highaltitude.log.20100831

[00:00] SpeedEvil (~user@tor/regular/SpeedEvil) left irc: Ping timeout: 240 seconds
[00:14] SpeedEvil (1000@tor/regular/SpeedEvil) joined #highaltitude.
[00:59] nlin (~norman@11.60.138.210.bn.2iij.net) joined #highaltitude.
[01:00] nlin (norman@11.60.138.210.bn.2iij.net) left #highaltitude.
[01:04] juxta_ (~blah@ppp203-122-193-94.static.internode.on.net) joined #highaltitude.
[01:08] <natrium42> juxta_, pix plz
[01:08] <natrium42> :D
[01:09] <juxta_> morning natrium42
[01:09] <natrium42> evening
[01:09] <juxta_> i out up some launch pics from horus 6 on the blog
[01:09] <juxta_> then fell asleep
[01:10] <natrium42> how did horus 7 pics turn out?
[01:10] <juxta_> i suppose i better hurry up and get the horus 7 flight pics up
[01:10] <natrium42> the sample one looks great
[01:10] <juxta_> good good
[01:10] <juxta_> www.bogaurd.net/horus7
[01:10] <juxta_> there's a few there
[01:11] <natrium42> haha --> http://www.bogaurd.net/horus7/IMG_2715.JPG
[01:11] <natrium42> chase cars, aren't they?
[01:12] <juxta_> yup :D
[01:12] <natrium42> :)
[01:13] <juxta_> okay - give me 15 minutes and i'll be back uploading the horus 7 pics, promise
[01:13] <juxta_> :)
[01:14] <natrium42> what mini cameras is adrian talking about?
[01:17] juxta_ (~blah@ppp203-122-193-94.static.internode.on.net) left irc: Ping timeout: 240 seconds
[01:31] juxta (Bogaurd@ppp203-122-193-94.static.internode.on.net) joined #highaltitude.
[01:32] <juxta> ping natrium42
[01:33] <natrium42> yo
[01:33] <juxta> so there are about 2000 photos to sort through :)
[01:34] <natrium42> eek
[01:38] Jasperw (~jasperw@93.89.81.29) left irc: Quit: Leaving.
[01:38] GeekShadow (~Antoine@reactos/tester/GeekShadow) left irc: Quit: The cake is a lie !
[01:39] TraumaPony (~TraumaPon@202-173-162-181.dyn.iinet.net.au) left irc:
[01:39] <juxta> I had 300b telemetry and a fair few listeners natrium42, which did evil things to the tracker :)
[01:40] <natrium42> yeah, i noticed
[01:40] <natrium42> should look into it
[01:40] <natrium42> i thought i fixed out-of-order positions
[01:41] <natrium42> might be on the dl-server end
[01:46] jasonb (~jasonb@dsl092-009-225.sfo1.dsl.speakeasy.net) left irc: Quit: Leaving
[01:46] <juxta> what's a good simple video editor these days?
[01:50] DrLuke (~luke@pD9E3843E.dip.t-dialin.net) joined #highaltitude.
[01:57] <natrium42> for what OS?
[02:00] <juxta> oh, windows
[02:10] <natrium42> try windows movie maker
[02:10] <natrium42> http://explore.live.com/windows-live-movie-maker
[02:11] <DrLuke> You mean windows slideshow maker?
[02:12] <natrium42> hehe
[02:12] <juxta> i just virtualdubmod usually
[02:12] <juxta> just use*
[02:13] <juxta> though it's not really authoring software
[02:13] Action: natrium42 uses adobe premiere pro :P
[02:15] <juxta> speaking of adobe, I don't know why lightroom still hasnt imported all those images
[02:20] <Darkside> hahahaha
[02:20] <Darkside> 1200 images?
[02:20] <Darkside> sorry, 2045
[02:20] <Darkside> or whatever it was
[02:24] <natrium42> juxta, you should autostich a nice panorama
[02:25] <juxta> yeah, I will do
[02:25] <juxta> I have autopano
[02:25] <Darkside> noice
[02:26] <Darkside> they move around a lot tho
[02:26] <Darkside> you might be able to do something witn the video
[02:27] <juxta> autopano scans the whole lot and looks for frames that match up Darkside
[02:28] <Darkside> nice
[02:28] <Darkside> very very cool
[02:29] <juxta> my neighbors have some building gong on
[02:29] <juxta> it sounds like someone is trying to penetrate a 2 foot concrete wall with a hamme
[02:29] <juxta> hammer
[02:32] shea_ (~shea@pool-71-164-86-204.albyny.fios.verizon.net) left irc: Ping timeout: 276 seconds
[02:34] DrLuke (~luke@pD9E3843E.dip.t-dialin.net) left irc:
[02:45] shea_ (~shea@pool-71-164-86-204.albyny.fios.verizon.net) joined #highaltitude.
[03:46] Cleo (~Cleo@unaffiliated/cleo) joined #highaltitude.
[04:16] juxta (Bogaurd@ppp203-122-193-94.static.internode.on.net) left irc: Ping timeout: 252 seconds
[04:25] BuffaloSouljah (~alberto@2002:ad13:5efb:0:20c:76ff:fefb:b0e2) joined #highaltitude.
[05:04] BuffaloSouljah (alberto@2002:ad13:5efb:0:20c:76ff:fefb:b0e2) left #highaltitude.
[05:06] BuffaloSouljah (~alberto@2002:ad13:5efb:0:20c:76ff:fefb:b0e2) joined #highaltitude.
[05:09] BuffaloSouljah (~alberto@2002:ad13:5efb:0:20c:76ff:fefb:b0e2) left irc: Client Quit
[05:10] BuffaloSouljah (~alberto@2002:ad13:5efb:0:20c:76ff:fefb:b0e2) joined #highaltitude.
[05:12] BuffaloSouljah (alberto@2002:ad13:5efb:0:20c:76ff:fefb:b0e2) left #highaltitude.
[05:28] Upu (~Upu@ubn.upuaut.net) joined #highaltitude.
[05:30] Simon-MPFH (~simon@phantom.mpfh.co.uk) joined #highaltitude.
[05:49] juxta (Bogaurd@ppp203-122-193-94.static.internode.on.net) joined #highaltitude.
[05:49] BuffaloSouljah (~alberto@2002:ad13:5efb:0:20c:76ff:fefb:b0e2) joined #highaltitude.
[05:54] smealum (~smealum@4.156.73-86.rev.gaoland.net) left irc: Ping timeout: 240 seconds
[06:01] natrium42 (~natrium@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) left irc: Quit: Leaving
[06:08] Cleo (~Cleo@unaffiliated/cleo) left irc: Quit: leaving
[07:07] Cleo (~Cleo@unaffiliated/cleo) joined #highaltitude.
[07:37] m1x10 (m1x10@ppp089210187222.dsl.hol.gr) joined #highaltitude.
[07:38] jasonb (~jasonb@adsl-66-124-73-250.dsl.sntc01.pacbell.net) joined #highaltitude.
[08:44] juxta_ (~blah@ppp203-122-193-94.static.internode.on.net) joined #highaltitude.
[08:45] juxta- (~blah@ppp203-122-193-94.static.internode.on.net) joined #highaltitude.
[08:46] juxta- (~blah@ppp203-122-193-94.static.internode.on.net) left irc: Client Quit
[08:48] juxta_ (~blah@ppp203-122-193-94.static.internode.on.net) left irc: Ping timeout: 240 seconds
[08:54] <juxta> oooh, footage is up! http://www.vimeo.com/14572413
[08:56] <m1x10> oh juxta, its beautiful
[08:56] <DanielRichman> very nice
[08:56] <m1x10> shows in detail how lands
[08:56] <m1x10> too smooth
[08:57] <m1x10> really congragulations
[08:57] <juxta> thanks :)
[08:57] <DanielRichman> juxta: you should have run for it, you could have caught that :P
[08:57] <juxta> haha
[08:58] <juxta> if you listen closely you'll hear me on the radio saying I'm on the other side of the paddock
[08:58] <DanielRichman> >.>
[08:58] <juxta> so I would have a big run :)
[08:58] <DanielRichman> :)
[08:59] <juxta> i'll get around to catching one some day ;p
[08:59] <m1x10> You knew the landing position from predictor and went there to wait for the balloon earlier?
[09:03] <juxta> well the predictor isn't that accurate
[09:03] <juxta> but it sure helps
[09:04] <juxta> ping Darkside
[09:06] GW8RAK (~chatzilla@host217-43-79-160.range217-43.btcentralplus.com) joined #highaltitude.
[09:08] <Darkside> juxta: ?
[09:09] <juxta> footage of the landing is up, thought you might like to see it :)
[09:09] <Darkside> cool
[09:10] <Darkside> very cool juxta
[09:45] fsphil (~phil@beastie.sanslogic.co.uk) joined #highaltitude.
[09:45] <fsphil> morning all
[09:48] <DanielRichman> hi
[09:51] <m1x10> hey fsphil
[09:54] <juxta> hey fsphil
[09:54] <fsphil> DanielRichman, looking at the gps stuffs -- in the fscan line, the time doesn't always have the floating point bit
[09:54] <DanielRichman> ah.
[09:55] <fsphil> that got me when I changed gps device last time, the old one did and the new one didn't
[09:55] <DanielRichman> I never read how to do optional chars with fscanf
[09:55] <DanielRichman> did you just remove the .%*2i on yours or have you already written something that accepts either?
[09:55] <fsphil> I changed the last %2u (for seconds) to a float
[09:56] <fsphil> %2u%2u%2f
[09:57] <fsphil> unfornatually fscanf can't really do optional, it's one way or nothing
[10:00] <fsphil> although I suppose %*[^,] would skip everything after the seconds
[10:01] <fsphil> %2u%2u%2u%*[^,],
[10:01] <DanielRichman> hmm, OK
[10:02] <DanielRichman> I could change it to a float and then immediately cast it to an int, but that's a bit hacky
[10:02] <DanielRichman> I don't feel like wiring up the fsa03 so I might see if I can find a gps later
[10:02] <DanielRichman> I think there's a usb cable for one
[10:02] <fsphil> the last one is probably best, parses the seconds as int and skips any floaty bits
[10:03] <DanielRichman> and I need to simulate a totally responseless one with a fifo or something to see if the pthread_kill works
[10:03] <fsphil> I've got a usb gps at home, going to test when I get outta here later
[10:03] <DanielRichman> oh, awesome :)
[10:03] <DanielRichman> I tested it by disabling rjh's baud rate setting stuff and having it read a dump I made from the lassen ages ago
[10:04] <DanielRichman> and it appeared to work. Was in offline mode since I didn't feel like telling the world where I live
[10:04] <fsphil> too late for that in my case :) my position is in a few places on the web
[10:05] <fsphil> I can stick the laptop in a back and take it out when I'm walking the dog
[10:05] <DanielRichman> I added a box to confdialog (<3 fluid) to set the "identity" of the thing.
