Camelia, the Perl 6 bug

IRC log for #openframeworks, 2013-08-19

| Channels | #openframeworks index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
01:18 seb_ly joined #openframeworks
02:48 jvcleave joined #openframeworks
02:49 seb_ly joined #openframeworks
07:11 andreas_ joined #openframeworks
07:26 rossc1 joined #openframeworks
07:42 rossc1 joined #openframeworks
08:47 rossc1 joined #openframeworks
09:54 okjake joined #openframeworks
10:54 andreas_ joined #openframeworks
11:08 seb_ly joined #openframeworks
11:34 neue joined #openframeworks
11:51 neue left #openframeworks
11:53 neue joined #openframeworks
12:17 andreas__ joined #openframeworks
12:38 superFluid joined #openframeworks
13:05 harpreet joined #openframeworks
13:10 harpreet left #openframeworks
13:10 harpreet joined #openframeworks
13:21 stephanschulz joined #openframeworks
13:41 harpreet joined #openframeworks
13:53 seb_ly joined #openframeworks
13:57 dantheman joined #openframeworks
14:22 workergnome joined #openframeworks
14:25 dantheman left #openframeworks
14:33 laserpilot joined #openframeworks
14:37 admsyn joined #openframeworks
14:43 admsyn stephanschulz: hey
14:43 stephanschulz hi
14:43 admsyn RE: aggregate devices and such, did you try making a multi-output device too?
14:43 admsyn I think that might be the trick
14:44 stephanschulz not sure about the details of that
14:44 stephanschulz i aggregated many fast tracks
14:44 stephanschulz and it shows in sys pref as a device
14:44 admsyn you can create two kinds of devices, an aggregate device or a multi-output device
14:45 stephanschulz i see
14:45 admsyn I *think* multi-output will solve it
14:45 stephanschulz let me try
14:45 admsyn you should see the two options after clicking the +
14:47 stephanschulz ok made one
14:47 stephanschulz but same problem
14:47 stephanschulz even with your busses example
14:47 stephanschulz also getOutChannelAmount() report 2 outputs
14:47 admsyn hm :/ I wonder if aggregate devices need some extra setup from the host
14:47 admsyn I thought they were supposed to just kinda figure it out
14:48 stephanschulz it also did not work with the matrixmixer test that i mailed you a while back
14:48 stephanschulz that only uses objective-c
14:48 admsyn yeah I'm guessing that either the aggregate device needs a box checked or something from the user's side, or some extra function call needs to happen to set it up properly
14:48 admsyn try your aggregate device in live or garageband or something
14:51 dantheman joined #openframeworks
14:53 admsyn another thing to try would be fiddling with the "clock source" in Audio MIDI Setup
14:53 admsyn since ultimately it seems like the problem is nothing is claiming ownership of the audio timing callbacks
14:54 admsyn maybe you need to add the computer's built in output to the aggregate device, and set it as the clock source
14:54 rossc1 joined #openframeworks
14:58 stephanschulz i tried that
14:58 stephanschulz adding the build in otuput and use it's clock
14:59 stephanschulz i tried max msp
14:59 stephanschulz and when i select the coreaudio aggregated device it's not playing the sound file
15:00 admsyn alright
15:00 admsyn well at least that narrows it down to the aggregate device..
15:01 admsyn IMO it'd be way easier to figure out the aggregate device path, since trying to manage all of the multichannel sound with 11 discrete output units is going to be WAAAAYYY harder (i think)
15:01 admsyn it's probably just a box, or a preference or something like that that's tucked away somewhere
15:03 stephanschulz ok max seems to work
15:03 stephanschulz i reduces the amount of devices to make the tests easier
15:03 stephanschulz i have 1 and 2 fast tracks
15:03 stephanschulz and it is playing through at least one speaker
15:04 admsyn oh, and then when you add a few more it drops out?
15:04 stephanschulz ok this works also with the matrix mixer
15:04 stephanschulz i will add a more now
15:09 stephanschulz mh
15:09 stephanschulz adding more then 2 device per usb port seems to stop things from working
15:09 admsyn interesting
15:09 stephanschulz i now have 2 on port a and 2 on port b
15:09 stephanschulz and that works
15:10 stephanschulz but when i add onemore it doe not
15:10 admsyn no matter which one? As in, it's not just a misbehaving device?
