[00:04] Upu (~Upu@ubn.upuaut.net) left irc:
[00:06] RocketBoy (steverand@ left #highaltitude.
[00:06] Nick change: SpeedEvil1 -> SpeedEvil
[00:08] LazyLeopard (~irc-clien@chocky.demon.co.uk) left irc: Quit: Bye
[00:31] MoALTz (~no@ left irc: Quit: Leaving
[00:35] SpeedEvil (~user@tor/regular/SpeedEvil) left irc: Ping timeout: 245 seconds
[01:10] SpeedEvil (~user@tor/regular/SpeedEvil) joined #highaltitude.
[01:23] SpeedEvil (~user@tor/regular/SpeedEvil) left irc: Ping timeout: 260 seconds
[01:33] SpeedEvil (1000@tor/regular/SpeedEvil) joined #highaltitude.
[01:45] natrium42 (~natrium@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) joined #highaltitude.
[02:27] DaveyC (~IceChat7@188-221-51-13.zone12.bethere.co.uk) left irc: Quit: Now if you will excuse me, I have a giant ball of oil to throw out my window
[02:33] juxta (juxta@ppp203-122-193-94.static.internode.on.net) joined #highaltitude.
[02:44] jedahan (~jedahan@ool-45701ce6.dyn.optonline.net) left irc: Ping timeout: 265 seconds
[02:54] natrium42 (~natrium@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) left irc: Quit: Leaving
[03:01] SpeedEvil (1000@tor/regular/SpeedEvil) left irc: Quit: Leaving.
[03:07] natrium42 (~natrium42@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) joined #highaltitude.
[03:15] SpeedEvil (~user@tor/regular/SpeedEvil) joined #highaltitude.
[03:30] SpeedEvil (~user@tor/regular/SpeedEvil) left irc: Quit: Leaving.
[03:30] SpeedEvil (1000@tor/regular/SpeedEvil) joined #highaltitude.
[05:03] GeekShadow (~Antoine@reactos/tester/GeekShadow) left irc: Quit: The cake is a lie !
[05:03] juxta_ (~blah@ joined #highaltitude.
[05:12] juxta_ (~blah@ left irc: Ping timeout: 240 seconds
[06:14] codetiger (7aa41f22@gateway/web/freenode/ip. joined #highaltitude.
[06:15] <codetiger> Hi everyone
[06:15] codetiger (7aa41f22@gateway/web/freenode/ip. left #highaltitude.
[06:15] codetiger (7aa41f22@gateway/web/freenode/ip. joined #highaltitude.
[06:19] codetiger (7aa41f22@gateway/web/freenode/ip. left irc: Client Quit
[07:55] natrium42 (~natrium42@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) left irc: Quit: My other car is a cdr.
[08:10] Upu (~Upu@ubn.upuaut.net) joined #highaltitude.
[08:47] akawaka (~akawaka@icculus.org) left irc: Ping timeout: 276 seconds
[08:48] akawaka (~akawaka@icculus.org) joined #highaltitude.
[08:49] Darkside (~Darkside@hpavc/darkside) joined #highaltitude.
[08:49] <Darkside> nwas anyone here at science alive today?
[08:52] ContraSF (email@ left irc: Quit: Time left until the Epochalypse: 27yrs 23wks 3days 1hr 21mins 21secs
[09:25] jcoxon (~jcoxon@host86-163-184-138.range86-163.btcentralplus.com) joined #highaltitude.
[09:26] <jcoxon> morning all
[09:42] juxta (juxta@ppp203-122-193-94.static.internode.on.net) left irc: Quit: Leaving
[09:44] juxta (juxta@ppp203-122-193-94.static.internode.on.net) joined #highaltitude.
[09:44] <juxta> morning jcoxon
[09:55] Darkside (~Darkside@hpavc/darkside) left irc: Ping timeout: 248 seconds
[10:07] <jcoxon> hey juxta
[10:12] <DanielRichman> Randomskk: I've pulled get_grib.py out of the git history for the predictor; it uses http://nomads.ncep.noaa.gov/cgi-bin/filter_gfs_hd.pl . It seems to have been left behind after a changeover to OpeNDAP, but there's no mention of why you changed to ndap
[10:15] <fsphil> morning
[10:20] <jcoxon> interesting: http://www.balloonmission.fr/Balloonmission/Balloonmission.html
[10:23] <jcoxon> fsphil, you should enter your altitude record into the ARHAB records
[10:23] <jcoxon> not altitude i mean distance
[10:24] <jcoxon> as its 345miles
[10:24] <jcoxon> as it would be 16th
[10:25] <fsphil> ah cool, didn't see that
[10:26] <jcoxon> also include that its 10mW
[10:26] <jcoxon> :-p
[10:27] <juxta> jcoxon, did you see the nmea thing m1x10 updated the FSA03 page with?
[10:27] <jcoxon> yeah
[10:27] <jcoxon> i'm quite excited about that
[10:27] <juxta> yeah
[10:27] <jcoxon> my breakout boards have been shipped
[10:27] <juxta> i'll put up a modified tinygps version that handles it shortly
[10:27] <jcoxon> oh nice
[10:27] <jcoxon> i think i'll move to that to tell the truth
[10:28] <jcoxon> just request the data instead of filling up the buffer
[10:28] <juxta> yeah me too
[10:28] <juxta> esp seeing as it returns straight away with whatever the latest data is
[10:28] <fsphil> will be handy not having to wait around
[10:28] <jcoxon> exactly
[10:29] <jcoxon> juxta, so the new lib are you just using the parsing abilities then?
[10:29] <juxta> yeah - will just add in the needed bits for the new string
[10:29] <juxta> the public methods will stay the same
[10:30] <juxta> though I'll add one for sats
[10:30] <jcoxon> great
[10:32] <juxta> jcoxon, that was really quick for your boards to be shipped by the way
[10:33] <juxta> what was the cost again?
[10:34] <jcoxon> $23
[10:34] <jcoxon> :-p
[10:34] <jcoxon> for 10
[10:34] <jcoxon> they quote 10-30days shipping though
[10:35] <fsphil> from china?