[10:05] <fsphil> does the tracker not need a password for uploading chase positions?
[10:05] <DanielRichman> when it posts a string it sends the identity of the gps and your own callsign aiui so I don't know which it will display under
[10:06] <DanielRichman> I don't know fsphil. Actually, this might be a totaly new format
[10:06] <fsphil> aah
[10:06] <DanielRichman> RJH suggested using ZC,IDENTITY,LAT,LON,ALT
[10:06] GeekShadow (~Antoine@reactos/tester/GeekShadow) joined #highaltitude.
[10:06] <DanielRichman> but you can configure identity, so, chase_dog would be cool :P
[10:06] <fsphil> lol
[10:06] <DanielRichman> tell you what, I'll try posting some
[10:08] <fsphil> is the normal coordinates still posted when the gps is enabled?
[10:08] <fsphil> (is == are)
[10:08] <DanielRichman> yeah the coords you configure in the operator page get posted periodically anyway
[10:08] Laurenceb (~laurence@host86-137-224-95.range86-137.btcentralplus.com) joined #highaltitude.
[10:09] <DanielRichman> if you saw the habhub design we've changed that
[10:09] <fsphil> I only had a quick look, busy weekend
[10:09] <DanielRichman> we've proposed that chase cars and static listeners send the same message type to tell the server where they are, just, if you're staying stil it sends it periodically and not very often
[10:09] GW8RAK_ (~chatzilla@host217-43-79-160.range217-43.btcentralplus.com) joined #highaltitude.
[10:10] <fsphil> aaah yea that's perfect
[10:12] <DanielRichman> no, I don't think rjh has implemented the ZC string yet
[10:13] GW8RAK (~chatzilla@host217-43-79-160.range217-43.btcentralplus.com) left irc: Ping timeout: 245 seconds
[10:13] Nick change: GW8RAK_ -> GW8RAK
[10:13] Action: DanielRichman doesn't know what the chase car format was before.
[10:13] <DanielRichman> Could hack it in for the purposes of your dog walk :P
[10:18] <fsphil> I had the old format somewhere,, it's gone walkies .. (pardon the dog pun)
[10:19] <DanielRichman> Found it.
[10:19] <fsphil> http://gist.github.com/479637
[10:19] <DanielRichman> There does appear to be a password required. And it has to be posted to /tracker/track.php
[10:20] <DanielRichman> hmm. Hacking that in might be a pain
[10:20] <fsphil> yep, not much point
[10:23] <DanielRichman> well... if you're not planning on receiving any strings via fldigi I could just change listener/listen.php to tracker/track.php
[10:23] <DanielRichman> and comment out the other calls to dl_fldigi_post (or w/e I named that function)
[10:24] <fsphil> yea that'll do
[10:36] timbobel (~timboebl@5ED040EE.cable.ziggo.nl) joined #highaltitude.
[10:36] <timbobel> oh hi
[10:37] <timbobel> if someone wants a lauch, i posted it yesterday, check out my superbright led rtty transmission
[10:37] <timbobel> http://hollandshoogte.wordpress.com/
[10:37] <SpeedEvil> Random thought.
[10:38] <SpeedEvil> Never mind, it was silly.
[10:41] <fsphil> ah ha timbobel, I'm putting a really bright IR led on my payload - same sort of thing
[10:42] <fsphil> only yours is cooler
[10:42] <SpeedEvil> I have contemplated a similar thing.
[10:42] <SpeedEvil> But using several CDR lasers.
[10:43] <SpeedEvil> For a couple of watts light out.
[10:43] <fsphil> holy cow that's bright
[10:43] <SpeedEvil> And a 2nm half-width filter.
[10:43] <fsphil> what power is there in a cdr laser?
[10:44] <SpeedEvil> actually - 250mW I think is the top
[10:44] <fsphil> ah, not much but if you could aim it well then it would cover a good distance
[10:45] <fsphil> what's the filter for?
[10:46] <SpeedEvil> On the reciever
[10:47] <SpeedEvil> to block out 99.9% of the ambient light
[10:47] <SpeedEvil> http://cgi.ebay.co.uk/780nm-80mW-laser-OEM-module-3-2VDC-w-adj-lens-780nm-/360168513403?pt=LH_DefaultDomain_0 - for example
[10:47] <fsphil> the same sort of thing that covers the top of an IR remote? only better
[10:48] <SpeedEvil> no
[10:48] <SpeedEvil> not really
[10:48] <SpeedEvil> It's a complex dielectric filter
[10:48] <SpeedEvil> using layers of dfferent refractrive index to make a interference filter
[10:50] <fsphil> ah -- I was just going to hack an old remote, but your way is better
[10:55] <fsphil> timbobel, are you going to fly that or was it just a test?
[10:58] <DanielRichman> arrr, it's a bug!
[10:58] <DanielRichman> progdefaults.myLat.at(progdefaults.myLat.length() - 1)
[10:58] <Randomskk> ohno, a bug!
[10:58] <Randomskk> this is why unit testing
[10:58] <fsphil> killz it!
[10:59] <DanielRichman> d/w it'z dead
[10:59] <Randomskk> squash
[10:59] <Randomskk> now write a test case in case it comes back
[10:59] <DanielRichman> have you seen the fldigi codebase?
[10:59] <Randomskk> heck no
[10:59] <Randomskk> I don't think it even has a test library or anything linked to it
[10:59] <Randomskk> no time like the present to start. google test is great
[11:00] <Randomskk> oh, and it uses autotools
[11:00] <DanielRichman> wat
[11:00] <Randomskk> that'd be loads of fun to do automated testing on make with. ouch.
[11:00] <DanielRichman> oh, fldigi uses autotools?
[11:00] <Randomskk> pretty sure it deos?
[11:00] <Randomskk> does*
[11:01] <DanielRichman> yeah, thought you meant google test uses autotools for a sec
[11:01] <Randomskk> no, it uses cmake
[11:01] <DanielRichman> it's like, if i want to do anything, I spend more time ./configuring than compiling
[11:01] <Randomskk> heh
[11:01] <DanielRichman> though obviously not if I keep the objects lying around
[11:02] <DanielRichman> and like, what is up with -O2 by default? I have to make CFLAGS="-g -O0" -j4 -l
[11:02] <Randomskk> cmake is a lot nicer
[11:02] <Randomskk> though I'm still not a massive fan of its syntax
[11:03] <Randomskk> or .txt on its files
[11:03] <Randomskk> but like, it makes things much nicer.
[11:03] <DanielRichman> cmake - hard to get it to do anything other than build C
[11:03] <Randomskk> hmm maybe
[11:03] <Randomskk> I've only had to do C++ with it
[11:03] <Randomskk> which it is very easy to do
[11:10] <timbobel> fpshil: just a test, i would like to use, but have not thought of a good application
[11:12] <fsphil> just for looks surely ;)
[11:15] <fsphil> the range of an optical system follows the same rules as radio doesn't it? them being the same thing
[11:15] <Randomskk> same thing at different wavelengths
[11:15] <Randomskk> but the technology used tends to vary quite massively
[11:15] <Randomskk> it's a bit hard to modulate optical frequencies. or generate them in the traditional sense
[11:15] <fsphil> yea
[11:15] <Randomskk> but you can OOK them :P
[11:16] <fsphil> though that would be cool .. all that lovely bandwidth to play with :)
[11:16] <Randomskk> I know right, terahertz+ of bandwidth
[11:16] <fsphil> I wonder how much gain my telescope had, in db
[11:16] <Randomskk> haha
[11:17] <Randomskk> it's just like a yagi right
[11:17] <Randomskk> (actually someone has made an optical frequency yagi)
[11:17] <fsphil> I remember reading that
[11:17] <SpeedEvil> fsphil: It can be really quite high.
[11:17] <Darkside> man thats a cool concept
[11:18] <Darkside> optical yagis :P
[11:18] <SpeedEvil> fsphil: For example - a 0.5cm lens has a beamwidth of a milliradian. This is about a gain of 10 million over an isotropic emitter
[11:18] <SpeedEvil> fsphil: Or 70dB
[11:18] <fsphil> nice!
[11:19] <SpeedEvil> 5cm of course is 20dB higher, ...
[11:20] <fsphil> my small scope has about a 5cm lens
[11:21] <fsphil> I bet the receiver would be fairly inefficient though
[11:21] <Randomskk> photodiodes aren't awful
[11:21] <SpeedEvil> You get single photon detectors
[11:21] <Randomskk> APDs and things
[11:21] <SpeedEvil> the real problem is the background noise.
[11:22] <SpeedEvil> hence the above mention of really, really tright filters.
[11:22] <Randomskk> 2nm is pretty tight
[11:24] <SpeedEvil> yeah - slightly wider - 20nm - filters are quite cheap
[11:24] <SpeedEvil> http://cgi.ebay.co.uk/IR-30mm-30-mm-Filter-Infrared-Lens-780nm-780-Standard-/150485276756?pt=Camera_Filters
[11:28] <fsphil> shame Luxeon don't make IR leds
[11:30] <SpeedEvil> They dio
[11:30] <SpeedEvil> do
[11:31] <Laurenceb> you can do very high frequency prn
[11:31] <Laurenceb> but you also need an FPGA to handle all the data
[11:32] <fsphil> I'd be keeping this very slow
[11:32] <Laurenceb> and fast adc to digitise each photon
[11:37] <fsphil> SpeedEvil, do you have a link -- I'm not having much luck finding them
[11:37] <SpeedEvil> for?
