also b/c it is beyond the 16bit (short) value space
20130128 15:31:35 <xranby1> i mean CAFEBABE
20130128 15:31:52 <sgothel> and semantics are matching :)
20130128 15:57:22 * sgothel (~sven@anon) Quit (Ping timeout: 276 seconds)
20130128 16:12:55 * sgothel (~sven@anon) has joined #jogamp
20130128 16:12:55 * ChanServ sets mode +v sgothel
20130128 17:06:45 * EricB1 (~Adium@anon) Quit (Quit: Leaving.)
20130128 21:13:40 * EricB1 (~Adium@anon) has joined #jogamp
20130128 22:47:24 * EricB1 (~Adium@anon) Quit (Quit: Leaving.)
20130129 04:31:41 <Titanoum> xranby1
20130129 04:31:47 <Titanoum> no jpanels just are terrible
20130129 05:05:28 -CatOut- Continue @ http://jogamp.org/log/irc/jogamp_20130129050528.html