15:12 stephanschulz i tried a bunch
15:12 admsyn alright, so maybe it's a USB bandwidth thing? That'd suck
15:12 stephanschulz i will try an other computer
15:12 stephanschulz ok at least i know it's nothing that could be solved with software
15:12 admsyn yeah if you have a fireware sound card kicking around I'd give that a shot
15:13 admsyn *firewire
15:13 admsyn could also try dropping the sampling rate just to diagnose it
15:13 admsyn i.e. if you drop it to like 8000hz can you get a few more on there?
15:14 stephanschulz of each of the fast tracks in the midi setup
15:14 stephanschulz ?
15:14 admsyn yeah
15:14 admsyn but it definitely sounds like USB bandwidth issues
15:15 dantheman joined #openframeworks
15:15 stephanschulz yes that works
15:15 stephanschulz now i am at 2 x 3 devices
15:15 admsyn alright that's pretty conclusive then
15:15 admsyn I'd imagine 8000hz audio ain't gonna cut it
15:15 admsyn is a firewire sound card an option?
15:16 stephanschulz the min i can set it to is 32k
15:16 stephanschulz in the midi setup
15:16 stephanschulz i will see if with a mac mini it will work
15:16 stephanschulz because it has more usb ports
15:16 stephanschulz but then again they might be on the same buss
15:16 admsyn vOv
15:18 admsyn well, depending on how tolerant your installation is to low(ish) quality sound, I think you can set it to lower than 32k via Core Audio function calls, even if there's no interface for it in Audio MIDI Setup
15:18 stephanschulz can you tell me how
15:19 admsyn it'll be the AudioHardware* set of functions I think
15:19 admsyn something like AudioHardwareSetProperty()
15:19 stephanschulz AudioStreamBasicDescription ASBD = {
15:19 stephanschulz .mSampleRate       = 44100,
15:19 admsyn oh yeah that might do it too
15:19 admsyn if you set the ASBD of the output unit to use a lower sample rate it might figure it out
15:20 admsyn I think in Ableton Live or Logic or GarageBand or Max/MSP you might find a way to test it too
15:21 stephanschulz .mSampleRate       = 32000,
15:21 stephanschulz makes the files play slower
15:21 admsyn sounds about right :)
15:21 stephanschulz but even when setting it to 8000
15:21 stephanschulz and adding more devices
15:22 admsyn you'd probably need to set the same ASBD on the file players to make it work out
15:22 stephanschulz the fileplayers do not start
15:22 admsyn yeah so even super-low sampling rate might not fix it
15:22 stephanschulz about the AudioHardwareSetProperty
15:23 stephanschulz where do i find this?
15:23 admsyn it'll be in the core audio framework
15:24 admsyn I think if you expand it in your "Frameworks" group in Xcode, then look for somehting like AudioHardware.h you'll find the "documentation"
15:24 admsyn if the ASBD actually works though, that's definitely going to be the easier way to do it
15:24 admsyn the AudioHardware stuff is even lower level than audio units, it's not fun
15:25 stephanschulz i will first try a mac mini
15:25 stephanschulz and see what happens
15:25 stephanschulz on an other front:
15:26 stephanschulz you don't think it is easy to get the matrix mixer output connected to seperate devices
15:26 stephanschulz not using aggregated devices
15:27 admsyn no, because the entire audio unit system is based on one output unit driving the chain
15:27 admsyn so if you have 11 different units, you'd need 11 entirely isolated chains
15:27 admsyn you wouldn't be able to have a matrix mixer sitting in the middle
15:27 admsyn (because it would be told to render 11 times per callback, when it'd only be expecting 1)
15:28 admsyn there might be a clever way to just have 11 isolated stereo audio unit chains that does what you want though
15:28 admsyn you'd still run into the USB bandwidth issue though, I think
15:28 stephanschulz strange because i had a setup that had 10 or so fast tracks on two usb hubs
15:28 stephanschulz all playing different stuff
15:28 admsyn alright, so maybe that avoids the USB issue then
15:29 stephanschulz before i started with martix mixers
15:29 admsyn is there a way to do what you want with 11 isolated chains? Do you need to share (for example) one sound file between all 11 chains simultaneiously?