[11:38] <fsphil> luxeon IR led
[11:39] Action: Laurenceb rages at raster -> vector convertor code
[11:39] <Laurenceb> think i need to read this http://www.svgopen.org/2005/papers/Prasad_Abstract_R2V4SVG/
[11:40] <SpeedEvil> hmm
[11:40] <SpeedEvil> I may have been misremembering
[11:40] <fsphil> I've found a youtube video with one, they do seem to exist
[11:41] <fsphil> or maybe it's another brand and people are just calling them luxeon
[11:41] <SpeedEvil> http://www.dealextreme.com/details.dx/sku.14061
[11:42] <fsphil> haha... "I tested and play with it for about 30 minutes and after that my eyes were hurting"
[11:42] <SpeedEvil> http://www.dealextreme.com/search.dx/search.IR
[11:42] <SpeedEvil> also
[11:42] <SpeedEvil> :)
[11:43] <fsphil> so I figure pulsing something like that with the data should be detectable on the ground
[11:46] <Laurenceb> maybe modulate it at say 1MHz
[11:46] <fsphil> yea
[11:46] <Laurenceb> then have a filter
[11:46] <Randomskk> I wonder if you could transmit live images at a useful baudrate
[11:46] <Randomskk> I'm thinking a mirror
[11:46] <juxta> hey Randomskk
[11:46] <Randomskk> yo
[11:47] <Randomskk> a mirror and a telescope on the ground
[11:47] <juxta> did you build ferret, or was that someone else?
[11:47] <fsphil> that's my thought Randomskk
[11:47] <Randomskk> uh, jonsowman and I did
[11:47] Laurenceb (~laurence@host86-137-224-95.range86-137.btcentralplus.com) left irc: Remote host closed the connection
[11:47] <Randomskk> fsphil: I'm talking parallel, analogue transmission of live images :P
[11:47] <fsphil> we've got a mountain about 10km away, so I can test at distance
[11:48] <fsphil> aah I'm thinking of digital, at about 9600 baud :)
[11:48] <Randomskk> :P
[11:48] <juxta> righto - I was contacted by a ham in the UK hoping to launch rather soon who has been having trouble with their telemetry systems
[11:48] <Randomskk> heh
[11:48] <Randomskk> ferret is around somewhere in cambridge
[11:48] <Randomskk> where are they launching from?
[11:48] <Randomskk> (do they have a notam?)
[11:48] <juxta> not sure - I will ask
[11:48] <fsphil> could an LED be modulated enough to carry a PAL video signal anyway?
[11:49] <Randomskk> pal isn't that high frequency is it?
[11:49] <Randomskk> it is analogue though
[11:49] <Randomskk> juxta: I mean, we can dig it out, I believe it's in the lab somewhere
[11:49] <Randomskk> but it is really quite hacky, doesn't currently have an antenna and the software is a bit rubbish
[11:49] <Randomskk> it does work, admittedly
[11:49] <juxta> it had a GPS bug at some stage didn't it?
[11:49] <fsphil> wasn't there W/E bug in ferret?
[11:49] <Randomskk> yea, meridian crossing
[11:50] <fsphil> though easy to fix
[11:50] <Randomskk> but we fixed that after the first flight by taking it into a car, driving over the meridian, then sitting in a layby coding until it worked, then driving back
[11:50] <fsphil> lol
[11:50] <Randomskk> well jonsowman drove, I coded
[11:50] <SpeedEvil> fsphil: yes.
[11:50] <SpeedEvil> fsphil: The quesrtion would be SNR
[11:52] <juxta> Randomskk, where is EARS?
[11:52] <jonsowman> juxta: go onto the predictor and select EARS
[11:52] <juxta> good idea
[11:53] <juxta> oh, right next to Cambridge?
[11:53] <jonsowman> :)
[11:53] <jonsowman> not far away certainly
[11:54] <juxta> this guy's callsign is listed as Grimsby
[11:54] <juxta> looks like a fair distance
[11:55] <jonsowman> grimsby is a reasonably long way away from cam
[11:55] <Randomskk> dunno where he plans to launch from
[11:55] <Randomskk> atm I'm not in cam though jonsowman and I were thinking of driving up at some point before we go back to uni
[11:56] <Randomskk> so also depensd when he plans to launch I guess
[11:56] <juxta> he said within 2 weeks
[11:56] <Randomskk> that implies he either has a notma
[11:56] <Randomskk> or doesn't know he needs one
[11:56] <juxta> yeah
[11:57] <juxta> i'm asking in the email back, but shall I pass on your contact details too if you'd be happy to loan ferret?
[11:57] <Randomskk> happy to loan ferret if I'm in cambridge
[11:57] <jonsowman> ditto
[11:57] <juxta> alrighty
[11:57] <jonsowman> happy to drive to cambridge, infact
[11:58] <juxta> what contact details shall I give then?
[11:58] <Randomskk> tbh #highaltitude ?
[11:58] <jonsowman> cusf mailing list
[11:58] <Randomskk> that too
[11:58] <juxta> okay
[11:58] <juxta> will do :)
[11:58] <Randomskk> get them on irc anyway though
[11:58] <jonsowman> yea
[12:00] TraumaPony (~TraumaPon@202-173-162-181.dyn.iinet.net.au) joined #highaltitude.
[12:05] <juxta> hey jonsowman, what's the url for the CUSF hourly predictions setup?
[12:05] <jonsowman> www.cuspaceflight.co.uk/hourly-predictions
[12:06] <juxta> oh good, that one has the bug too, it's not just me :)
[12:06] earthshine (~earthshin@cpc3-orpi1-0-0-cust867.2-3.cable.virginmedia.com) left irc: Read error: Connection timed out
[12:06] <Randomskk> bug?
[12:06] <juxta> if you click on a launch, the date seems to be showing incorrectly at the moment
[12:06] <juxta> one month in advance
[12:06] <jonsowman> hum
[12:06] <jonsowman> that's interesting
[12:06] <jonsowman> fcvo interesting
[12:07] <Randomskk> well
[12:07] <Randomskk> august works
[12:07] <Randomskk> then 1st october
[12:07] <jonsowman> what the
[12:08] <juxta> hey we got a landing on film the other day, woo
[12:08] <jonsowman> ooo
[12:08] <jonsowman> can I see?
[12:08] <juxta> http://www.vimeo.com/14572413
[12:08] <jonsowman> :D
[12:11] <juxta> you can hear me on the radio saying I'm on the other side of the paddock
[12:11] <juxta> shortly before the farmer arrived
[12:11] <fsphil> 20 minutes to lunch, then I can watch it :)
[12:11] <juxta> by some freak coincidence the met office balloon launched the day before also landed in the same field, so he gave us that one
[12:12] <jonsowman> juxta: that's fantastic
[12:12] <juxta> it made our day :D
[12:13] Action: Randomskk wants to watch, will have to wait until after work :(
[12:13] <juxta> we have some pretty interesting ideas for future launches, after the success of the airborne repeater
[12:14] <fsphil> juxta, was the met balloon launched near your own launch site?
[12:15] <juxta> we're looking at relaying car telemetry (locations, speed etc) via the payloads telemtry stream to lever the listener network when we dont have net coverage
[12:16] <juxta> also considering having everything (the cars & payload) on an IP network over radio using DSTAR radios
[12:16] <juxta> fsphil, about 60km away I would say
[12:17] <Randomskk> that would be fun
[12:17] <juxta> and the balloon had not burst - it had a tear in the side a few cm long
[12:17] <Randomskk> hmm
[12:17] <Randomskk> as in, at failure?
[12:17] <jonsowman> that's odd
[12:17] <Randomskk> at altitude?
[12:17] <juxta> not my balloon - the met office one
[12:17] <Randomskk> oh, right
[12:17] <juxta> it was just found in the field like that
[12:18] <fsphil> that's weird
[12:18] <juxta> our balloons (the chinese ones) went really well
[12:18] <jonsowman> juxta: what size balloon was horus 7?
[12:18] <juxta> I topped the highest alt on the UKHAS record chart using a 1000g balloon
[12:18] <juxta> 1000g
[12:19] <jonsowman> 35km+ from a 1000g balloon
[12:19] <jonsowman> that's seriously impressive
[12:19] <SpeedEvil> what was the payload anad ascent rate?
[12:19] earthshine (~earthshin@cpc3-orpi1-0-0-cust867.2-3.cable.virginmedia.com) joined #highaltitude.
[12:20] <juxta> SpeedEvil, 500g payload, 4.5m/s asc
[12:20] <juxta> the day before we flew ~950g with about the same asc and got 34.something km
[12:20] <SpeedEvil> It would be interesting to graph top at 100,250,500,750g payload with the same balloons
[12:20] <juxta> yeah
[12:20] <juxta> I will be ordering more soon
[12:21] <juxta> I have some 600g ones
[12:21] <juxta> and a 2kg one too
[12:21] <fsphil> 2kg!!
[12:21] <Randomskk> juxta: bear in mind the burst calculator won't work with those balloons until we get new data for it
[12:21] <juxta> Randomskk, yup, i know
[12:21] <juxta> I just treated it as I would a kaymont/totex one
[12:21] <juxta> wind speed at max alt is negligible anyway
[12:22] <Randomskk> true I guess
[12:22] <Randomskk> mostly things like its predicted burst alt and time and such will be a bit wrong
[12:22] <juxta> in fact the manufacturer claims less of their balloons than totex/kaymont do
[12:22] <Randomskk> curious
[12:22] <Randomskk> well if you keep a record of how all your flights do we can sart making data :P
[12:22] <juxta> I expected the burst alt to be slightly lower
[12:23] <juxta> they're all archived, including the flight paths :)
[12:23] <Randomskk> excellent
[12:23] <jonsowman> give that data to eureqa
[12:24] <juxta> who/what is eureqa?
[12:24] <jonsowman> wonder if it would find anything interesting
[12:24] <jonsowman> juxta: http://ccsl.mae.cornell.edu/eureqa
[12:24] <Randomskk> it might well come up with some good correlations
[12:24] <jonsowman> there's the potential for it
[12:24] <Randomskk> depends if there's a manufacturing process difference for the different weights that would account for the different bursts though
[12:24] <jonsowman> there's also the potential for it to tell us everything we already know
[12:24] <juxta> oh right yes I think I stumbled upon this once
[12:28] <fsphil> DanielRichman, ever used sqlite -- think it might be useful for queueing up data in dl-fldigi when offline?
[12:29] <Randomskk> it's definitely an option
[12:30] <Randomskk> the new plan involves some kind of spool but no hard choices with implementation yet
[12:30] <Randomskk> sqlite is fairly neat but other solutions might be better for this, it's a pretty simple problem
[12:30] <fsphil> means even if it crashes the data should persist
[12:30] <fsphil> true
[12:31] <jonsowman> hehe this motherboard has little blue LEDs on it to indicate current processor laod
[12:32] <jonsowman> how stunningly unnecessary
[12:33] <Randomskk> pretty though
[12:33] Action: SpeedEvil sighs.
