[01:33] <brousch> mrgoodcat: I'm curious to see where you're taking this tactic
[01:43] <mrgoodcat> it's a no-go
[01:43] <mrgoodcat> you had 2 choices and you chose correct
[01:43] <mrgoodcat> oh well, mistakes are just mistakes
[01:46] <mrgoodcat> brousch: i was hoping to invade since you don't have 2 eyes
[01:46] <mrgoodcat> but its just not a good idea
[01:47] <brousch> heh
[01:47] <brousch> I saw 2 choices but they ended up mostly the same
[01:48] <mrgoodcat> it would have required a pretty bad slip upp
[01:59] <mrgoodcat> brousch: i keep missing those simple vulnerabilities
[02:09] <brousch> a lot of foam makes for a lot of places to look
[02:14] <mrgoodcat> apparently
[04:31] <mrgoodcat> brousch: great game
[04:32] <brousch> Yeah, ended up close
[04:33] <mrgoodcat> you just took a huge chunk of the bottom right
[04:34] <mrgoodcat> and the middle wasn't decisive enough. I wish i had won the bottom left or been able to hold you in the top right. that would have been the differnece
[04:35] <brousch> I was quite happy with my top right
[04:36] <mrgoodcat> i thought i could hold it
[04:38] <brousch> Ok, bedtime
[04:39] <mrgoodcat> alright. night
[11:50] <brousch> My fame spreads http://www.blog.pythonlibrary.org/2014/10/13/pydev-of-the-week-ben-rousch/
[11:50] <bookiebot> http://goo.gl/FG5xPl - PyDev of the Week: Ben Rousch « The Mouse Vs. The Python
[12:09] <PhillyMelt02> Excellent article.  And a fellow UM alum.  GO BLUE!
[12:16] <brousch> PhillyMelt02: Are you near Detroit?
[12:17] <PhillyMelt02> Yeah, just north of Mount Clemens in Chesterfield Township
[12:22] <brousch> Going to MUG tonight?
[12:23] <rick_h_> wheee party
[12:28] <PhillyMelt02> Yep, going to MUG.  It will be my first meeting
[12:32] <brousch> Excellent. Another ubuntu-mi person to meet
[12:33] <PhillyMelt02> I guess this is a silly question to you veterans, but is there anything I need to bring?
[12:37] <mrgoodcat> today will be my first MUG too :)
[12:40] <brousch> I've never been to MUG, but I'm told they have wifi
[12:43] <mrgoodcat> i think every public library has wifi
[12:43] <mrgoodcat> how late does MUG usually go?
[12:44] <PhillyMelt02> the site lists the meeting going until 8:45
[12:50] <mrgoodcat> sounds good to me
[13:54] <cmaloney> Good morning
[13:55] <rick_h_> party
[13:55] <rick_h_> save me, I'm in marketing/sales hell
[13:55] <cmaloney> rick_h_: Welcome to the other side. :)
[13:56]  * rick_h_ dives on the floor looking for a dropped pistol
[13:59] <cmaloney> heh
[14:34] <cmaloney> brousch: Congrats on being the dev of the week. :)
[14:40] <brousch> Now if i could get the rpi to actually play a video it makes ...
[14:40] <cmaloney> heh
[14:40] <brousch> Damn 700MHz CPU
[14:40] <cmaloney> yeah, it's unfortunately a bit underpowered as a "desktop" computer.
[14:43] <brousch> web browsing is painful
[14:50] <mrgoodcat> put arch and i3 on mine
[14:50] <mrgoodcat> runs just fine
[15:04] <jrwren> you are using the wrong video player.
[15:04] <mrgoodcat> oh god damnit
[15:04] <mrgoodcat> brousch: me--
[15:04] <jrwren> I playback 1080p h264 with rpi.
[15:05] <mrgoodcat> jrwren: same here
[15:05] <mrgoodcat> no problems
[15:05] <jrwren> hardware assist h264 rules.
[15:05] <jrwren> now we just need hardware assist browser rendering.
[15:05] <mrgoodcat> avi drooles?
[15:06] <jrwren> avi is just a container. can put h264 in there.