[10:35] <jcoxon> yeah seeedstudio
[10:35] <jcoxon> juxta, when they arrive i'll get some ntx2s and post them and a couple of boards to you
[10:35] <fsphil> shipping from there is strange .. I've had stuff arrive within a week, and other things take 2 months
[10:35] <jcoxon> fsphil, yeah
[10:35] <juxta> cheers jcoxon
[10:35] <juxta> how large are you boards?
[10:36] <juxta> i have some balloons on the way over to me from China
[10:36] <fsphil> jcoxon, are these different from the esawdust boards?
[10:36] <juxta> I think they left China yesterday, if all goes to plan they should be here in a few days
[10:37] <jcoxon> 41mm x 27mm
[10:37] <jcoxon> fsphil, yes
[10:37] <jcoxon> stuch the backup battery on the back and they are all thru hole
[10:37] <jcoxon> stuck*
[10:38] <jcoxon> so its smaller
[10:39] <DanielRichman> http://pastebin.com/wAvLD3GC
[10:40] <jcoxon> fsphil, http://www.flickr.com/photos/jcoxon77/4871446692/
[10:42] <fsphil> oh sweet
[11:14] timbobel (~timboebl@5ED040EE.cable.ziggo.nl) joined #highaltitude.
[11:14] <timbobel> verry good morning everyone
[11:18] <jcoxon> morning timbobel, congratulations for the launch last weekend
[11:18] <jcoxon> very cool
[11:20] <timbobel> jcoxon, thanks
[11:20] <timbobel> i have been getting some mediaattention, lol
[11:20] <jcoxon> hehe
[11:21] <juxta> hey timbobel
[11:21] <timbobel> i had >>11.000 visits on friday, which is too much for me. very nice comments though.
[11:21] <Upu> hey timbobel did it transmit its location when it was in the sea ?
[11:21] <timbobel> upu: yes but the antenna was getting shorted because of the water
[11:22] <Upu> still keep watching the video :)
[11:22] <DanielRichman> hey there Upu
[11:22] <Upu> you've given me a great idea for my launch :)
[11:22] <Upu> hi DanielRichman
[11:22] <timbobel> because in the sea it did get a GPS lock (not of the falcom since that one was dead). and it logged that to SD.. though the phone's data cable snapped loose so it didnt transmit an sms, otherwise it would have.
[11:23] <DanielRichman> Upu: do you have an account on that adaptec site? They want people to register their cards to download software etc.
[11:23] <DanielRichman> and I thought it'd be easier if you already had an account
[11:23] <timbobel> i actually think landing in the sea is ideal, but you have to have a boat though
[11:23] <Upu> doubt it we haven't used Adaptec for years
[11:23] <Upu> 1 sec I'll check
[11:23] <jcoxon> timbobel, yeah i understand what you mean
[11:23] <jcoxon> no worries about landing on anyone etc
[11:24] <jcoxon> but if you don't get to it quickly it might drift
[11:24] <DanielRichman> http://www.adaptec.com/en-US/speed/raid/aac/sm/asm_linux_x64_v4_30-16038_rpm.htm
[11:24] <timbobel> very true
[11:25] <Upu> 1 sec DanielRichman
[11:25] <Upu> stupid
[11:26] <Upu> I hate websites like this they want a serial number
[11:26] <Upu> which could cause an issue :)
[11:26] <DanielRichman> oh if you don't have an account don't worry
[11:28] <Upu> yeah sorry and I don't have the serial number what a stupid website
[11:28] <fsphil> if the water is calm it makes spotting it easier too
[11:28] <timbobel> who can give me points about silkscreen logo prints in eagle??
[11:28] <Upu> http://www.instructables.com/id/Adding-Custom-Graphics-to-EAGLE-PCB-Layouts/
[11:28] <timbobel> You are quick.
[11:28] <Upu> I know :)
[11:28] <fsphil> very quick lol
[11:28] <Upu> favourites :)
[11:30] <Upu> Next time i'm out I'll pop the lid off and get the serial number DanielRichman
[11:30] <Upu> ok afk now
[11:30] <DanielRichman> don't you dare go out there on our behalf :P
[11:30] <DanielRichman> but thank you :)
[11:30] <Upu> no don't worry I won't :)
[11:31] <jcoxon> no way of getting the device to report its serial number?
[11:44] GeekShadow (~Antoine@reactos/tester/GeekShadow) joined #highaltitude.
[11:47] <DanielRichman> jcoxon: looking into it
[11:48] <Randomskk> are you trying to install arcconf?
[11:48] <DanielRichman> I *could* download this software from an alternative sauce
[11:48] <Randomskk> arcconf can report device serial numbers
[11:48] <DanielRichman> yeah
[11:48] <Randomskk> :P
[11:48] <DanielRichman> :D
[11:48] <Randomskk> I have arcconf installed now, but it needs libstdc++5
[11:48] <Randomskk> and 10.04 only has 6
[11:48] <DanielRichman> damnit that's what I'm doing
[11:48] <DanielRichman> where did you get it from?
[11:48] <Randomskk> probably the same place you were about to get it from
[11:48] <Randomskk> hwraid
[11:48] <DanielRichman> hehe
[11:48] <DanielRichman> yeah I'll see if I can get stdc5 from lenny
[11:49] <Randomskk> http://www.hackourlives.com/ubuntu-10-04-lucid-lynx-libstdc-so-5/ looks promising
[11:49] <DanielRichman> Do you trust hwraid?
[11:49] <DanielRichman> http://packages.debian.org/lenny/amd64/libstdc++5/download ?
[11:49] <Randomskk> it works! http://mirrors.kernel.org/ubuntu/pool/universe/g/gcc-3.3/libstdc++5_3.3.6-17ubuntu1_amd64.deb actually
[11:50] <DanielRichman> arcconf runs?
[11:50] <Randomskk> yup!
[11:50] <Randomskk> $ sudo arcconf GETCONFIG 1
[11:51] <DanielRichman> oh you actually installed it >.>
[11:51] <Randomskk> Controller Status : Optimal
[11:51] <Randomskk> :D
[11:51] <Randomskk> gets us the serial # too
[11:51] <DanielRichman> allright, do you wanna sign up an adaptec account and get the real binary or shall I?