[12:33] <SpeedEvil> I want to make a clock for ebay.
[12:34] <Randomskk> as in, a clock to sell on ebay, or a clock that monitors ebay items?
[12:34] <SpeedEvil> clock to sell
[12:34] <SpeedEvil> An olde-fashioned CRT
[12:35] <Randomskk> ah cool
[12:35] <SpeedEvil> blown glass and mahogony
[12:35] <Randomskk> yea I've seen some nice CRT clocks
[12:35] <Randomskk> would be pretty
[12:37] <SpeedEvil> http://members.chello.nl/~h.dijkstra19/page3.html - the braun tube at the top
[12:37] <SpeedEvil> Similar to - but with antique looking wound deflection coils
[12:37] <Randomskk> pretty
[12:37] <SpeedEvil> And more brass
[12:46] Darkside (~Darkside@hpavc/darkside) left irc: Ping timeout: 276 seconds
[12:46] <Randomskk> you can never have too much brass
[12:46] <Randomskk> actually you can
[12:46] <Randomskk> someone makes steampunk sex toys
[12:46] <Randomskk> they're shaped like brass laser pistols but with barrels that are perhaps a bit unusual
[12:47] G4TNX (8a20b0fe@gateway/web/freenode/ip.138.32.176.254) joined #highaltitude.
[12:47] G4TNX (8a20b0fe@gateway/web/freenode/ip.138.32.176.254) left irc: Client Quit
[12:47] Darkside (~Darkside@hpavc/darkside) joined #highaltitude.
[12:47] <SpeedEvil> ...
[12:48] <Randomskk> but that edge case aside
[12:48] <Randomskk> brass is great
[12:50] alikins_ (~adrian@cpe-066-057-101-225.nc.res.rr.com) left irc: Quit: alikins_
[12:51] <SpeedEvil> But a CRT clock, with a fake 'driver' going to the coils, comprising a fake radiometer rotating at 1hz
[12:54] gonzo_ (~gonzo_@cpc1-pool8-0-0-cust154.15-1.cable.virginmedia.com) joined #highaltitude.
[13:02] <fsphil> expensive enough? http://uk.farnell.com/osram/sfh4750/ir-ostar-850nm/dp/1827022
[13:02] <jonsowman> :o
[13:07] <SpeedEvil> Well...
[13:07] <SpeedEvil> 3.5W out in a 5mm package is very handy
[13:08] <SpeedEvil> Combine that with a teeny reflector to give you a 10mm or so reflector hitting a 150mm lens, and you've got a 4 degree quite bright IR spot
[13:53] <griffonbot> @nearsys: The NearSpace Ultralight PCbs are on their way. I'll need to run a test and get the radios (two are on order). Kit out soon. #ARHAB [http://twitter.com/nearsys/status/22617855498]
[13:57] <DanielRichman> fsphil: http://gist.github.com/559051 uses the (temporary) testing script that implements the ZC, string
[13:57] <DanielRichman> http://spacenear.us/tracker/
[13:59] <fsphil> aah gotcha
[13:59] <m1x10> is it possible to do arduino soft reset?
[13:59] <m1x10> sorry for interrupting
[14:02] <DanielRichman> m1x10: easiest way is to turn on the watchdog with a very low timeout and nothing that will reset it
[14:02] <DanielRichman> Iirc the arduino bootloader will turn it off aswell just to make sure
[14:02] <DanielRichman> fsphil: http://github.com/danielrichman/dl-fldigi/tree/testing
[14:03] <m1x10> ok, watchdog is not a solution for me right now.
[14:03] <m1x10> Something else
[14:03] <m1x10> When asking permission for NOTAM
[14:03] <m1x10> they may respond in a month. ok.
[14:04] <m1x10> do they give a certain day to launch?
[14:04] <fsphil> DanielRichman, applied -- will test as soon as i get home
[14:04] <DanielRichman> awesome fsphil :). I might see if I can get this gps hooked up but I don't have a laptop that I can run around with atm
[14:04] <fsphil> they may not respond at all :/
[14:05] <fsphil> DanielRichman, wifi is a bit patchy in areas, but it should work for most of the route
[14:05] <DanielRichman> you and your epic win wifi >.>
[14:06] <Randomskk> 3g's where it's at
[14:06] <m1x10> fsphil, assuming they respond. do they give you a certain day to launch?
[14:07] <DanielRichman> m1x10: why can't you use the watchdog?
[14:07] <juxta> m1x10, check with your local authorities
[14:07] <fsphil> in the UK at least, you pick the day or days when you submit the form.
[14:07] <juxta> the procedure for NOTAMS will vary from country to country
[14:08] <m1x10> ok, thx
[14:09] <m1x10> DanielRichman, I already use WDT for other things.
[14:09] <fsphil> actually this would be a pretty good way to map the wifi coverage
[14:09] <DanielRichman> m1x10: so what? whatever resets the wdt - when you want to soft reset, turn that bit off
[14:09] <DanielRichman> and the wdt will reset the cpu
[14:09] <m1x10> hmm
[14:09] <DanielRichman> what are you using it for?
[14:10] <m1x10> but wdt is resetting only the ¼C. Not whole arduino.
[14:10] <DanielRichman> ...
[14:10] <fsphil> :)
[14:10] <DanielRichman> well there's not much to the arduino other than the uc
[14:11] <m1x10> yes, I was thinking of doing actually a hardware reset to the shields above
[14:11] <m1x10> to cut the power and turn it on again
[14:11] <DanielRichman> sounds like you'd have to do something ugly with a relay
[14:12] <m1x10> Ok its going to far if im going to use a relay
[14:12] <DanielRichman> idk if you would be able to use some solid state components but if you want to actually cut the power you actually have to cut the power
[14:13] <DanielRichman> the power delivered to the shilds is not controlled by the uC
[14:13] <m1x10> y sure
[14:13] <m1x10> so i need something mechanical
[14:13] <m1x10> got it
[14:13] <m1x10> ppl check this plz: www.eeeke.gr
[14:13] <m1x10> it has english title too
[14:14] <m1x10> that should be the correct website to ask for notam
[14:14] <m1x10> can anyone send me an example notam text?
[14:14] <m1x10> I mean how did u explain the hab thing to them
[14:14] <m1x10> If anyone has his text I would love to read it :)
[14:15] <juxta> that doesn't look like the right place to apply. a quick google suggests this is more like it: http://www.hcaa.gr
[14:16] <juxta> you need to get in touch with them and let you know you want to launch - submitting a NOTAM directly is likely not what they will want you to do
[14:16] <juxta> it's certainly not what I do here, and I don't think that's the way it's handled in the UK either
[14:16] <Randomskk> you'll also want a list of things they require you comply with, too
[14:17] <Randomskk> e.g. they may need radar reflectors, certain maximum payload weights, emergency cutdown or tracking
[14:17] <juxta> let them know*, even
[14:17] <juxta> best to give them a call or send them an email and see what's required
[14:17] <Randomskk> yea
[14:17] <m1x10> yeah
[14:18] <m1x10> better phone call for first time
[14:19] <m1x10> Im sure they will not understand me when Ill tell them that Im lauching a balloon at near space
[14:19] <m1x10> they might close the phone
[14:19] <m1x10> :)
[14:31] <timbobel> juxta:great vid
[14:31] <juxta> thanks timbobel
[14:32] <fsphil> how many people where at the landing? sounded like a crowd :)
[14:33] <juxta> 3 cars worth
[14:33] m1x10 (m1x10@ppp089210187222.dsl.hol.gr) left irc: Ping timeout: 272 seconds
[14:36] <Darkside> juxta: VK5PFAW?
[14:36] <Darkside> thats on the post :P
[14:38] <juxta> fixed
[14:38] <juxta> too tired
[14:38] <juxta> I'm off to bed!
[14:39] m1x10 (m1x10@79.167.238.143) joined #highaltitude.
[14:45] timbobel (~timboebl@5ED040EE.cable.ziggo.nl) left irc: Ping timeout: 240 seconds
[15:00] rharrison_ (~rharrison@80.4.177.225) joined #highaltitude.
[15:01] <rharrison_> Hey Randomskk
[15:05] <Randomskk> hi
[15:33] jasonb (~jasonb@adsl-66-124-73-250.dsl.sntc01.pacbell.net) left irc: Ping timeout: 252 seconds
[15:53] jasonb (~jasonb@m3b0536d0.tmodns.net) joined #highaltitude.
[15:53] BuffaloSouljah (~alberto@2002:ad13:5efb:0:20c:76ff:fefb:b0e2) left irc: Ping timeout: 272 seconds
[15:55] <jshriver> good morning
[15:55] <jshriver> what can cause asterisk to constantly drop calls, even though there is ample bandwidth
[15:55] Laurenceb (~laurence@host86-137-224-95.range86-137.btcentralplus.com) joined #highaltitude.
[15:55] <Laurenceb> yo
[15:56] <jshriver> calls meaning site to site using iaxy
[15:56] <Laurenceb> http://www.seeedstudio.com/depot/ethernet-shield-pcb-p-179.html?cPath=101_102 <- the black n white looks really nice
[15:57] DrLuke (~luke@pD9E3A7CB.dip.t-dialin.net) joined #highaltitude.
[15:59] <Randomskk> Laurenceb: yea. some of my stm32 boards are in white
[15:59] <Randomskk> it's good
[15:59] <Randomskk> fairly easy to see traces too, unlike black
[15:59] <Laurenceb> yeah
[16:00] <Laurenceb> i was thinking youd miss the traces, but it doesnt look like it
[16:00] <Laurenceb> ive got the logo vectorised... but my code is buggy
[16:00] <jshriver> Is there a way of finding out why a call gets dropped?
[16:00] <Laurenceb> had to manually fix it
[16:00] <Randomskk> jshriver: not really the room to ask.
[16:00] <Laurenceb> im going to give up on the image2gerber.py code - might finish it properly later
[16:01] <Laurenceb> as itd be quite useful for a few things
[16:01] <jshriver> lol sorry wrong channel sorry
[16:02] <Laurenceb> Randomskk: have you tried asking seeedstudio to do a specific thickness pcb?
[16:02] <jshriver> <-- burned out sorry
[16:02] <Laurenceb> imo 0.8 or 1mm would be best
[16:04] <Randomskk> nope, I haven't
[16:08] BuffaloSouljah (~alberto@173-19-94-251.client.mchsi.com) joined #highaltitude.