15:29 stephanschulz no
15:29 stephanschulz but it would have been so nice to use the matrix mixer code
15:29 admsyn it would, yeah
15:30 stephanschulz and not to split everything up 11 times
15:30 stephanschulz the mic will need to be attached to all output devices
15:30 stephanschulz but i guess that's just a tap that gets connected to each mixer
15:30 admsyn ah right
15:31 admsyn you *might* be able to do it with juts 11 ofxAudioUnitInputs
15:31 admsyn at some point though, you're going to make Core Audio very angry :p
15:31 admsyn but that might work
15:31 stephanschulz ok well thanks so far
15:32 stephanschulz i will try a mac mini
15:32 stephanschulz then i will mess around with sample rates
15:32 stephanschulz see if i can get the AudioHardwareSetProperty stuff to do  something
15:32 stephanschulz otherwise i will make 11 seperate stereo mixers
15:32 stephanschulz 11 seperate outputs etc
15:48 jacres joined #openframeworks
16:01 harpreet joined #openframeworks
16:07 workergnome The documentation for ofVideoGrabber.listDevices() says, "Prints to the console a list of available capture devices with the device ID of each device. The device ID can then be used with setDeviceID() to specify a specific device to capture from."
16:08 admsyn workergnome: are you running into a bug with it?
16:08 workergnome It doesn't actually print to the console, as far as I can tell—it just returns a vector of the devices.  Are "the console" and the xCode output window different things?
16:09 admsyn no they should be the same
16:09 admsyn can post code or a screen shot in xcode?
16:11 workergnome Sure—how do people typically post screenshots?
16:11 admsyn imgur is what I usually do
16:11 admsyn cmd-4 + space will get you a screen of just xcode
16:11 workergnome https://dl.dropboxusercontent.com/u/3034651/Scre​en%20Shot%202013-08-19%20at%2012.09.55%20PM.png
16:12 workergnome Does that work for people who aren't me?
16:12 admsyn No I think you're doing it right :)
16:12 admsyn I'm not terribly familiar with oF video, though
16:12 admsyn maybe move the init grabber to before the list devices call? I don't think that'll be the trick though
16:13 admsyn and setVerbose too
16:14 workergnome The thing is, the code works.  I'm seeing the devices I have plugged in in the debugger, and the program works.  I'm just wondering if it's a documentation error.
16:14 admsyn ah maybe
16:14 workergnome (Changing the code order has no effect)
16:14 admsyn yeah, cause your devices vector is getting populated just fine
16:15 workergnome OK.  Time to read OF source code again.
16:17 admsyn ah looks like the list devices is happening at the verbose level
16:17 admsyn yeah the quicktime grabber does it at notice level, and the QTKit grabber does it at verbose
16:18 admsyn that's no good
16:19 admsyn you're on 10.7+, right?
16:21 workergnome 10.8
16:23 dantheman joined #openframeworks
16:29 admsyn alright I opened an issue about that, you should be able to get console logs by doing ofSetLogLevel(OF_LOG_VERBOSE)
16:29 workergnome Yup.  That does it.   S
16:29 workergnome Thanks
16:51 stephanschulz left #openframeworks
17:21 laserpilot joined #openframeworks
17:39 dantheman joined #openframeworks
17:45 rossc1 joined #openframeworks
17:58 dantheman joined #openframeworks
17:59 superFluid joined #openframeworks
18:31 rossc1 joined #openframeworks
19:08 seb_ly left #openframeworks
19:12 dantheman joined #openframeworks
19:39 seb_ly joined #openframeworks
19:51 workergnome joined #openframeworks
20:52 okjake joined #openframeworks
21:05 dantheman anyone around
21:06 admsyn o hai
21:09 dantheman_ joined #openframeworks
21:10 dantheman_ admsyn: sup
21:10 dantheman_ this is weird
21:10 admsyn what is weird
21:10 dantheman_ https://gist.github.com/danthemellowman/6274203
21:10 dantheman_ that is weird
21:11 admsyn use macports?
21:11 admsyn *are you using macports
21:12 admsyn "error: ‘FILE’ was not declared in this scope" is definitely weird though
21:14 admsyn wait, cstdio is throwing FILE was not declared?
21:14 admsyn isn't…isn't that where FILE is declared?