[11:51] <Randomskk> if you want to, go for it
[11:51] <DanielRichman> then we can uninstall these dodgey debs and put it under /opt/adaptec perhaps
[11:51] <Randomskk> dunno if needed
[11:51] <Randomskk> true
[11:53] <DanielRichman> ugh they want to know where I live
[11:53] <Randomskk> lie
[11:54] <DanielRichman> hmm they didn't notice that D0NT J0K isn't a valid postcode
[11:54] <jcoxon> the stratosphere!
[11:55] <Upu> lol
[11:55] <Upu> mail address bill@microsoft.com check all boxes for spam pls
[11:56] <jcoxon> steve@apple.com
[11:56] <Upu> haha
[11:56] <Upu> ok looks like I can afford a radio :) Does anyone know if the 817 comes with a usable whip or do I need to buy one ?
[11:56] <jcoxon> the whip is good
[11:56] <Upu> comes with one ?
[11:57] <DanielRichman> um it doesn't work
[11:57] <jcoxon> its a 2m/70cm and can also do 6m
[11:57] <jcoxon> Upu, yes
[11:57] <Randomskk> check what you're buying, if it's a new boxed one it comes with aiui
[11:57] <Randomskk> if it's second hand it may or may not but probably will
[11:57] <Randomskk> either way they should list it :P
[11:57] <Upu> Well I think the one from Radioworld is new
[11:57] <jcoxon> yeah it'll include a whip
[11:57] <Randomskk> I wanna get a second radio to leave on at home all day that I can rigcontrol
[11:57] <Upu> unless anyone has any preference for suppliers they see fairly cheap
[11:57] <fsphil> I used the supplied whip in the car, worked well
[11:57] <Randomskk> getting a vertical installed on the roof
[11:58] <Randomskk> wanna make a fully auto tracking station that'l just pick up anything that's flying
[11:58] <Upu> http://www.radioworld.co.uk/~radio/catalog/a430s10r-diamond-430440mhz-70cm-yagi-p-3962.html?osCsid=64f1baa3b074168a3500e80dac6c693a
[11:58] <Upu> will get one of those too for the top of the car
[11:58] <jcoxon> Randomskk, ouch on you electicity bill
[11:58] <Randomskk> hehe
[11:58] <Randomskk> really though? ft817 on receive with no sound can't use all that much power
[11:58] <jcoxon> Upu, i just got one of those
[11:58] <Upu> oh JackNorris is still working on a automated vehicle mounted tracker
[11:58] <jcoxon> Randomskk, its still a bit of a waste really
[11:59] <Randomskk> true
[11:59] <Randomskk> well thenr
[11:59] <fsphil> ft817 won't use much, computer might
[11:59] <Randomskk> rigcontrol and power control :P
[11:59] <Randomskk> depends what computer
[11:59] <jcoxon> i'm just challenging you to do power control ;-)
[11:59] <Upu> make a remote on/off :)
[11:59] <Randomskk> the main thing is something I can control entirely away from home, including turning it on/off
[11:59] <fsphil> an IP-enabled power switch
[11:59] <fsphil> oooh
[11:59] <Randomskk> yea, could make an IP enabled power switch
[11:59] <Randomskk> arduino, ethernet shield
[11:59] <Upu> easy arduino and a sheild
[11:59] <Upu> great minds...
[11:59] <Randomskk> :P
[12:00] <fsphil> Controlling a station at home, and one while chasing is a recipe for madness :)
[12:00] <Randomskk> probably better ways
[12:00] <Randomskk> wake on lan?
[12:00] <Randomskk> fsphil: not if the home one can be massively automated though
[12:00] <Randomskk> I want it to pretty much not need controlling
[12:00] <Upu> ok off to get some food bbl
[12:00] <DanielRichman> I think their website is broken.
[12:00] <fsphil> true -- the trick would be tracking the frequency drift
[12:00] <Randomskk> yea
[12:00] <jcoxon> just need to get some one to do the interface to dl-fldigi...
[12:00] <fsphil> haha
[12:00] <Randomskk> fldigi's AFC is okay
[12:00] <Randomskk> but doesn't do rigcontrol
[12:01] <jcoxon> Randomskk, thats actually not a challenge
[12:01] <fsphil> some kind of upper and lower limit, and have it nudge the radio a few hz
[12:01] <Randomskk> yea
[12:01] <Randomskk> just hack it a little to nudge the radio when it gets to the edges
[12:01] <jcoxon> we can control rigcontrol and dl-fldigi so
[12:01] <Randomskk> then you just need to find it to start with
[12:01] <jcoxon> can easily nudge the radio
[12:01] <Randomskk> tempted to have it parse emails to ukhas to find launch announcements
[12:01] <jcoxon> i think an online dl-fldigi interface will solve this
[12:01] <Randomskk> but probably best to have a simple web interface or something
[12:01] <jcoxon> which is on the list of jobs
[12:01] <Randomskk> I want it to work without me being there at all
[12:02] <Randomskk> like, if I'm away that day, or in lectures at launch, or something
[12:02] <Randomskk> manually scheduling start would be okay
[12:02] <jcoxon> Randomskk, you could always get someone else to kick it off for you
[12:02] <Randomskk> true
[12:02] <jcoxon> Randomskk, of course there is a xml-rpc lib for the android...
[12:02] <Randomskk> hehe
[12:03] <Randomskk> android client =D
[12:03] <jcoxon> slightly lower down the jobs list
[12:03] <Randomskk> this is all assuming a vertical on the roof will get a decent signal, but it seems it should
[12:03] <jcoxon> once i've finished moving will start get on with this
[12:03] <Randomskk> this is the same antenna fsphil used to get our record, though he has a fair bit more altitude
[12:04] <jcoxon> bbia
[12:04] <jcoxon> b
[12:04] <fsphil> Randomskk, I'm not sure yet how well it works when the balloon nearby
[12:04] <fsphil> most of it's gain is toward the horizon
[12:04] <Randomskk> balloons don't tend to get very nearby me though
[12:04] <Randomskk> me at home anyway
[12:05] <Randomskk> obviously they do get a bit closer at cambs, but this is for home
[12:05] <fsphil> should work well then
[12:07] <Randomskk> ground level is about 23m though
[12:07] <DanielRichman> Randomskk: I've uninstalled the 2 debs and put it in /opt to keep it out of the way
[12:07] <DanielRichman> sudo LD_LIBRARY_PATH=/opt/arcconf /opt/arcconf/arcconf GETCONFIG 1
[12:07] <fsphil> are there any large buildings nearby?