[16:10] jasonb (~jasonb@m3b0536d0.tmodns.net) left irc: Ping timeout: 245 seconds
[16:12] <W0OTM> anyone used FRS radios in a balloon?
[16:12] <W0OTM> or GMRS
[16:14] jasonb (~jasonb@m3b0536d0.tmodns.net) joined #highaltitude.
[16:21] Laurenceb (~laurence@host86-137-224-95.range86-137.btcentralplus.com) left irc: Remote host closed the connection
[16:47] jasonb (~jasonb@m3b0536d0.tmodns.net) left irc: Ping timeout: 276 seconds
[16:59] <alikins> sent up an sd850 in a balloon last week, and logs the lens/sensor/batt temperatures. But looking at the numbers, they seem off. lens temp never got below 18C despite air temps at altitude being 0C and lower
[16:59] jasonb (~jasonb@dsl092-009-225.sfo1.dsl.speakeasy.net) joined #highaltitude.
[17:06] <SpeedEvil> cameras get really quite toasty at ground level if you keep taking pics
[17:06] mixio (m1x10@79.167.238.143) joined #highaltitude.
[17:08] m1x10 (m1x10@79.167.238.143) left irc: Ping timeout: 240 seconds
[17:08] <alikins> and it was pretty well insulated (aside from the lens sticking out into the air)
[17:09] <alikins> It got pretty toasty once it hit the ground and continued to take pics (up to about 70C for the batter, and 45C for the lens on a 30C day
[17:10] Hiena (~Hiena@81.93.195.181.datatrans.hu) joined #highaltitude.
[17:10] <alikins> I was expecting the lens temp to get a lot lower though, so trying to make sense of the numbers
[17:11] <alikins> https://spreadsheets.google.com/pub?key=0AnbpMcx9jw2KdHctQ21kSlhIY0N3VWwyUWdTTVFsM3c&hl=en&gid=1
[17:15] <rharrison_> For them who are interested pics from esquire launch last may http://www.flickr.com/photos/30721501@N05/sets/72157624725057385/
[17:19] fsphil|eee (~fsphil@2001:470:1f09:483:215:afff:fe9f:bf8e) joined #highaltitude.
[17:19] <W0OTM> rharrison_: was you camera lense our in the open
[17:19] <fsphil|eee> DanielRichman, seg faults when I change any of the gps conf boxes
[17:19] <W0OTM> grr, I can't type
[17:19] <DanielRichman> graaah :( fsphil|eee so sorry :P
[17:19] <DanielRichman> any chance of a backtrace?
[17:20] <fsphil|eee> rebuilding atm, bit slow -- will try
[17:20] <DanielRichman> fsphil|eee: 32 or 64bit?
[17:20] <rharrison_> no behind uv filter
[17:20] <fsphil|eee> 32
[17:20] <DanielRichman> ok. Two things; 1 I can't build binaries for you, 2 that might have caused something
[17:22] <W0OTM> rharrison_: how did your prevent fogging?
[17:22] <fsphil|eee> crash is somewhere in dl_fldigi_gps_setup
[17:22] <W0OTM> rharrison_: I launch one this past sat, and got serious fogging 15min into the flight
[17:23] <alikins> me too
[17:23] <W0OTM> rharrison_: all my picts at altitude were blurry
[17:23] <DanielRichman> rharrison_: did you flush that one with He?
[17:23] <W0OTM> DanielRichman: what do you mean flush?
[17:23] <DanielRichman> helium
[17:23] <DanielRichman> to get rid of moisture inside
[17:24] <W0OTM> DanielRichman: oh really, is that a way to prevent fogging?
[17:24] <DanielRichman> I don't know, it might be
[17:24] <W0OTM> I see most people just seem to leave their cam lense exposed
[17:26] <DanielRichman> fsphil|eee: I see it
[17:27] <alikins> W0OTM: do you have in flight pics posted?
[17:27] <W0OTM> the foggy ones?
[17:28] <alikins> W0OTM: sure. ours ended up pretty foggy on the first launch (http://www.flickr.com/photos/alikins/sets/72157624837242750/)
[17:28] <rharrison_> DanielRichman, Randomskk I'm going to have to cancel 8 pm as teenager has gone missing
[17:28] <rharrison_> Sorry
[17:28] <DanielRichman> fsphil|eee: I have pushed a "fix"; though I'm not 100% sure it's correct
[17:28] <DanielRichman> OK rharrison_
[17:28] <fsphil|eee> k
[17:28] <W0OTM> alikins: let me get you a link
[17:29] <DanielRichman> fsphil|eee: I'm guessing it's getting NULL strings - could you try it? I didn't think .c_str() returned NULL
[17:29] <fsphil|eee> building
[17:30] <fsphil|eee> no, segging
[17:30] <W0OTM> http://www.w0otm.com/iHAB/iHAB-1/Payload_Photos/
[17:30] <DanielRichman> grah. Sorry about this :'(
[17:30] <alikins> W0OTM: some of our pics actually look like icing on the lens
[17:30] <fsphil|eee> it's crashing in the thread, that's all I've found so far
[17:30] <DanielRichman> fsphil|eee: are you using gdb or fldigis' built in stacktrace?
[17:30] <W0OTM> alikins: diddo
[17:30] <fsphil|eee> built in
[17:30] <W0OTM> http://www.w0otm.com/iHAB/iHAB-1/Payload_Photos/IMG_3244.JPG
[17:30] <fsphil|eee> actually I'm ignoring that too, just peppering the code with fprintf's
[17:31] <DanielRichman> do you have gdb installed? did you compile with -g or -O0?
[17:32] <DanielRichman> also, <3 the fprintfs :D
[17:32] <fsphil|eee> not sure about build flags, segs instantly in gdb
[17:32] <fsphil|eee> this keyboard is too small for hacking :)
[17:32] <W0OTM> alikins: what you think?
[17:33] <DanielRichman> fsphil|eee: you're not out in the field or anything are you?
[17:33] <LA3QMA> coffetime
[17:33] <fsphil|eee> nah, still inside
[17:34] <fsphil|eee> itsopening the port fine
[17:34] <alikins> W0OTM: condensation/freezing would be my guess. Maybe even tried to focus on the fog/ice?
[17:35] <alikins> W0OTM: does the exif focus info look like it might of tried to focus close?
[17:35] <W0OTM> alikins: so i guess ill remove my lens filter on the outside, and let the lense just stick out
[17:35] <DanielRichman> ok cool. 32bit VM is booting
[17:35] <DanielRichman> fsphil|eee: if you can find which line it is that'd be awesome. iirc gps_setup is quite a small function ?
[17:35] <fsphil|eee> dl_fldigi: dl_fldigi_gps setup begin
[17:35] <fsphil|eee> Signal of Class 10 received. Exiting
[17:36] <alikins> W0OTM: yeah, ours was just sticking out. It got a little fog, but not as much as yours. you can see the fog/ice on ours a little better in the time lapse http://www.youtube.com/watch?v=7iL__ed8Eyc
[17:36] <fsphil|eee> haha, SIGUSR1 == 10
[17:36] <fsphil|eee> it's not killing itself is it?
[17:36] <DanielRichman> it better not be
[17:37] <alikins> W0OTM: it was pretty humid the day we launched, so flushing the camera or otherwise trying to dry it might of helped
[17:37] <DanielRichman> fsphil|eee: if that's the debug output; can you confirm that gps_init was called before that happened?
[17:37] <DanielRichman> it should announce itself to stderr
[17:38] <fsphil|eee> the thread is started, and the serial port is opened
[17:39] <DanielRichman> and *then* it dies?!
[17:39] <fsphil|eee> the crash happens as soon as I change anything in the dialog
[17:39] <DanielRichman> right that makes sense, it will send a signal when that happens.
[17:39] <DanielRichman> static void empty_handler(int sig)
[17:39] <DanielRichman> {
[17:39] <DanielRichman> if (!pthread_equal(pthread_self(), serial_thread_id))
[17:39] <DanielRichman> {
[17:39] <DanielRichman> raise(SIGUSR2);
[17:39] <DanielRichman> }
[17:39] <DanielRichman> }
[17:40] <DanielRichman> but it shouldn't do anything
[17:40] spenguin[work] (~penguin@59.162.86.164) left irc: Quit: leaving
[17:41] <DanielRichman> and even so, it definitely shouldn't die by sigusr1
[17:41] <fsphil|eee> is that meant to be SIGUSR2 in the code above?
[17:42] <DanielRichman> no. SIGUSR2 was a poor choice by me there
[17:42] <DanielRichman> fldigi uses SIGUSR2 for the same purpose as I'm using SIGUSR2 in some other code
[17:42] <fsphil|eee> it's not even when the signal is send, it's shortly after
[17:44] <DanielRichman> oops. I only let the 32bit vm use 1 cpu. Building might take a while
[17:47] <DanielRichman> CDs are so slow, I could have downloaded & used the iso in this time >.<
[17:51] mixio (m1x10@79.167.238.143) left irc: Quit: I eat electrons for breakfast and I vomit thunders
[17:51] <DanielRichman> fsphil|eee: so it's the signal that's defo. causing it to die?
[17:52] <fsphil|eee> looks like it survives the signal, it's whatever happens after that
[17:52] <DanielRichman> fsphil|eee: are you familiar with gdb?
[17:53] <fsphil|eee> used it, but far from a master
[17:54] <DanielRichman> fsphil|eee: did you build with make, or supply extra cflags like -O0 or -g?
[17:54] <fsphil|eee> whatever the defaults are, -g -O2 I think
[17:54] <DanielRichman> hmm; OK
[17:55] GW8RAK (~chatzilla@host217-43-79-160.range217-43.btcentralplus.com) left irc: Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716]
[17:58] <fsphil|eee> brb
[17:58] <fsphil|eee> rebuilding with what I think is -O0 and -g
[17:59] <rharrison_> DanielRichman, Randomskk can make it given up reported him missing to police
[17:59] <rharrison_> heading home
[17:59] rharrison_ (~rharrison@80.4.177.225) left irc: Quit: Leaving
[18:00] <DanielRichman> fsphil|eee: I use make CFLAGS="-O0 -g" -j4 -l
[18:00] <DanielRichman> also @rjh :o
[18:00] <DanielRichman> doesn't sound too good.
[18:01] natrium42 (~natrium@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) joined #highaltitude.
