18:08:43 <darci> #startmeeting
18:08:44 <tota11y> Meeting started Fri Mar  1 18:08:43 2013 CET.  The chair is darci. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:08:44 <tota11y> Useful Commands: #action #agreed #help #info #idea #link #topic.
18:08:58 <darci> Ok...updates anyone?
18:09:07 <darci> #topic Updates
18:09:10 <amber> ill go
18:09:29 <amber> so i found a uml of the opencv framework only that was in the mousetrap docs.
18:09:41 <amber> epydoc is not able to complete the diagram because of errors
18:09:58 <Stoney> install graphviz
18:10:00 <amber> so i am trying it in a vm using an old version of ubuntu but am still seeing an error
18:10:16 <amber> Stoney, it was because mousetrap couldn't compile
18:10:26 <Stoney> oh i see
18:10:32 <amber> so it couldnt traverse the dependencies
18:10:43 <amber> but that might be why im still seeing the error in ubuntu
18:10:50 <amber> ill try it, thanks!
18:10:52 <john_> It worked in Ubuntu 10.4, did you use that version?
18:11:00 <amber> yes im in it now
18:11:02 <amber> :)
18:11:10 <darci> #info amber found a uml of the opencv framework only that was in the mousetrap docs
18:11:21 <heidi> Nice work Amber!
18:11:22 <amber> i uploaded it to the current version of my git branch
18:11:34 <amber> in the mousetrap docs folder
18:11:49 <amber> along with an html version of the epydocs that did compile
18:12:19 <amber> so i think thats it for me
18:12:25 <darci> #info epydoc couldnt traverse the dependencies because mousetrap doesn't compile
18:12:56 <amber> it did do the html written docs, just not the uml diagram
18:12:58 <darci> #action amber will try graphviz
18:13:28 <Stoney> graphviz is needed to generate diagrams...
18:13:38 <Stoney> but the dependency issue is still a problem
18:13:47 <amber> i am geting a gdk_x_error()
18:13:58 <darci> #info amber did work on the html written docs
18:13:58 <amber> for a "bad window"
18:14:21 <Stoney> hmm...
18:15:06 <amber> i wonder if it has something to do with my camera not being enabled? bc im running a vm..
18:15:09 <heidi> Sorry, Firefox crased
18:15:34 <heidi> My apologies if you have already covered this, but do you know that your camera works in the VM?
18:15:49 <heidi> (outside of mousetrap that is?)
18:16:10 <amber> no, it doesnt
18:16:23 * john_ kinda wishes he hadn't returned the laptop
18:16:24 <john_> haha
18:16:30 <amber> which is way i suspect it could be the probelm
18:16:47 <heidi> Right, I have had much problems getting cameras to run in VMs.
18:17:00 <amber> i dont know how to even attempt it haha
18:17:01 <heidi> The problems tend to be driver-related.
18:17:08 <darci> John: you can always get the laptop back
18:18:01 <darci> So amber: you are doing all of this just to generate the UML diagrams, right?
18:18:09 <amber> yea
18:18:13 <amber> it didnt take long tho
18:18:30 <amber> but to go further, i dont think its worth it
18:18:40 <heidi> Right, I agree
18:18:51 <amber> but i can still do the manual docs i was working on
18:19:00 <amber> which will be worth it
18:19:13 <john_> I could toss Ubuntu on my external and generate the UML file if you like.
18:19:26 <john_> try to, anyway
18:19:44 <Stoney> very strange to have to spin up another OS just to generate a UML :)
18:20:07 <john_> because mousetrap (the original branch) compiles+ runs fine on Ubuntu 10.4
18:20:22 <john_> but wouldn't compile on Fedora (or any newer OSes I suspect)
18:20:29 <Stoney> ah
18:20:32 <amber> i could do it, but i don tknow if we want to spend more time on it
18:20:42 <darci> John: another thought...
18:20:56 <darci> You could stop the lab and try it quick.
18:21:13 <john_> Sadly the laptop now has Fedora 17 on it, not Ubuntu 10.4
18:21:18 <john_> so it wouldn't be quick :)
18:21:34 <darci> Oh yea...duh!
18:22:19 <amber> and you cant boot off a flash drive either, bc you have to be root in order to use apt-get right?
18:23:47 <darci> Ok...maybe we can think about this some more.
18:23:54 <darci> John: do you have any updates?
18:26:57 <john_> Not much -- I looked at the ocfw source and read Amber's docs.
18:27:07 <john_> as Amber asked me to do last week
18:27:37 <amber> did it make more sense?
18:27:47 <john_> Amber -- looking at my status re: Mousetrap install on Ubuntu, it ran great when there's no camera plugged in. Couldn't the UML be generated when no camera is present?
18:27:58 <john_> Kinda. Not really. :D
18:28:03 <darci> #info john has looked at the ocfw source and read Amber's docs
18:28:50 <amber> i thought it would work. ill look into the error. seems to be gdk related or maybe a problem with epydocs running on this os.
18:28:57 <heidi> As for the WNE students, Logan is still working on debugging OpenCV stuff.