[15:06] <mrgoodcat> idk anything about video codecs/containers/formats
[15:06] <jrwren> mrgoodcat: now you do. :p
[15:07] <mrgoodcat> i just know I play 1080p h264 since thats the settings i use for transcoding
[15:07] <mrgoodcat> i actually use 720 most of the time though
[15:23] <mrgoodcat> python requests vulnerability http://www.ubuntu.com/usn/usn-2382-1/
[15:23] <bookiebot> http://goo.gl/agZIAT - USN-2382-1: Requests vulnerabilities | Ubuntu
[15:34] <mrgoodcat> also an oxide vuln. a slightly bigger deal since chromium (and chrome?) uses it http://www.ubuntu.com/usn/usn-2345-1/
[15:34] <bookiebot> http://goo.gl/vkqs60 - USN-2345-1: Oxide vulnerabilities | Ubuntu
[15:37] <mrgoodcat> whelp, upon further reading i was wrong. chromium does not use oxide but oxide uses chromium. Oxide is a library that allows you to embed a Chromium-powered webview in QML applications. still should probably update though :)
[16:09] <brousch> jrwren: The codec and container selection is limited. I can generate avi files pretty well, but they're not h264. The built-in efficient RPi player will not play them. However VLC will play them if they are small enough not to use 100% of the CPU (240x180)
[16:09] <jrwren> brousch: i'm not really sure what you just said :)
[16:09] <brousch> I have not succeeded in generating h264 videos from the webcam, but I could from the official RPi cam
[16:10] <jrwren> brousch: ah. I see. You just need a better encoder :)
[16:11] <brousch> Yes, and on-the-fly h264 encorder for rpi
[16:12] <brousch> The videos from the official cam were quite nice, but something broke in the latest firmware or in my camera and it no longer works at all
[16:12] <brousch> The good news is the solution for a generic webcam are more useful generally, and on laptops and such
[16:16] <jrwren> does the pri cam encode to h264 by default or something?
[16:16] <jrwren> I guess I shouldn't be surprised. That is how cell phones do it.
[16:57] <mrgoodcat> jrwren: the raspicam supports raw and accellerated h.264 natively
[16:57] <mrgoodcat> that's actually a property of the software though, so you could theoretically output in other formats with enough effort
[17:42] <cmaloney> I love that Caribou sent me a survey about their new breakfast / lunch entrees.
[17:43] <cmaloney> Save for Caribou isn't in this state anymore I'm sure I'll be the best target audience.
[17:46] <cmaloney> I don't know what they were thinking.
[17:55] <mrgoodcat> anybody have any experience in stripping down a linux so that it only runs a single program?
[17:57] <jrwren> no. why would you want to?
[17:59] <mrgoodcat> i'm not sure i do yet
[17:59] <mrgoodcat> in fact i'm sure i dont
[18:00] <mrgoodcat> but it was just the first thing that came to my head when i was brainstorming ideas
[18:00] <mrgoodcat> i want to make a yubikey door lock
[18:00] <mrgoodcat> the problem is, in order to do that I have to expose a usb port
[18:00] <mrgoodcat> so nothing is stopping someone from plugging in a usb keyboard and hitting ctrl+c
[18:00] <mrgoodcat> or ctrl+z
[18:00] <mrgoodcat> or ctrl+\
[18:01] <mrgoodcat> or any number of keyboard interrupts i'm sure exist but i'm not yet aware of
[18:01] <jrwren> pretty easy to disable those.
[18:01] <mrgoodcat> so i'm trying to find a way to disable ALL of the possible interrupts
[18:01] <jrwren> you don't even have to do that.
[18:01] <mrgoodcat> what do i have to do then?
[18:02] <jrwren> write your program to detach from term and start on system start. Then only interrupt it ctrl-alt-del, just disabled that in inittab or modern equiv
[18:02] <mrgoodcat> but i need keyboard input in the program. can i do that detached?
[18:03] <jrwren> oh, right, its just yubi.
[18:03] <mrgoodcat> actually i suppose thats a stupid question
[18:03] <jrwren> how about... don't use yubi :)
[18:03] <mrgoodcat> since keyloggers have to work somehow
[18:10] <jrwren> keyloggers run in bg as root and "sniff" the device.
[18:11] <jrwren> your original idea to deal with signals might be easiest.
[18:11] <jrwren> just noop handle sigterm and sigpipe
[18:11] <mrgoodcat> yea
[18:11] <mrgoodcat> thats what i have so far
[18:11] <jrwren> err, rather sigint, not sigterm.
[18:11] <mrgoodcat> i've ignored both
[18:11] <mrgoodcat> and sigquit
[18:11] <jrwren> i don't think term or quit can be triggered from KB.
[18:12] <jrwren> probably not a great idea to noop them.
[18:12] <jrwren> oh, quit can be.  ctrl-\
[18:12] <mrgoodcat> it is
[18:12] <jrwren> LEARN SOMETHING NEW EVERYDAY!
[18:13] <mrgoodcat> i want to noop anything that could possibly be signaled
[18:13] <jrwren> no, you really don't
[18:13] <jrwren> signals exist for a reason.