[18:14] <DanielRichman> fsphil|eee: thank you for your help but I've gtg and will be back in a bit; if I find the bug I'll give you a shout ;)
[18:23] <fsphil|eee> DanielRichman, likewise
[18:48] timbobel (~timboebl@5ED040EE.cable.ziggo.nl) joined #highaltitude.
[18:48] <timbobel> whooh tomorrow radioexam
[18:48] <jonsowman> good luck :)
[18:48] <timbobel> easy as $#!7
[18:49] <timbobel> do you need a licence in the UK to send 434mhz on 10mw
[18:49] <DanielRichman> no
[18:49] <timbobel> thought so
[18:49] <DanielRichman> most of us have them anyway
[18:49] <timbobel> same here.
[18:50] <timbobel> daniel, hows that nim2?
[18:52] <DanielRichman> nim2 no you don't need a license for that either
[18:52] <DanielRichman> fsphil|eee: well, I can replicate the bug :D
[18:52] <timbobel> yeah i got it
[18:53] <timbobel> but how is it working
[18:53] <timbobel> any good?
[18:53] <DanielRichman> oh right, I haven't started messing with the receive bit yet tbh
[18:53] <DanielRichman> I probably should
[18:53] <DanielRichman> jonsowman I believe used one doing OOK
[18:53] <DanielRichman> I'm going to try and do rtty >.>
[18:55] <jonsowman> yep it was OOK
[18:55] <Randomskk> see if you can find a way to do it that doesn't involve long periods of it being off waiting to see if it gets a signal
[18:56] <jonsowman> yeah that'd be good
[18:56] <jonsowman> that screwed things a bit
[18:56] <Randomskk> so blimey, what happened to rjh's kid?
[18:56] <jonsowman> though not aided by terrible radio thermal insulation
[18:56] <jonsowman> hmm yea
[18:56] <jonsowman> :\
[18:56] <Randomskk> I just get emails of the one line with my name in it and no context or following lines for a few minutes
[18:56] <Randomskk> so like
[18:56] <Randomskk> <soandso> Randomskk:
[18:57] <Randomskk> tempted to try turning the script into a buffer that records for x lines then emails
[18:57] <Randomskk> reply by email would be good too
[18:58] <DanielRichman> Randomskk: I was thinking of having a quite-short ~2s period to wait for a signal and having the transmitting station rig controlled so it can jump right in there
[18:58] <DanielRichman> automagically
[18:58] <Randomskk> sounds good
[18:58] <timbobel> who here has experience with receiving on hab ballons
[18:58] <jonsowman> that'd work
[18:59] <DanielRichman> fsphil|eee: now I can't reproduce the error. Something bad is happening
[18:59] <Randomskk> if the balloon's AGC is fast enough anyway
[19:00] <DanielRichman> fsphil|eee: the first time/try I compiled with -O2/default flags and laucnhed it and it just segv'd instantly
[19:00] <DanielRichman> other times it appears to work
[19:00] <DanielRichman> Randomskk: there's a pin that measures current signal amplitude
[19:00] Darkside (~Darkside@hpavc/darkside) left irc: Read error: Operation timed out
[19:00] <DanielRichman> s/measures /tells you /
[19:00] <Randomskk> rssi?
[19:00] <DanielRichman> yeah
[19:01] <jonsowman> so remember i emailed radiometrix about custom ntx2 frequencies
[19:01] <jonsowman> they can do them - MOQ 10 at £34 each
[19:01] <jonsowman> Randomskk: maybe it's worth considering 10 on a non-standard frequency for CUSF
[19:02] <jonsowman> they're likely to give some edu. discount too
[19:02] timbobel (~timboebl@5ED040EE.cable.ziggo.nl) left irc:
[19:02] <Randomskk> yea would be excellent to try
[19:03] <jonsowman> see your email
[19:03] Darkside (~Darkside@hpavc/darkside) joined #highaltitude.
[19:06] LazyLeopard (~irc-clien@chocky.demon.co.uk) joined #highaltitude.
[19:08] <fsphil> back
[19:08] <fsphil> DanielRichman, going to try on my desktop machine
[19:08] <DanielRichman> fsphil|eee: I'm suspicious of - 1sec -
[19:08] <DanielRichman> "warnign format %i expects int, but arg 3 is time_t"
[19:16] <fsphil> crashes without that line
[19:18] shauno (soneil@pdpc/supporter/professional/shauno) left #highaltitude.
[19:19] <fsphil> changing the %i to %li silences it here
[19:20] shauno (~soneil@pdpc/supporter/professional/shauno) joined #highaltitude.
[19:22] <fsphil> in saying that, the immediate crash in gdb is in vfprintf
[19:22] <DanielRichman> well the crash in vfprintf is what made me suspect that $i
[19:22] <DanielRichman> but it only happened once
[19:22] natrium (~natrium@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) joined #highaltitude.
[19:22] <fsphil> #4 0x00000000004f32dc in serial_thread (a=<value optimized out>)
[19:22] <fsphil> at misc/dl_fldigi_gps.cxx:136
[19:24] <DanielRichman> fsphil: SET_THREAD_ID(DL_FLDIGI_GPS_TID);
[19:24] <DanielRichman> ?
[19:24] <fsphil> yea
[19:24] natrium42 (~natrium@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) left irc: Ping timeout: 245 seconds
[19:24] <fsphil> I've followed it to: void linux_log_tid(void)
[19:24] <fsphil> in misc/threads.cxx
[19:24] <DanielRichman> it's ok fsphil; just comment it out
[19:25] <fsphil> instant crash
[19:25] <DanielRichman> if you comment it out you get an instant crash?
[19:26] <fsphil> yea, line 136
[19:29] <DanielRichman> I um... dont ;X
[19:29] <DanielRichman> I'll try with -O2
[19:30] <DanielRichman> All right, got it to break :D
[19:30] <fsphil> yay .. ish :)
[19:31] <fsphil> syscall(SYS_gettid) seems to be the cause
[19:33] jcoxon (~jcoxon@cpc1-lanc4-0-0-cust720.3-3.cable.virginmedia.com) joined #highaltitude.
[19:34] <jonsowman> hi jcoxon
[19:35] <jcoxon> internetz!
[19:35] <Randomskk> jcoxon: joooin uuuus
[19:35] <Randomskk> on skype
[19:35] <jonsowman> we have cookies
[19:35] <natrium> :D
[19:36] <fsphil> yay for internet
[19:39] <DanielRichman> fsphil: does it insta-crash every single type?
[19:39] <DanielRichman> even if you don't recompile or touch it
[19:39] <fsphil> hold on
[19:40] smeaway (~smealum@85-170-63-39.rev.numericable.fr) joined #highaltitude.
[19:40] smeaway (~smealum@85-170-63-39.rev.numericable.fr) left irc: Read error: Connection reset by peer
[19:41] <DanielRichman> ok fsphil I suspect a race condition. I'll try something
[19:41] <fsphil> hah, no its working now
[19:41] <fsphil> still crashing on conf change
[19:41] smeaway (~smealum@85-170-63-39.rev.numericable.fr) joined #highaltitude.
[19:41] <fsphil> the last message is always dl_fldigi: dl_fldigi_gps setup begin
[19:41] smeaway (~smealum@85-170-63-39.rev.numericable.fr) left irc: Read error: Connection reset by peer
[19:42] smeaway (~smealum@85-170-63-39.rev.numericable.fr) joined #highaltitude.
[19:43] smeaway (~smealum@85-170-63-39.rev.numericable.fr) left irc: Read error: Connection reset by peer
[19:43] smeaway (~smealum@85-170-63-39.rev.numericable.fr) joined #highaltitude.
[19:50] <fsphil> the thread never seems to leave pthread_cond_wait(&serial_info_cond, &serial_info_mutex);
[19:54] <DanielRichman> fsphil: http://github.com/danielrichman/dl-fldigi/commit/4ea63ef32fb617c87075d9646dd303fdbfd97df9 ?
[19:55] <DanielRichman> fsphil: unless it manages to connect to a gps then it will sit in pthread_cond_wait, waiting for you to change the settings
[19:55] jcoxon (~jcoxon@cpc1-lanc4-0-0-cust720.3-3.cable.virginmedia.com) left irc: Quit: Leaving
[19:56] <fsphil> does changing the settings not cause it to try opening the port?
[19:56] <DanielRichman> yeah it does
[19:56] <fsphil> crashes too
[19:56] <DanielRichman> :'(
[19:57] <DanielRichman> you can reliably get it to crash,right?
[19:57] <fsphil> it crashes everytime with gdb, as soon as it runs
[19:57] <fsphil> same with native, without gdb, but it only crashes once I've changed one of the text boxes in the GPS dialog
[19:58] Jasperw (~jasperw@skeleton2.london.iofc.org) joined #highaltitude.
[19:59] <DanielRichman> every time it runs it will receive a SIGUSR1 and gdb will break but not crash
[19:59] <DanielRichman> if you type continue; it should carry on
[19:59] <DanielRichman> *should*
[19:59] <fsphil> d'oh -- that's it
[20:00] <fsphil> gdb newbie
[20:00] <fsphil> now lets see
[20:01] <fsphil> hmm.. got it as far as the dialog crash but the backtrace isn't very inspiring
[20:02] RocketBoy (~steverand@217.47.75.8) joined #highaltitude.
[20:02] <DanielRichman> oh?
[20:02] <fsphil> http://pastebin.com/kfcSQb9w
[20:05] <DanielRichman> oh lovely
[20:05] <DanielRichman> is that with -g -O0
[20:05] <fsphil> whatever the default is
[20:06] <DanielRichman> could you try make CFLAGS="-g -O0"
[20:06] <DanielRichman> you'd have to make clean first
[20:06] <fsphil> yea, building now
[20:09] <fsphil> no better
[20:11] <DanielRichman> still no backtrace?
[20:11] <fsphil> nope, same as previous
[20:11] <fsphil> crash seems to be in libpthread itself
[20:11] <DanielRichman> eeh. I am confuse, sorry ;'(
[20:12] <DanielRichman> you could try installing libportaudio-dbg
[20:12] <DanielRichman> hmm.. could be some sort of overrun
[20:14] <DanielRichman> fsphil: you've tried this on two totally different boxes with the same results?
[20:15] <fsphil> yea
[20:15] <fsphil> one 32bit, the other 64
[20:15] <DanielRichman> ok, so that's just weird. Why on earth do we have different results?
[20:15] RocketBoy (~steverand@217.47.75.8) left irc: Read error: Connection reset by peer
[20:16] RocketBoy (~steverand@217.47.75.8) joined #highaltitude.