[12:07] <DanielRichman> I couldn't get adaptec's site to work so I've just used the binaries from hwraid
[12:08] <fsphil> the low angle of most of the balloons from here mean the buildings get in the way for some, with you being closer it shouldn't be as much of a problem
[12:08] <Randomskk> fsphil: nothing higher than my house as far as I can see
[12:08] <Randomskk> plenty the same height, but once I get onto the current antenna mast it gets above them
[12:09] <Randomskk> should be a mostly all clear view to the horizon from where it'l be going
[12:10] <Randomskk> not the best possibly position country-wise as london is in the way for cambridge and it's further south for the people launching up north (and for some reason there aren't many launches down south it seems)
[12:10] <DanielRichman> bbl
[12:10] <Randomskk> but hopefully not bad
[12:10] <Randomskk> (mainly I guess because APEX and ALIEN both launched from cambridge despite being from surrey, D:)
[12:10] <DanielRichman> skript: $ sudo /opt/arcconf/arcstatus
[12:10] <Randomskk> DanielRichman: cool
[12:11] <fsphil> that should change if rjh gets his permanent notam
[12:11] <Randomskk> isn't rjh also up north though?
[12:12] <fsphil> ah, I read that backwards
[12:12] <fsphil> from my point of view cambridge is very south :)
[12:14] <Randomskk> :P
[12:16] <fsphil> my only advice would be to keep the cable short, and use as low loss as possible
[12:16] <DanielRichman> hmm the latest version of adaptec storage manager doesn't require registration to download
[12:16] <Randomskk> fsphil: yea
[12:16] <DanielRichman> but it wasnt listed on the page for our card so I wonder if it works
[12:17] <Randomskk> I've got 5m or so of nick thick coax
[12:17] <Randomskk> which helpfully doesn't say what it's called
[12:17] <Randomskk> not sure if it'l be long enough though
[12:17] <Randomskk> hoping so
[12:18] SpikeUK (~chatzilla@host86-161-154-254.range86-161.btcentralplus.com) joined #highaltitude.
[12:20] jcoxon (~jcoxon@host86-163-184-138.range86-163.btcentralplus.com) left irc: Quit: This computer has gone to sleep
[12:20] jcoxon (~jcoxon@host86-163-184-138.range86-163.btcentralplus.com) joined #highaltitude.
[12:22] <Randomskk> it is either RG213 or URM67
[12:23] <Randomskk> not sure how to tell
[12:23] <Randomskk> both have 7/0.75ish inner conductors, 10.3mm outer diameter
[12:24] <Randomskk> and about 0.6dB/10m loss at 100MHz
[12:26] <DanielRichman> Randomskk: I can confirm that the hwraid version doesn't have any more warez than adaptec put in it
[12:27] <Randomskk> excellent
[12:27] <Randomskk> it did seem unlikely
[12:27] <Randomskk> :o it looks like libopenstm32 has working i²c
[12:27] <Randomskk> excellent
[12:28] <Randomskk> and ADC!
[12:28] <Randomskk> but not timer, not yert
[12:28] <Randomskk> or ethernet
[12:29] <DanielRichman> Randomskk: should we have monit watching the hw raid status?
[12:29] <Randomskk> if it can, I guess that's quite a good idea
[12:29] <Randomskk> if nothing else it can email us if anything goes wrong
[12:30] <DanielRichman> well if anything we could use the return code of grep:
[12:30] <DanielRichman> daniel@nessie:~$ sudo /opt/arcconf/arcstatus | grep "Status of logical device" | grep Optimal
[12:30] <Randomskk> true
[12:30] <DanielRichman> if that doesn't return 0 bad things are happening
[12:32] <DanielRichman> LD_LIBRARY_PATH=/opt/arcconf /opt/arcconf/arcconf-real GETCONFIG 1 | grep "Status of logical device" | grep Optimal > /dev/null || exit 1
[12:33] <DanielRichman> ( /opt/arcconf/monit )
[12:37] <DanielRichman> I don't think you can ask monit to exec a command periodically and act on the result...
[12:42] <DanielRichman> allright; how about executing this every day:
[12:42] <DanielRichman> LD_LIBRARY_PATH=/opt/arcconf /opt/arcconf/arcconf-real GETCONFIG 1 > /opt/arcconf/report
[12:42] <DanielRichman> grep "Status of logical device" /opt/arcconf/report | grep Optimal > /dev/null || mail root -s "arcconf report" < /opt/arcconf/report
[12:44] Laurenceb (~laurence@host86-152-35-235.range86-152.btcentralplus.com) joined #highaltitude.
[12:44] <Laurenceb> hi
[12:44] <SpeedEvil> hoi
[12:45] <DanielRichman> Randomskk: only remaining thing that probably should be dealt with is backups
[12:46] <Randomskk> yea
[12:46] <Randomskk> need to work out what to backup and where to I guess
[12:46] <DanielRichman> Filesystem Size Used Avail Use% Mounted on
[12:46] <DanielRichman> /dev/sda1 1.4T 1.6G 1.3T 1% /
[12:46] <DanielRichman> there's not much on disk atm. We *could* just do a file level backup of the whole thing
[12:47] <DanielRichman> depends if we want to worry about consistency etc.
[12:48] <Randomskk> a file level backup of everything might not be all that useful
[12:48] <Randomskk> depends what we want to guard against
[12:49] <DanielRichman> well providing we're doing it right, everything deployed to there will be on github
[12:49] <DanielRichman> so we want to backup application data, flight archives
[12:49] <Randomskk> yea
[12:50] <Randomskk> system config?
[12:50] <DanielRichman> encrypt and store /etc/ & /etc/.git somewhere?