18:29:38 <amber> mousetrap runs just fine tho, cool to finally see it work :)
18:30:36 <darci> Very cool!
18:30:37 <heidi> Stoney, do you remember what Nikki is doing?
18:30:41 <heidi> Yes, very cool!
18:31:32 <john_> you all saying I was the first to see it working way back in December? lol
18:31:39 <Stoney> (sorry person at door... i'm back... reading)
18:31:49 <darci> #info Logan is working on debugging OpenCV
18:32:30 <Stoney> I'm not sure what nikki is working on specifically, no
18:33:11 <amber> i know she was working on looking through the opencv meeting notes
18:33:19 <heidi> I think we set her to figure out if she could find any more recent design docs for MouseTrap.
18:33:22 <amber> and i suggested to her that we try to track down the ones for mousetrap instead
18:33:37 <amber> jinx :)
18:33:44 <heidi> :-)
18:34:28 <darci> FYI...I just found this on stackoverflow:  Epydoc is a tool to generate API documentation from Python source code. It also generates UML class diagrams, using Graphviz in fancy ways.
18:34:43 <Stoney> yup
18:34:45 <darci> So, Epydoc uses Graphviz?
18:34:48 <amber> yes
18:34:49 <Stoney> yup
18:35:02 <amber> but thats the part i cant get to work
18:35:14 <darci> :-)
18:36:48 <Stoney> amber, silly question, but did you use yum or apt to install graphviz?
18:36:49 <darci> #info Nikki is working on looking through the opencv meeting notes and tracking down design docs for MouseTrap
18:37:01 <amber> apt is on ubuntu
18:37:05 <amber> so apt
18:37:11 <amber> yum is fedora
18:37:23 <amber> so i tried both
18:37:36 <amber> it worked on fedora but Mousetrap didnt
18:37:36 <Stoney> hm
18:37:40 <Stoney> right
18:37:41 <amber> and now i have the opposite
18:37:50 <Stoney> ah, ok
18:38:39 <darci> Any other thoughts on sending an email out to the MouseTrap list serve?
18:39:12 <darci> Actually, John may not have heard about this before?
18:39:56 <heidi> We should first check with joanie to see if that list has anyone else except us on it.
18:40:10 <joanie> how far should I scroll back?
18:40:10 <joanie> :)
18:40:17 <john_> apt-get graphviz should work? https://launchpad.net/ubuntu/lucid/+package/graphviz
18:40:31 <amber> yea i tried it
18:40:33 <amber> just now
18:40:38 <amber> same gdk error tho
18:40:42 <heidi> Hi Joanie! Just wanting to know if the MouseTrap list was resurrected for us or if the previous developers might be on it.
18:40:50 <john_> I'm on the list serv but never get anything
18:41:02 <joanie> previous developer(s) might be on it
18:41:15 <heidi> Right, but the question is who would we reach if we send email.
18:41:17 <heidi> OK, thanks.
18:41:23 <heidi> So we should ping the list.
18:41:27 <amber> mousetrap says to report bugs to: flaper87@flaper87.org
18:41:28 <joanie> no
18:41:42 <joanie> amber: that should be changed as well
18:41:49 <joanie> Flavio may or may not respond
18:41:59 <joanie> he apparently did not respond to heidi's gnome account access
18:42:05 <john_> flapper was ~21 when he wrote this. Probably moved to bigger/better things.
18:42:05 <amber> ah ok
18:42:12 <heidi> OK, thanks. The core issue is that we have some questions about the design of MouseTrap
18:42:16 <joanie> and he definitely did not respond to my list access
18:42:24 <heidi> Ah, OK.
18:42:29 <heidi> So might not be there any longer.
18:42:35 <joanie> heidi: if it's design related and you have a BetterIdea(tm), go for it
18:42:41 <joanie> it's your/our module now ;)
18:42:42 <heidi> Ah, OK.
18:43:11 <joanie> I'll get the list resurrected (i.e. API, me, and heidi as admins)
18:43:22 <joanie> if Flavio feels like chiming in, we certainly would not mind
18:43:26 <joanie> ;)
18:43:39 <joanie> anything else I can do for y'all in the meantime?
18:43:50 <heidi> The design appears to have a class that might serve as an interface to the OpenCV stuff
18:44:16 <heidi> and we're just curious as to the purpose. But we can go ahead around that as needed if it's OK to do a significant rewrite.
18:44:33 <heidi> Joanie, the previous is the core of the issue.
18:44:47 <joanie> if a significant rewrite is called for, what I would recommend is:
18:45:00 <joanie> do it all in a branch (it could be an upstream gnome branch if you wanted)
18:45:05 <joanie> finish it
18:45:09 <joanie> test the heck out of it
18:45:20 <joanie> and if it is everything you dreamed it would be, merge it into master
18:45:32 <joanie> the older versions/releases will still be there
18:45:35 <heidi> Yes, OK.
18:45:38 <heidi> :-)
18:46:21 <heidi> Anyone have any other questions for joanie?