[20:18] <fsphil> oh man nildram are rubbish now that talk talk bought them out
[20:19] <fsphil> http://pastebin.com/Ywn6F5pg
[20:19] <fsphil> new BT, with added goodieness
[20:19] <fsphil> why it's got anything to do with portaudio, I've no idea
[20:21] <DanielRichman> haha got it to reproduce in gdb
[20:21] <DanielRichman> (gdb) handle SIGUSR1 nostop
[20:23] <fsphil> ah handy
[20:23] <fsphil> brb, I've to put up curtains. fun fun
[20:23] cant_copypaste_w (5610ae56@gateway/web/freenode/ip.86.16.174.86) joined #highaltitude.
[20:23] <cant_copypaste_w> http://pastebin.com/i8jh3gRt
[20:23] cant_copypaste_w (5610ae56@gateway/web/freenode/ip.86.16.174.86) left irc: Client Quit
[20:24] <DanielRichman> that was meant to say can't copypaste with my vm :X
[20:27] Laurenceb (~laurence@host86-137-224-95.range86-137.btcentralplus.com) joined #highaltitude.
[20:27] Simon-MPFH (~simon@phantom.mpfh.co.uk) left irc: Quit: Leaving
[20:28] <DanielRichman> maybe that bug is in fldigi and is induced by my enabling of fldigi-debug output
[20:32] <fsphil> tried it with debug disabled?
[20:39] <DanielRichman> pushed to testing
[20:39] <DanielRichman> You're the one that can make it crash :P
[20:39] <DanielRichman> http://github.com/danielrichman/dl-fldigi/commit/5d49b0a592e70913bd8a164546f77766792cec3d
[20:40] <DanielRichman> make CFLAGS="-g -O0" && gdb src/dl-fldigi
[20:40] <DanielRichman> handle SIGUSR1 nostop
[20:40] <DanielRichman> run
[20:42] <fsphil> still hanging curtains
[20:43] <fsphil> btw, I'm running it with the --hab option
[20:43] <DanielRichman> oh; I'm not ;P
[20:45] <DanielRichman> no, I still can't get it to crash
[20:46] <fsphil> crash, backtrace still landing me in pulseaudio libs now
[20:47] <fsphil> -still
[20:47] <DanielRichman> oh.. can't believe I didn't think of this earlier
[20:47] Action: DanielRichman installs valgrind
[20:47] <fsphil> muhaha
[20:47] <fsphil> brb
[20:47] <DanielRichman> Need to get 58.6MB of archives. After unpacking 175MB will be used.
[20:48] <DanielRichman> >.>
[20:48] <Randomskk> valgrind's great
[21:05] Laurenceb (~laurence@host86-137-224-95.range86-137.btcentralplus.com) left irc: Remote host closed the connection
[21:15] <fsphil> no more diy for me
[21:30] Jasperw (~jasperw@skeleton2.london.iofc.org) left irc: Quit: Leaving.
[21:36] <fsphil> valgrind is pretty verbose
[21:36] <DanielRichman> that's because fldigi is a mess (?)
[21:37] <DanielRichman> well, to be fair, it's most of the libraries that are causing the mess and not fldigi itself
[21:40] <DanielRichman> fsphil: at the moment we don't mind about the losing of pointers; I'm more interested in the Invalid writes
[21:40] <DanielRichman> if you find any, that is
[21:42] Nick change: natrium -> natrium42
[21:44] Hiena (~Hiena@81.93.195.181.datatrans.hu) left irc: Quit: -=Got bored from the net. Gone blowing up things.=-
[21:50] jcoxon (~jcoxon@cpc1-lanc4-0-0-cust720.3-3.cable.virginmedia.com) joined #highaltitude.
[21:51] <W0OTM> m running my own igate on 144.33. any reason why I can't beacon as often as I want
[21:51] <W0OTM> like excessively (every 5-10 sec)
[21:51] <W0OTM> aprs.fi is saying "This station is transmitting packets at a high rate, which can cause congestion in the APRS network."
[21:51] <W0OTM> but im not on the national aprs freq
[21:56] MrCraig (~Craig@cpc2-craw3-0-0-cust571.croy.cable.virginmedia.com) joined #highaltitude.
[21:57] gonzo_ (~gonzo_@cpc1-pool8-0-0-cust154.15-1.cable.virginmedia.com) left irc: Quit: (Connection reset by beer)
[21:59] BuffaloSouljah (~alberto@173-19-94-251.client.mchsi.com) left irc: Quit: Leaving.
[22:00] RocketBoy (~steverand@217.47.75.8) left irc: Quit: RocketBoy
[22:02] <fsphil> plenty of invalid reads, no writes
[22:03] Upu (~Upu@ubn.upuaut.net) left irc:
[22:05] <DanielRichman> fsphil: I see. ugh; I'm out of ideas
[22:06] <fsphil> I'll keep looking, it's likely something silly
[22:06] <DanielRichman> so basically you're reliably able to produce a crash, and the backtrace you get has got some crazy portaudio stuff in it?
[22:06] <fsphil> one backtrace ended up in portaudio, another in pulseaudio
[22:06] <fsphil> I suspect that bit's random
[22:07] <DanielRichman> ok. That sort of makes me think that it might be me corrupting the memory of something else and thus causing it to crash
[22:07] <fsphil> the only common point is in libpthreads
[22:07] <DanielRichman> but I don't know for sure. That could be total BS
[22:07] <DanielRichman> Well, everything calls libpthreads?
[22:07] <DanielRichman> fsphil: you can get it to crash every single time you run the program?
[22:07] <fsphil> yep
[22:07] <DanielRichman> fsphil: and it crashes as soon as you do something that may cause gps_setup() to be called?
[22:08] <fsphil> I'm totally unable to get past configuring the serial device
[22:08] <fsphil> yep
[22:08] <fsphil> the signal is fired, then it dies
[22:08] <DanielRichman> can you try commenting out the pthread_cond_signal or pthread_kill functions (one, the other, or both) and see if, or which, has any effect?
[22:08] <DanielRichman> tyvm
[22:10] <fsphil> commenting out pthread_cond_signal, no change
[22:10] jcoxon (~jcoxon@cpc1-lanc4-0-0-cust720.3-3.cable.virginmedia.com) left irc: Quit: Leaving
[22:10] <fsphil> ahh, no crash if I comment out the kill
[22:10] <DanielRichman> right. And you should be able to configure the serial stuff
[22:10] <fsphil> yep, set and saved
[22:11] <DanielRichman> the kill is there to interrupt blocking IO
[22:11] <DanielRichman> it should tell you that it's oopened something in the debug output
[22:11] <DanielRichman> (and if it's waiting for settings it's in pthread_cond_wait not any blocking io)
[22:11] <fsphil> it's repeatedly opening it, yea
[22:11] <DanielRichman> repeatedly? must be some error while trying to open it
[22:12] <DanielRichman> OK. So the question is why the signal causes spontaneous death, and only on your box
[22:12] <fsphil> oh wait, wrong serial device -- got three
[22:14] <fsphil> same -- no indication of an error
[22:15] <DanielRichman> can you pastebin the debug log please?
[22:17] <fsphil> http://pastebin.com/EywWUyLJ
[22:17] <fsphil> it's not getting data from the serial port
[22:17] <fsphil> looking at the init bits
[22:17] <DanielRichman> oh, right.
[22:17] <DanielRichman> dl-fldigi: fgetc gps: Success
[22:17] <DanielRichman> that's perror printing something when fgetc returned EOF
[22:17] <DanielRichman> I don't think fgetc sets errno :X
[22:18] <fsphil> setting up the serial port is a pain in linux
[22:18] RocketBoy (~steverand@217.47.75.8) joined #highaltitude.
[22:19] <DanielRichman> I'm going to replace perror("...fgetc") with
[22:19] <DanielRichman> fprintf(stderr, "dl-fldigi: fgetc returned EOF; feof() == %i, ferror() == %i", feof(f), ferror(f));
[22:20] RocketBoy (steverand@217.47.75.8) left #highaltitude.
[22:24] <DanielRichman> fsphil: could you uncomment the pthread_kill, strace -f dl-fldigi and pastebin the output (though, it might be quite large)
[22:24] <DanielRichman> oh, and don't forget to cause a massive crash.
[22:24] <fsphil> that I can do :)
[22:24] <fsphil> one sec, messing with the serial init
[22:25] <DanielRichman> If you're feeling up to it you could start a dl-fldigi process in strace, attach gdb, cause it to crash and produce a simultaneous backtrace and strace
[22:25] <DanielRichman> now that would be awesome ;)
[22:26] <DanielRichman> I think that (gdb) attach pid will do it
[22:26] <fsphil> yay, gps data being read in -- there's a but in the serial setup somewhere
[22:26] <DanielRichman> though you may need to load dl-fldigi and handler SIGUSR1 nostop handler SIGUSR2 nostop aswell
[22:26] <fsphil> I'll paste bin my code, but I'm setting a fixed baud rate -- you'd need to fix that
[22:26] <fsphil> but=bug
[22:26] <DanielRichman> Ah, now I can safely say that one is not mine :)
[22:26] <DanielRichman> RJH wrote that code >.>
[22:27] <fsphil> haha
[22:27] <DanielRichman> then again, I could have messed it up
[22:27] <DanielRichman> also, awesome hostname
[22:27] <fsphil> ta :)
[22:27] GeekShadow (~Antoine@reactos/tester/GeekShadow) left irc: Read error: Connection reset by peer
[22:27] <fsphil> chase_cat .. how'd I know that would be there :)
[22:28] <DanielRichman> i wrote LOL on the map but it's since dissappeared
[22:28] <fsphil> un-killing now
[22:29] <fsphil> lolcatz .. he knowz where you live
[22:29] <fsphil> okie, crash is back if I change config
[22:29] <fsphil> shouldn't I be on the map though?
[22:30] <fsphil> ack, wrong post url
[22:31] <fsphil> okie -- the configuration is right, and if I don't change it I get no crash -- but it doesn't seem to be doing anything
[22:35] <fsphil> strace output -> https://www.sanslogic.co.uk/files/out.gz
[22:37] <DanielRichman> excellent, thank you
[22:37] <DanielRichman> :o vim strace syntax hilighting
[22:37] <fsphil> yea. snazzy :)
[22:37] <fsphil> it even ungzips it
[22:40] alikins_ (~adrian@cpe-066-057-101-225.nc.res.rr.com) joined #highaltitude.