[12:50] <Randomskk> could do
[12:51] <Randomskk> how do we want to do app data backups?
[12:51] <Randomskk> also where? Upu's server? S3?
[12:51] <DanielRichman> mysqldump, gzip, encrypt, upload?
[12:51] <DanielRichman> well depends again what you want to guard against
[12:51] <DanielRichman> if his server is in the next rack...
[12:51] <Randomskk> encrypt with what? pgp? what key? symmetric? encrypt to several of our public keys?
[12:51] <DanielRichman> encrypting to several of our pgp keys sounds good.
[12:52] <DanielRichman> it also depends on how much data we have
[12:52] <Randomskk> except it means old data becomes unusable if we all bugger off
[12:52] <Randomskk> depends how long old backups are kept for
[12:52] <DanielRichman> beacuse if it's small enough we could just email it to ourselves
[12:52] MoALTz (~no@ joined #highaltitude.
[12:52] <Randomskk> if we add a new admin, they'd only be able to decrypt from then on
[12:52] <Randomskk> maybe that's good too, and maybe backups won't be stored for long enough in time for it to be an issue
[12:52] <DanielRichman> well we want to keep some mega-old backups
[12:52] <DanielRichman> ie if we delete or corrupt something but dont notice for a month
[12:53] <Randomskk> do we want any of us individually to be able to restore the backups, or require at least two keys or such?
[12:53] <DanielRichman> well it's not exactly sensitive data
[12:53] <DanielRichman> all the flight archives are going to be publicly available anyway
[12:53] <DanielRichman> the config has some password hashes and ssh public keys in it, but apart from that...
[12:54] <DanielRichman> perhaps if we identify files like /etc/shadow, the mysql password database, /home/*/.ssh/ and put those in an encrypted tgz, and put everything else in a plain tar.gz?
[12:55] <DanielRichman> if we end up with user logins, would have to consider those
[12:55] <Randomskk> user login hashes shouldn't matter hugely but worth keeping as secure as possible I suppose
[12:55] <Randomskk> equally ssh public keys aren't as much of an issue but private ones are
[12:56] <DanielRichman> I would hope that everyones used fresh new passwords for nessie
[12:56] <Randomskk> anything in ~/ could potentially be considered sensitive though e.g. user generated private keys
[12:56] <Randomskk> I certainly have, but whether everyone has or not...
[12:57] <DanielRichman> well I'm not going to spike `sudo`, but on a server with so many people having root access it'd be crazy not to
[12:57] <DanielRichman> bbl, food
[12:57] <DanielRichman> *crazy not to choose a new p/w
[12:57] <Randomskk> yea exactly
[13:12] <timbobel> what are you guys working on
[13:12] DaveyC (~IceChat7@188-221-51-13.zone12.bethere.co.uk) joined #highaltitude.
[13:12] wonderchook (~wondercho@53.sub-75-197-252.myvzw.com) joined #highaltitude.
[13:30] wonderchook (~wondercho@53.sub-75-197-252.myvzw.com) left irc: Ping timeout: 245 seconds
[13:40] juxta|console (~juxta@ppp203-122-193-94.static.internode.on.net) joined #highaltitude.
[13:48] wonderchook (~wondercho@15.sub-75-197-95.myvzw.com) joined #highaltitude.
[13:55] wonderchook (~wondercho@15.sub-75-197-95.myvzw.com) left irc: Ping timeout: 240 seconds
[13:59] <timbobel> something secret, surely.
[14:00] <jcoxon> timbobel, they are working on combining the tracker/ predictor etc onto one server
[14:07] DaveyC (~IceChat7@188-221-51-13.zone12.bethere.co.uk) left irc: Ping timeout: 240 seconds
[14:08] wonderchook (~wondercho@108.sub-75-192-30.myvzw.com) joined #highaltitude.
[14:17] <DanielRichman> ping Randomskk
[14:19] jcoxon (~jcoxon@host86-163-184-138.range86-163.btcentralplus.com) left irc: Quit: This computer has gone to sleep
[14:37] Darkside (~Darkside@hpavc/darkside) joined #highaltitude.
[14:38] <Randomskk> DanielRichman: hi
[14:39] <DanielRichman> Randomskk: have a look in ~daniel/urls
[14:39] <DanielRichman> or run it if you want; it only takes about ~15s
[14:39] <Randomskk> is this grib vs gfs?
[14:40] <Randomskk> what am I looking at?
[14:40] wonderchook (~wondercho@108.sub-75-192-30.myvzw.com) left irc: Ping timeout: 240 seconds
[14:41] <Randomskk> (also, yay, I can get a vertical on the roof and have a radio about 2m cable away in the loft
[14:43] <DanielRichman> Randomskk: the point is that if you download all three components of the ndap data simultaneously, for a 5x5 latxlon square all alts one time, ndap is actually faster
[14:45] <Randomskk> so that would potentially be why we now use ndap instead of grib?
[14:45] <DanielRichman> that is indeed possible
[14:45] <Randomskk> do you see any advantage to grib over dap?
[14:46] <DanielRichman> grib seems to feature some sort of compression
[14:46] <DanielRichman> the data is much smaller for the same area
[14:46] <DanielRichman> and for some reason it depends on jpeg
[14:46] <Randomskk> more so than gzipping the dap data?
[14:46] <DanielRichman> well I get the impression that the grib is the master copy
[14:47] <DanielRichman> but yeah. I haven't looked to closely at that one
[14:48] <Randomskk> I see
[14:49] <Darkside> hmm
[14:49] <Darkside> was anyone here at science alive today?
[14:49] <Randomskk> not me
[14:51] <DanielRichman> Randomskk: in no case is bandwidth an issue here; http download seems to be capped at 500KB/s (if you want the whole file), perl cripples itself to about 4KB/s and opendap... is just doing something silly
[14:52] DaveyC (~IceChat7@188-221-51-13.zone12.bethere.co.uk) joined #highaltitude.