[18:13] <jrwren> you don't want to un-unixify your process.
[18:13] <mrgoodcat> sorry
[18:14] <mrgoodcat> anything that could close the program
[18:14] <jrwren> you just want to stop the KB from doing things to your process.
[18:14] <mrgoodcat> yes
[18:14] <mrgoodcat> i want keyboard signals ignored
[18:15] <mrgoodcat> ctrl+y also sends sigquit
[18:15] <mrgoodcat> nope wrong
[18:15] <mrgoodcat> ignore that
[18:15] <jrwren> stop
[18:16] <mrgoodcat> stop what?
[18:16] <jrwren> it sent sigstop.
[18:16] <mrgoodcat> oh yea
[18:18] <jrwren> mrgoodcat: you doing this in python
[18:18] <mrgoodcat> found a simple way to list all signals that can be triggered by kb stty -a| grep intr
[18:19] <mrgoodcat> yes i am
[18:19] <jrwren> because a lot of what we just talked about are shellisms which may not be true in python
[18:19] <jrwren> mrgoodcat: http://stackoverflow.com/questions/5475456/python-built-in-keyboard-signal-interrupts
[18:19] <bookiebot> http://goo.gl/7ql0FI - linux - Python: Built-in Keyboard Signal/Interrupts - Stack Overflow
[18:31] <mrgoodcat> SIGINT and SIGQUIT are the only 2 that can kill python programs from the keyboard. Here is a much simpler solution: http://hastebin.com/ezufitutif.py
[18:31] <bookiebot> http://goo.gl/fnEAQt - hastebin
[18:31] <mrgoodcat> it will print the pid on start so you can kill -9 $PID from another terminal
[18:33] <jrwren> would be interesting to toggle the disable of quit, because you really do want the nice cleanup of quit.
[18:33] <jrwren> at_exit is your friend.
[18:34] <jrwren> you could use USR to turn QUIT disable on/off
[18:40] <mrgoodcat> i was thinking of overriding the handler for one of them, and putting it into a sort of 'command' mode
[18:40] <mrgoodcat> ctrl+\ then a valid yubikey
[18:40] <mrgoodcat> then i could program new keys/terminate the program/anything else
[18:41] <mrgoodcat> the program should never quit so i dont really need at_exit
[18:42] <jrwren> hehe, *should*
[18:43] <jrwren> you going to log to disk?
[18:43] <mrgoodcat> probably
[18:44] <mrgoodcat> not that i think anybody is likely to try to break into my usb door lock, but it would be nice to have notifications of unauthorized attempts
[18:44] <mrgoodcat> i've already got a twilio account set up
[18:44] <mrgoodcat> .sms mrgoodcat hey there from twilio :)
[18:44] <bookiebot> Message sent
[18:49] <jrwren> mrgoodcat: any disk writes is a good reason to allow propery quit instead of kill.
[18:54] <mrgoodcat> yea i'll probably enable quit in command mode or something. after you're already authenticated. it's very easy to toggle ignoring the signals
[18:57] <cmaloney> mrgoodcat: You could also make your thing a daemon
[18:57] <cmaloney> that way it won't be subject to foreground signals
[18:57] <mrgoodcat> cmaloney: it needs keyboard input
[18:57] <cmaloney> though I'd question the use of Linux for this. :)
[18:58] <mrgoodcat> it doesn't need to be professional grade security. i've still got a key for the back door and we all know what a joke keys are
[18:58] <mrgoodcat> i think it's highly unlikely anybody skilled enough to worry about will be trying to break into my house
[18:59] <cmaloney> WEll, you'll want it so if something does happen the default state is locked.
[18:59] <cmaloney> Though again, if there's a fire or something similar you'll want something that can be easily defeated from the inside so you can escape
[19:00] <cmaloney> eg: don't make it so you'll need ssh in order to get out of your house.
[19:00] <mrgoodcat> lol
[19:00] <mrgoodcat> my house has multiple doors
[19:01] <cmaloney> Right, but I'm cautioning you that you don't want the "no wait, one sec" between life and death.