[22:42] MrCraig (~Craig@cpc2-craw3-0-0-cust571.croy.cable.virginmedia.com) left irc: Quit: Leaving
[22:48] timbobel (~timboebl@5ED040EE.cable.ziggo.nl) joined #highaltitude.
[22:48] <timbobel> i got that hellschreiber workin'!
[22:48] <fsphil> sweet!
[22:48] <fsphil> it's a great mode
[22:48] <fsphil> any vids?
[22:48] <timbobel> pretty basic. had to teach arduino all the capitals
[22:48] <timbobel> http://hollandshoogte.wordpress.com/
[22:49] <fsphil> nice and clear too
[22:49] <timbobel> i would like to amaze myself with some test with static
[22:49] <timbobel> i had to make the font myself :)
[22:49] <timbobel> Pain is my friend
[22:49] <timbobel> PainT
[22:50] <timbobel> but fldigi works with it quite nicely
[22:50] <timbobel> even with very low signal, it adjusts the contract automatically
[22:50] <DanielRichman> this string
[22:50] <DanielRichman> "Signal of Class <x> received. Exiting"
[22:50] <DanielRichman> I do not know what program that comes from
[22:50] <fsphil> the last ballasthalo flight had hell but I wasn't able to receive it
[22:50] <fsphil> DanielRichman, nor me - had a good grep for it
[22:51] <DanielRichman> damn, I was about to start grepping
[22:51] <DanielRichman> Binary file /usr/lib/libxmlrpc_server_abyss++.so.3.06 matches
[22:51] <fsphil> ooh
[22:51] <DanielRichman> however. I have no idea what that pos library is doing on *my* thread
[22:51] <timbobel> why couldnt you receive it?
[22:51] <DanielRichman> fsphil: does ./configure have a --disable-xmlrpc option?
[22:52] <fsphil> timbobel, it was pretty far away and I think a building was in the way. I could just about see it on the waterfall, but not enough to decode
[22:52] <DanielRichman> the xmlrpc lib has those ... wait... maybe xmlrpc uses SIGUSR1 !?
[22:52] <fsphil> !!
[22:52] <DanielRichman> if it does I'mma gonna rage
[22:52] <timbobel> how can you not-decode hellschreiber
[22:52] <fsphil> timbobel, the signal was too weak for me -- others decoded it fine
[22:53] <timbobel> so ok
[22:53] <timbobel> anyway
[22:53] <timbobel> tomorrow radio exam
[22:53] <timbobel> wish me a little luck
[22:53] <timbobel> i wont need much
[22:53] <fsphil> ah they're fun! good luck!
[22:53] <timbobel> X
[22:53] <DanielRichman> daniel@desktop:xmlrpc-c-1.06.27$ grep SIGUSR . -R
[22:53] <DanielRichman> ./src/cpp/server_abyss.cpp: sigaction(SIGUSR1, &mysigaction, NULL);
[22:53] <DanielRichman> ./lib/abyss/src/main.c: signal(SIGUSR1,sigterm);
[22:53] <DanielRichman> RARRRRRRRRGHHHHHHHHHH
[22:53] <timbobel> ht
[22:53] timbobel (~timboebl@5ED040EE.cable.ziggo.nl) left irc:
[22:53] <fsphil> SIGRAGE !!!
[22:53] <DanielRichman> cerr << "Signal of Class " << signalClass << " received. Exiting" << endl;
[22:53] <DanielRichman> what is this
[22:53] <fsphil> I take it you don't have xmlrpc compiled in?
[22:54] <DanielRichman> that's in the function
[22:54] <DanielRichman> sigterm(int const signalClass) {
[22:54] smeaway (~smealum@85-170-63-39.rev.numericable.fr) left irc: Ping timeout: 258 seconds
[22:54] <DanielRichman> I think I do fsphil
[22:54] <DanielRichman> let me check
[22:54] <fsphil> building without
[22:54] <DanielRichman> sigterm(int const signalClass) {
[22:54] Last message repeated 1 time(s).
[22:54] <DanielRichman> grr. that's like, the second time I've tried to cut and paste from my vm
[22:54] <DanielRichman> yes, I have it compiled in
[22:55] <fsphil> --without-xmlrpc does the trick
[22:56] <DanielRichman> allright. Well I downloaded the xmlrpc code so I'm now going to have great fun rm-Rfing it
[22:56] <DanielRichman> Using my damn signals.
[22:56] <DanielRichman> It doesn't even have a good use for it!
[22:56] <DanielRichman> /* These signals abort the program, with tracing */
[22:56] <DanielRichman> mysigaction.sa_handler = sigterm;
[22:56] <DanielRichman> sigaction(SIGTERM, &mysigaction, NULL);
[22:56] <DanielRichman> sigaction(SIGINT, &mysigaction, NULL);
[22:56] <DanielRichman> sigaction(SIGHUP, &mysigaction, NULL);
[22:56] <DanielRichman> sigaction(SIGUSR1, &mysigaction, NULL);
[22:57] <DanielRichman> also I'm sure fldigi uses a bunch of those signals
[22:57] <fsphil> HAH, no crash!
[22:57] <DanielRichman> High five!
[22:57] <DanielRichman> fsphil: I was right: it's a race condition
[22:57] <DanielRichman> whichever thread starts last gets it
[22:57] <DanielRichman> I believe the eventual SIGSEGV is caused by failing to fail
[22:58] <DanielRichman> I've found that before; when fldigi calls its diediedie function it prints a stacktrace and then segfaults
[22:58] <DanielRichman> interestingly the sigsegv intially happened
[22:58] <DanielRichman> [pid 6676] --- SIGSEGV (Segmentation fault) @ 0 (0) ---
[23:02] LA3QMA (~kgb@150.120.189.109.customer.cdi.no) left irc: Ping timeout: 265 seconds
[23:03] <fsphil> there we are, on the map
[23:03] <fsphil> had to tweak the fscanf line a bit
[23:03] <fsphil> I'll pastebin all my changes
[23:03] <DanielRichman> all right ;)
[23:03] <DanielRichman> fsphil: I'll make this thing use SIGUSR2
[23:04] <fsphil> use this: "GPGGA,%2u%2u%2u%*[^,],%2u%f,%c,%3u%f,%c,%*u,%u,%*f,%u,M,%*u,%*c,,*%*2x\n"
[23:04] <fsphil> should worth with and without floating point time
[23:04] <DanielRichman> cool
[23:05] <fsphil> and any fix value, you had it fixed to 1
[23:05] <fsphil> mine is 2 for some reason
[23:06] LA3QMA (~kgb@150.120.189.109.customer.cdi.no) joined #highaltitude.
[23:06] <DanielRichman> oh right. I have the definitions of those somewhere
[23:07] <fsphil> might be worth reading the value, and only uploading if it's > 0
[23:08] <fsphil> serial port setup -> http://pastebin.com/zCibDPV3
[23:08] <fsphil> only tested with my own -- and I'm not even sure what some of that does :p
[23:09] <DanielRichman> fsphil: just pushed some code to use SIGUSR2 -well.. I say pushed code, I pushed a commit that removes a load of stuff
[23:09] <DanielRichman> that xmlrpc stuff is just silly. Why do they need all those signals to do exactly the same thing - terminate the program?
[23:10] <fsphil> it is .. odd
[23:10] <DanielRichman> the default actions for those terminate it anyway and their handler doesn't do anything other than print a stupid message
[23:10] <fsphil> esp. the SIGUSR* signals .. lots of apps use those
[23:11] <DanielRichman> indeed. Where's the mailing list where I can write a poorly worded and rudly undertoned rant?
[23:11] LazyLeopard (~irc-clien@chocky.demon.co.uk) left irc: Quit: Bye
[23:14] <DanielRichman> fsphil: I'll start comparing that stuff with the code I've got etc.
[23:14] <DanielRichman> fsphil: but tyvm for your help debugging
[23:15] <DanielRichman> feels good to have finally found it. What a silly bug; I checked that fldigi wasn't using SIGUSR1 but didn't think for a second (duh) that another library would be using it >.>
[23:15] <DanielRichman> and the race cond. just made it worse
[23:15] <DanielRichman> I think I should disable fldigi debug again cause that turned up some strange bugs as well? I'm too tired to look into those as well :X
[23:15] <DanielRichman> I'll just merge that commit
[23:15] <fsphil> I knew it would be simple
[23:15] <DanielRichman> it always is ;'(
[23:16] <fsphil> the valgrind stuff I found was the same one repeated as it loaded the payload data
[23:16] <fsphil> simple and fixed
[23:16] <fsphil> a few others in fltk but not much we can do about those
[23:16] <DanielRichman> you fixed it? I was looking at the invalid reads but my fixes broke the parsing
[23:16] <DanielRichman> then again was trying to talk on skype at the same time :P
[23:16] <fsphil> lol
[23:16] <DanielRichman> did you just strdup and free?
[23:17] <fsphil> dropped the second xml->read
[23:17] <fsphil> uncommented the first one
[23:17] <fsphil> unless that's broke something I hadn't noticed
[23:18] <fsphil> nah they're all there
[23:19] <DanielRichman> I lost coding a couple of times
[23:19] <DanielRichman> I don't know what the bug is. Are you not allowed to keep xml->getNodeData or are you only not allowed to use it after calling xml->read?
[23:19] <DanielRichman> was that the only one that did that?
[23:19] <fsphil> the second xml->read is free()'ing the string
[23:20] <DanielRichman> ahh, and that's the only one that does that?
[23:20] <fsphil> yea
[23:20] <DanielRichman> oh, cool
[23:21] <fsphil> though a few payload don't come through with a coding
[23:23] <fsphil> ah, they don't have any in the xml anyway
[23:24] <fsphil> right, bed time. I'll do the dog walk test tomorrow :)
[23:24] fsphil (~phil@beastie.sanslogic.co.uk) left irc: Quit: zzzzzzzz
[23:38] fsphil|eee (~fsphil@2001:470:1f09:483:215:afff:fe9f:bf8e) left irc: Ping timeout: 276 seconds
[23:48] BuffaloSouljah (~alberto@2002:ad13:5efb:0:20c:76ff:fefb:b0e2) joined #highaltitude.
[23:59] BuffaloSouljah (~alberto@2002:ad13:5efb:0:20c:76ff:fefb:b0e2) left irc: Read error: Connection reset by peer
[00:00] --- Wed Sep 1 2010