[14:52] <Randomskk> heh
[14:52] <DanielRichman> I think that I will probably divide the world up into 5x5 latlon chunks
[14:53] <DanielRichman> there isn't *that* much point caching data on disk since in ~6 hours it will be obsolete
[14:53] <DanielRichman> so I could cache in /dev/shm. Then if vmem gets low it can be swapped out
[14:56] <DanielRichman> It might also be possible to download an extremely low resolution dataset so the program knows roughly where the balloon is going, and can start downloading gfs in threads accordingly
[14:56] <Randomskk> can you download at low resolutions?
[14:57] <Randomskk> balloon flights don't really tend to go that far anyway though
[14:57] <DanielRichman> well the [a:b:c] syntax means give me points between a and c in steps of b
[14:57] <DanielRichman> [0:1:1] means 0,1 [0:1:3] 0,1,2,3 [0:2:3] 0,2
[14:58] <DanielRichman> afaik
[14:59] <Randomskk> other way around isn't it
[14:59] <Randomskk> [a:b:c] is a to b in steps of c
[14:59] <Randomskk> on python anyway
[14:59] <Randomskk> maybe they are annoying and do it differnetly
[14:59] DaveyC (IceChat7@188-221-51-13.zone12.bethere.co.uk) left #highaltitude.
[14:59] <Randomskk> >>> range(20)[1:10:2]
[14:59] <Randomskk> [1, 3, 5, 7, 9]
[15:00] <DanielRichman> http://nomads.ncep.noaa.gov:9090/dods/gfs_hd/gfs_hd20100807/gfs_hd_12z.asc?time[0:1:3]
[15:01] <DanielRichman> http://nomads.ncep.noaa.gov:9090/dods/gfs_hd/gfs_hd20100807/gfs_hd_12z.asc?time[0:2:3]
[15:01] <Randomskk> go them
[15:01] <Randomskk> how annoying
[15:02] spenguin[work] (~penguin@ left irc: Read error: Connection reset by peer
[15:02] spenguin[work] (~penguin@ joined #highaltitude.
[15:02] <DanielRichman> :P
[15:02] <DanielRichman> who's silly idea was it to write a data server in java
[15:02] <DanielRichman> *whose?
[15:02] <Randomskk> I know right
[15:02] <Randomskk> that said, I did not know you could access URLs like that and get ASCII data out of it
[15:03] <Randomskk> I wonder if there is anything besides GRIB and DAP
[15:03] <DanielRichman> http://nomads.ncep.noaa.gov/
[15:03] <DanielRichman> I think there are only the three options
[15:04] <Randomskk> http://en.wikipedia.org/wiki/Global_Forecast_System
[15:04] <DanielRichman> hmm where's the ECMWF's site?
[15:04] <DanielRichman> let's see what they do
[15:05] <Randomskk> "This is the only global model for which all output is available, for free in the public domain, over the internet" (GFS)
[15:06] <Randomskk> "(The IFS only has a limited amount of its output availiable for free.)"
[15:06] <DanielRichman> I go past the ECMWF on my way to school; I could pick it up
[15:06] <Randomskk> they give it out in hard copy?
[15:07] <DanielRichman> dunno, I've never tried
[15:07] <DanielRichman> "Information from the IFS is proprietary and copyrighted, though the ECMWF has made a limited amount of the model's most important calculations available to the public; this public data is, by declaration of the ECMWF, in the public domain."
[15:07] <DanielRichman> I can't make any sense of that sentence
[15:08] <Randomskk> it is a bit odd
[15:08] <Randomskk> if you go past them you could always pop in and be like "hi, we want to predict balloons, do you guys have wind velocity data we could use please?"
[15:09] <DanielRichman> I think Alex's dad works there
[15:10] <DanielRichman> All I can get out of their site is pikturez: http://www.ecmwf.int/products/forecasts/d/charts/medium/eps/ensm/essential/
[15:10] <Randomskk> haha his dad works there? what're the odds
[15:11] <DanielRichman> Alex keeps on going on about some massive computing cluster
[15:13] <DanielRichman> Anyway I think I'll - when I get around to it - try adding a dap/dods client to the predictor itself
[15:14] Laurenceb (~laurence@host86-152-35-235.range86-152.btcentralplus.com) left irc: Remote host closed the connection
[15:14] <DanielRichman> Randomskk: will `uprecords` tell you when you hit each milestone?
[15:14] <Randomskk> http://nomads.ncdc.noaa.gov/thredds/catalog/gfs-hi/201008/20100807/catalog.html
[15:15] <DanielRichman> nice historical data going back a loong way
[15:16] <Randomskk> they give near-real-time data apparently
[15:16] <Randomskk> and contain the archive
[15:16] <Randomskk> http://nomads.ncdc.noaa.gov/
[15:17] <DanielRichman> wait... what?
[15:17] <DanielRichman> so how is this different to ncep?
[15:17] <Randomskk> http://nomads.ncep.noaa.gov/txt_descriptions/servers.shtml
[15:19] <Randomskk> http://www.cpc.ncep.noaa.gov/products/wesley/fast_downloading_grib.html
[15:19] <Randomskk> Downloading meteorological data can be a pain. Servers are under-powered, connections are slow and the "bean counters" figure that 80 GB will store a trillion spreadsheets so who would want more disk space? Can't help with the last problem but downloading data in GRIB files can be made faster.
[15:20] <DanielRichman> yeah I saw that
[15:21] <DanielRichman> the grib files on ncep all have those header files
[15:22] <Randomskk> looks like DAP is still the best way though?
[15:22] <DanielRichman> that largely means that you can download the u component of wind, worldwide, for one pressure/altitude level in one go
[15:23] <DanielRichman> for example http://nomads.ncep.noaa.gov/pub/data/nccf/com/gens/prod/gefs.20100808/06/pgrb2/gep01.t06z.pgrb2f00.idx
[15:23] <DanielRichman> specfies byte offsets for UGRD VGRD at 10mb, 20mb etc
[15:24] <DanielRichman> I haven't yet tried doing that
[15:25] <Randomskk> what about the choice between GFS Ensemble HD, GFS Ensemble HD bias-corrected, GFS Ensemble NDGD bais-corrected, GFS 1x1 degree and GFS 0.5x0.5 degree?