21:15 dantheman lol
21:15 dantheman yup
21:15 dantheman its on the PI
21:15 dantheman doing a firmware update to see if that fixes it
21:16 admsyn mind_blown.gif
21:16 dantheman yah
21:16 dantheman cause it builds on my Mac
21:16 dantheman hmm I wonder if ofThreads is supported
21:18 admsyn it just wraps poco threads I think? So yeah?
21:18 * admsyn does not know
21:18 dantheman yer it just extends Poco's threads I thought
21:19 dantheman or is that ofEvents?
21:19 dantheman IDK where is arturo when you need him
21:19 admsyn no ofEvents wraps Poco events
21:19 dantheman lol
21:19 dantheman haha
21:19 dantheman right
21:19 admsyn ofThread wraps Poco threads
21:19 admsyn natch
21:19 dantheman natch gotcha
21:19 dantheman admsyn: did you look at the http requests stuff I push yesterday?
21:19 admsyn I did, looks good :)
21:19 dantheman did you test it?
21:20 admsyn I don't do much with http though, so I don't really know how to judge it
21:20 admsyn I did not
21:20 dantheman ahh
21:20 dantheman gotcha
21:20 dantheman I'm a data freak
21:20 dantheman Mr. Data
21:20 admsyn how would I test it?
21:20 dantheman or so I've been called
21:20 admsyn you pushed a dev app IIRC?
21:20 dantheman yah
21:21 dantheman so like put in a call to some webby service
21:21 dantheman using a POST or PUT
21:21 admsyn it's in your branch only right now right?
21:21 dantheman yah only on my branch
21:22 dantheman (or you have the power to merge it don't you now)
21:22 dantheman haha
21:22 dantheman MERGE IT
21:22 dantheman (don't)
21:22 admsyn :p
21:22 admsyn I'm always worried now whenever I do a git push that I'm not 100% sure what I'm doing
21:22 admsyn like I'm going to explode oF entirely
21:23 dantheman lol
21:23 dantheman noob
21:23 admsyn iknorite
21:24 admsyn so what've you tested this on so far?
21:24 admsyn the HHTP
21:24 admsyn HTTP
21:24 dantheman HTTPBin
21:24 dantheman its a site for testing HTTP clients
21:24 admsyn oh I meant win/osx/linux
21:24 dantheman http://httpbin.org/
21:24 dantheman oh just OSX
21:24 dantheman trying it on linux now
21:24 dantheman on the Pi
21:24 dantheman I need to test on windows
21:25 dantheman which I can this week
21:25 admsyn httpbin looks handy
21:25 admsyn yeah I can only test osx right now
21:25 dantheman yer it is very very handy
21:25 dantheman wordy
21:25 admsyn I killed my arch install :(
21:25 admsyn pacman -Syohshit
21:25 dantheman lol
21:26 dantheman ugs I need a faster machine for work
21:26 dantheman my air ain't cracking it
21:27 dantheman I neeeeeeeeed my distcc image but I don't have enough space on my AIR for it
21:27 dantheman :(
21:28 admsyn first thought: it'd be cool to have some constants for user agents
21:28 admsyn like .setUserAgent(OF_USER_AGENT_WHATEVER)
21:28 okjake joined #openframeworks
21:28 admsyn but that's kinda bikeshedding
21:28 admsyn looks good
21:32 admsyn works for me http://i.imgur.com/eApNxqe.png
21:34 dantheman success
21:34 dantheman yeah that is what i was thinking
21:35 dantheman admsyn:
21:35 dantheman yah
21:35 dantheman yeah having const strings for OF_USER_AGENT_ANDROID
21:35 dantheman and like IOS
21:35 dantheman would be interesting
21:36 admsyn yeah that'd be good
21:36 admsyn maybe even a default user agent that'll work 99% of the time
21:36 admsyn but that might be pushing it
21:44 dantheman yah
22:02 dantheman admsyn: it was ofxJSON
22:02 dantheman :(
22:02 dantheman you left me
22:23 workergnome joined #openframeworks
22:43 dantheman jvcleave: I'm getting seg faults allllll day on the pi now
23:10 simensen joined #openframeworks

| Channels | #openframeworks index | Today | | Search | Google Search | Plain-Text | summary