[19:01] <mrgoodcat> lol right
[19:02] <mrgoodcat> thats the trade-off between fail-safe and fail-secure magnetic locks
[19:02] <mrgoodcat> i have a fail-secure one in my closet though so that's more likely to be the one used. I'm not even sure i'm putting it on the front door though. Might use it for a locking cabinet or something instead
[19:03] <mrgoodcat> right now I just want to do it for fun. I'm not actually all that worried about the security aspect
[19:05] <cmaloney> You'll probably want to try it on a non-outward-facing door first
[19:07] <mrgoodcat> lol yes
[19:07] <mrgoodcat> i'm not putting an untested product on my front door. it may eventually end up on an outside door, but even that isn't guaranteed
[19:31] <jrwren> wow. https://github.com/rminnich/u-root
[19:31] <bookiebot> http://goo.gl/GWqp2c - rminnich/u-root · GitHub
[19:47] <mrgoodcat> jrwren: what is that used for?
[19:48] <jrwren> mrgoodcat: i don't know, but it is a cool hack.
[19:48] <mrgoodcat> yea it is lol
[19:57] <mrgoodcat> i can't get it to work :<
[20:01] <mrgoodcat> woah i got it working
[20:01] <mrgoodcat> it's pretty wild watching it compile the commands as they are used
[20:01] <jrwren> ha! nice!
[20:01] <jrwren> I didn't try it.
[20:03] <mrgoodcat> not sure what use this is though...
[20:03] <mrgoodcat> maybe super tiny linux
[20:03] <mrgoodcat> build only the tools you use
[20:03] <jrwren> right
[20:04] <jrwren> AND a dev environment
[20:04] <jrwren> its the start of an all Go linux.
[20:04] <jrwren> what are the 5 bins by default?
[20:05] <mrgoodcat> http://www.ubuntu.com/usn/usn-2372-1/ http://www.ubuntu.com/usn/usn-2383-1/
[20:05] <bookiebot> http://goo.gl/4LONqH - USN-2372-1: Firefox vulnerabilities | Ubuntu
[20:05] <mrgoodcat> hmm interesting... he only found 1
[20:06] <mrgoodcat> there are no bins by default
[20:07] <jrwren> must be the go compilers and linker then.
[20:10] <brousch> cmaloney: I'm heading out. See you around 6-6:30
[20:12] <mrgoodcat> jrwren: just created a new root and ls'd all the directories in the path
[20:12] <mrgoodcat> sh and ls are the only 2 there
[20:13] <mrgoodcat> and it had to build sh on start, and ls the first time i used it
[20:15] <mrgoodcat> i'd suggest NOT running this on a machine you really care about
[20:15] <mrgoodcat> it sort of borked my session and i had to reboot
[20:16] <mrgoodcat> now i don't know what else might be screwed up :<
[20:16] <jrwren> mrgoodcat: OH MY!
[20:18] <mrgoodcat> oh no it was just my path
[20:18] <jrwren> that would have been my guess.
[20:21] <mrgoodcat> jrwren: where did you find that?
[20:21] <jrwren> twitter
[20:22] <mrgoodcat> friend or someone well known?
[20:25] <cmaloney> brousch: Awesome. Thanks again!
[20:30] <cmaloney> I'm not sure I understand that code
[20:30] <cmaloney> for the all go Linux.
[20:30] <mrgoodcat> what don't you understand about it
[20:32] <cmaloney> Point? Setup? :)
[20:32] <cmaloney> http://humblebundle.com
[20:32] <bookiebot> http://goo.gl/ng1UK - Humble Mozilla Bundle: Powered by asm.js (pay what you want and help charity)
[20:33] <cmaloney> That. Is. Cool.
[20:36] <mrgoodcat> cmaloney: it doesn't seem to have a point... the setup is pretty simple though. it requires your go install to be at /go
[20:36] <mrgoodcat> then you run ./scripts/getgo
[20:36] <mrgoodcat> then you run ./README
[20:36] <mrgoodcat> and it will boot the example in a chroot
[20:36] <mrgoodcat> it can also be booted as a KVM image though
[20:37] <mrgoodcat> ok time to go home
[20:37] <mrgoodcat> see you at MUG
[20:49] <cmaloney> Have fun!
[21:13] <rick_h_> party
[21:20] <cmaloney> Evening
[21:25] <jrwren> ugh.
[21:25] <jrwren> rick_h_: isn't it like midnight there? :)
[22:35] <mrgoodcat> exactly, party time
[22:38] <mrgoodcat> cmaloney: that accident wasn't you was it?
[22:55] <cmaloney> mrgoodcat: No, just slow driving over here
[22:55] <cmaloney> Rain + Michigan driving = slowmofo
[22:59] <tony-smlr> MUG - Michigan!/usr/group (www.mug.org)  Live Meeting - Raspberry Pi and Owncloud http://www.youtube.com/watch?v=3v_M6OmuOBU
[22:59] <bookiebot> http://goo.gl/sHCTUe - October 14th 2014 - Raspberry Pi and OwnCloud - YouTube