[15:26] <DanielRichman> I have absolutely no idea
[15:26] <Randomskk> NDGD appears to just be lower res
[15:26] <Randomskk> ensemble appears to mean it uses some new tricks (feb2010) to reduce compound error
[15:26] <Randomskk> bias corrected appears to mean it also ties in with some of their other data sets to help improve accuracy
[15:27] <Randomskk> Bias corrected products are produced from the NCEP global ensemble and from the NAEFS ensemble, which includes the NCEP global ensemble in all cycles and the CMC global ensemble at 00Z and 12Z.
[15:27] <Randomskk> in other words I think we should be using the GFS Ensemble high-resolution Bias-corrected dataset
[15:27] <DanielRichman> :)
[15:27] <Randomskk> GENS rather than GFS
[15:28] <DanielRichman> all right
[15:28] <Randomskk> http://nomads.ncep.noaa.gov:9090/dods/gens_bc
[15:28] <Randomskk> that one
[15:28] <Randomskk> it looks like it is Better
[15:28] <Randomskk> though I am curious as to how different is it
[15:28] <DanielRichman> I'm going to see how fast using http part downloading to grab the data would be
[15:28] <SpeedEvil> Plus - at some point you don't care about better.
[15:29] <SpeedEvil> If your modeling is let down by the baloon ascent model and burst time forex
[15:29] <Randomskk> yea
[15:29] <timbobel> daniel can you check a small pcbsheet for me ;)
[15:29] <DanielRichman> what each one of these means, I have no idea http://nomads.ncep.noaa.gov/pub/data/nccf/com/gens/prod/gefs.20100808/06/
[15:30] Action: SpeedEvil can also have a look at PCBs.
[15:30] <SpeedEvil> I'm bored.
[15:30] <DanielRichman> timbobel: I dunno, but I'm definitly not the no1 authority on pcbs around here
[15:30] <SpeedEvil> Going through a list of 98 hits on my browser for my phone number.
[15:30] <Randomskk> just post the link
[15:30] <SpeedEvil> Saying 'take me of your DB, I am not an accounting firm'
[15:30] <timbobel> allright can pleaple check the NTX2 schematic, its for 3v3 as well as 5v; changeable with a switch
[15:30] <timbobel> http://img401.imageshack.us/img401/8894/checkplz.jpg
[15:31] <Randomskk> generally on schematics nets should be all at 0, 90, 180, 270 or 360 degrees
[15:31] <Randomskk> lots of resistor comining going on
[15:31] <timbobel> noted
[15:31] <Randomskk> you can usually just buy resistors for the value you want
[15:31] <DanielRichman> hey Randomskk look at this one http://nomads.ncep.noaa.gov/pub/data/nccf/com/gens/prod/gefs.20100808/06/pgrb2a_bc/gep01.t06z.pgrb2a_bcf00.idx
[15:31] juxta|console (~juxta@ppp203-122-193-94.static.internode.on.net) left irc: Quit: Leaving.
[15:32] <timbobel> randomskk: im using SMD's so not for this one..
[15:32] <DanielRichman> found that totally by chance
[15:32] <Randomskk> timbobel: hm?
[15:32] <timbobel> surfacemount resistors
[15:32] <Randomskk> DanielRichman: lol the number of random datsets they have
[15:32] <Randomskk> timbobel: yes, they are available in even more options than through hole
[15:32] <DanielRichman> Randomskk: but look; it's got all the data we need
[15:33] <DanielRichman> apart from TMP
[15:33] <DanielRichman> but that's like 2mb and bam you've got all the data for 00hours
[15:33] <timbobel> rly? couldnt find on farnell
[15:33] <Randomskk> timbobel: really?
[15:33] <timbobel> Really
[15:33] <DanielRichman> farnell has got so much crap on it I was struggling to choose resistors
[15:33] <Randomskk> DanielRichman: as in, download that entire dataset, rather than asking it for {u,v}grds?
[15:33] <Randomskk> timbobel: what value and what size smd?
[15:33] <DanielRichman> yeah because that's all that's in that dataset
[15:33] <Randomskk> DanielRichman: is that faster?
[15:34] <timbobel> R2512...
[15:34] <timbobel> values you can see in schematic
[15:34] <DanielRichman> well we don't have to seek around the file; in the other one UGRD for 10mb is at the start, then there's some other crazy shzzle, then UGRD for 20mb
[15:34] <DanielRichman> etc.
[15:34] <Randomskk> so e.g. 22k+2k2+2k2? you want exactly 26.4k?
[15:34] <DanielRichman> if we didn't fancy downloading the whole dataset we'd have to part download each UGRD table for each altitude
[15:34] <DanielRichman> (whole was 90mb)
[15:35] <Randomskk> isn't the whole point of the DAP thing that we just get what we ask for?
[15:35] <Randomskk> timbobel: the trick as well with resistors is that often what you need is relative
[15:35] <Randomskk> you really should never have to stack resistors to get the desired value when designing something you're buying parts for
[15:36] <DanielRichman> Randomskk: yeah but DAP is crazy slow. This is just a http server
[15:36] <Randomskk> ah, okay
[15:36] <DanielRichman> Randomskk: any idea what ENS=+2 means?
[15:36] <Randomskk> nope. ens is presumably ensemble
[15:37] <timbobel> randomskk: i know but this is best fit
[15:37] <Randomskk> http://nomads.ncep.noaa.gov/txt_descriptions/GFS_Ensemble_bias_corrected_doc.shtml http://nomads.ncep.noaa.gov/txt_descriptions/GFS_half_degree_doc.shtml is the comparison of the one I'm suggesting vs the one we used to use
[15:37] <Randomskk> timbobel: what values do you require?
[15:37] <Randomskk> seriously, you should not have to stack resistors
[15:38] <Randomskk> what size SMD are you doing? 0603? 0804?
[15:39] <Randomskk> basically your schematic is confusing: I don't understand what any of those four input pins are meant to be, what the resistors are trying to accomplish, or anything
[15:39] <Randomskk> the radiometrix is fairly obvious and I imagine the top connector is your antenna socket
[15:40] <Randomskk> I don't know what the bottom part is, and I'm guessing CONN has a vcc pin 1, GND pin 2, and then what, two pins for the two transmit frequencies of RTTY for 3 and 4?