18:46:24 <joanie> on a related note heidi, we should probably see if there are many (any) commits since the last release
18:46:29 <john_> joanie: newbie question about GIT: if you delete an entire directory (say, theoretically, the ocfw directory) and merge it, how does that work? Do the directory get removed in master?
18:46:30 <heidi> Ah, good idea.
18:46:52 <joanie> if you git rm it yes
18:46:55 <Stoney> only in the new version john_
18:46:57 <joanie> if you just rm it no
18:47:09 <john_> thanks
18:47:15 <Stoney> the old versions still have the deleted stuff.
18:47:20 <Stoney> so you can always get them back
18:48:22 <joanie> heidi: to finish my last thought:
18:48:43 <joanie> if there are a lot of uncommitted releases, we should branch
18:48:57 <joanie> i.e. make a branch for version foo
18:49:01 <joanie> so that master is the new stuff
18:49:44 <joanie> anyhoo, didn't mean to take over your meeting (sorry!!!!). I'll go back to my corner until I'm needed. ;)
18:49:48 <heidi> Ah, OK.
18:49:55 <heidi> You're a huge help Joanie!
18:50:00 <heidi> We appreciate all of your input!
18:50:21 <Stoney> yes yes huge help!
18:51:20 <darci> Should we move on to next steps?
18:51:28 <heidi> #action Heidi look into the status of releases in the next week.
18:51:35 <heidi> Sure,
18:51:42 <darci> #topic Next Steps
18:53:08 <darci> What is planned for the upcoming week?
18:53:48 <amber> im going to continue to work on docs..?
18:53:54 <darci> Also, are we planning to meet next Friday? Some of us will be in Denver?
18:54:19 <heidi> Ah, I think that those who can, should meet. But yes, most of the profs may not be able to make it.
18:54:25 <amber> I think ghislop wont be able to make it
18:54:41 <amber> we can do a student meeting, if enough of us can meet
18:55:30 <darci> Heidi: do you know if Logan could make the meeting on Fridays if the time changed?
18:55:44 <heidi> I don't know.
18:56:03 <heidi> I don't know the hours he works for OIT.
18:56:10 <heidi> Maybe a doodle poll?
18:57:36 <amber> that sounds good
18:57:52 <darci> I don't mind putting one together...
18:58:32 <heidi> OK, thanks.
18:58:41 <darci> Should I look for another time on Friday?
18:58:51 <darci> Or open it up to other days?
18:59:05 <amber> both?
18:59:11 <heidi> Yes, I would suggest both.
18:59:15 <heidi> OK, I need to run. Sorry!
18:59:17 <heidi> Bye
18:59:22 <darci> bye
18:59:56 <darci> Ok...it is almost 1:00
18:59:59 <darci> Any last thoughts?
19:00:16 <amber> when should we start reworking the framework?
19:00:48 <darci> It sounds like that is up to us.
19:00:53 <darci> you?
19:01:18 <amber> i think if we have at least two of us ready to code, we could start soon
19:01:28 <Stoney> then go go go :)
19:01:34 <darci> I agree!
19:01:52 <amber> i would like someone I can bounce ideas off of?
19:01:57 <amber> on how to rewrite
19:02:35 <john_> darci -- do you know of any simple/easy Python projects I could do to learn Python?
19:02:56 <darci> I sure I can find some.
19:03:21 <darci> amber:
19:03:29 <darci> oops I didn't finish that
19:03:46 <darci> Why not throw ideas out to the group?
19:04:03 <Stoney> excellent idea
19:04:17 <amber> ok, ill try to put something together or mess around in a branch everyone can look at
19:04:45 <darci> I'm thinking via email, so that those you can (either because they have an idea and/or the time) will reply
19:05:02 <darci> "you" who
19:05:30 <amber> ok
19:05:47 <Stoney> yup, email is better, as it doesn't require checking in on a website
19:06:06 <darci> #action Darci will find John some easy Python projects to work on.
19:06:28 <Stoney> I'll do my best to respond
19:06:35 <darci> Me too
19:06:40 <amber> sounds good
19:06:50 <john_> From scratch (think of it as an assignment :)) is fine, too, to start off
19:07:14 <darci> I can do that...
19:07:27 <amber> well i doubt any of us know how to use haars effectively enough to rewrite that part..
19:07:41 <darci> Ok...I need to head out. Anything else?
19:07:42 <john_> doesn't OpenCV have haars stuff built-in?
19:07:49 <john_> Could we just call on it?
19:08:00 <amber> yes, but we have mousetrap specific ones too
19:08:10 <amber> that arent in opencv's goodfeaturestotrack
19:08:14 <Stoney> haars?
19:08:43 <amber> they are xml files with calculations used to create a template for tracking features
19:09:00 <Stoney> thx
19:09:05 <Stoney> I think I just found a page on it
19:09:41 <Stoney> #link http://opencv.willowgarage.com/wiki/FaceDetection
19:10:13 <Stoney> adjourn?
19:10:23 <darci> Yes...
19:10:27 <darci> #endmeeting