[15:40] <Randomskk> so the enable pullup is fine
[15:41] <Randomskk> I still don't know what that part below the NTX2 is meant to be?
[15:41] <Randomskk> ah, some sort of mix in?
[15:42] <Randomskk> the two voltages from the two potential dividers connect through to the TX pin?
[15:42] <Randomskk> what is the actual thing?
[15:42] <Randomskk> there are much clearer ways to draw dividers that make it much more immediately obvious what you're trying to do
[15:43] <DanielRichman> http://nomads.ncep.noaa.gov/txt_descriptions/Help_Desk_doc.shtml "Ask a Science Question"
[15:43] DaveyC_ (~IceChat7@188-221-51-13.zone12.bethere.co.uk) joined #highaltitude.
[15:44] <DanielRichman> there are so many datasets in bazillions of subdirectories, I have absolutely no clue what each one does
[15:44] <DanielRichman> or whether it is worthy of use
[15:45] <DanielRichman> I wonder if they'd respond if I asked them which was the best one to use
[15:45] <Randomskk> probably
[15:45] <Randomskk> explain what we do, give a link to the current predictor, say "so.... what dataset?"
[15:46] <DanielRichman> do you think it's worth giving them a technical overview and asking them to consider how fast we can downlnoad the data when asking?
[15:46] <DanielRichman> *suggesting a set
[15:46] <Randomskk> no hurt
[15:46] <DanielRichman> allright. I'mma compose an email
[15:48] <Randomskk> timbobel: what is the 3/6 pin part at the bottom of your schematic?
[15:49] <griffonbot> @nearsys: Pluging away on the presnetation for my consulting job. I get to visit DC for it. #ARHAB [http://twitter.com/nearsys/status/20633302258]
[15:50] <Randomskk> DanielRichman: how did you generate the two voltages for the NTX2?
[15:50] <Randomskk> I'm not sure what timbobel is trying to do electrically
[15:51] <DanielRichman> what I did I wouldn't do again
[15:51] <Randomskk> I'd have thought two potential dividers would suffice but there is something more complicated going on here
[15:51] <Randomskk> ah
[15:51] <Randomskk> do two potential dividers not suffice?
[15:51] <DanielRichman> I'll go grab the schematic of what we did
[15:51] <DanielRichman> Randomskk: three resistors woudl suffice
[15:51] <DanielRichman> Randomskk: I investigated it, afterwards and I reckon I can do it, 5V, with 3 e12 resistors
[15:51] <Randomskk> why do we need two pins from the µC?
[15:51] <DanielRichman> we don't
[15:51] <Randomskk> I'm wondering why not one pin, given as it has a low and a high output level
[15:52] <Randomskk> not even PWM, but one pin and some resistors
[15:52] <DanielRichman> exactly. I had a design to use one pin a while ago but I lost it :(
[15:53] <DanielRichman> http://alienproject.files.wordpress.com/2009/06/radiometrix-circuit-diagram.png
[15:54] <Randomskk> that's what you wouldn't do again?
[15:54] <DanielRichman> yeah
[15:54] <Randomskk> why?
[15:55] <Randomskk> (by the way, for repeated pins in eagle you can use @ syntax to name them, e.g. RFGND@1, RFGND@2, both show up on schematic as RFGND)
[15:55] <DanielRichman> yeah that's what I did on the alien 2 schematic
[16:08] <Randomskk> DanielRichman: does sending email to root@habhub.org actually work
[16:08] <DanielRichman> Randomskk: I think I set it up
[16:08] <DanielRichman> on gapps
[16:09] <Randomskk> ah okay, cool
[16:09] <DanielRichman> haven't tried it yet though
[16:09] <Randomskk> hah
[16:09] <Randomskk> the HD in gfs_hd is not high definiton
[16:09] <Randomskk> it is half degree
[16:10] DaveyC_ (IceChat7@188-221-51-13.zone12.bethere.co.uk) left #highaltitude.
[16:13] <Randomskk> bloody exim
[16:13] <DanielRichman> ?
[16:13] <DanielRichman> oh right
[16:26] <Upu> we have Adaptec Software very nice
[16:35] Nick change: SpeedEvil -> Termana-
[16:35] Nick change: Termana- -> SpeedEvil
[16:42] <timbobel> randomssk: the 3pinx2 thing is a switch with 2 poles
[16:43] <Randomskk> oh, okay.
[16:43] <Randomskk> your 5v/3v3 select?
[16:43] <DanielRichman> Randomskk: does this cover everything? http://pastebin.com/0WKqHUFW
[16:43] <timbobel> right.
[16:44] <Randomskk> DanielRichman: perhaps a sentence preamble who you are and what you do
[16:44] <DanielRichman> all right; but apart from that? does the question sort of make sense
[16:45] <Randomskk> yea, looks good
[16:56] <DanielRichman> looking at that schematic... I don't think a dpdt switch will do what you want there
[16:56] <DanielRichman> because the 3v3 resistors will interfere with the 5v rtty even if only 5v is selected
[16:57] <DanielRichman> there will be ~50k extra randomly between the mark and space nets
[16:57] <Randomskk> aren't the voltages all ratiometric anyway?
[16:57] <Randomskk> I mean, evidently not I guess
[17:00] GeekShad0w (~Antoine@ joined #highaltitude.
[17:01] <Randomskk> hmm
[17:01] <Randomskk> bootloading my ARM over ethernet
[17:01] <Randomskk> what an exciting prospect
[17:02] <Randomskk> so long as you don't brick it that could be excellent
[17:03] GeekShadow (~Antoine@reactos/tester/GeekShadow) left irc: Ping timeout: 260 seconds
[17:06] GeekSh4dow (~Antoine@185.52.199-77.rev.gaoland.net) joined #highaltitude.
[17:08] GeekShad0w (~Antoine@ left irc: Ping timeout: 240 seconds
[17:13] juxta (juxta@ppp203-122-193-94.static.internode.on.net) left irc: Ping timeout: 240 seconds
[17:13] <DanielRichman> how about bootloading it over nim2
[17:14] <Randomskk> it doesn't have a radio on it, normally
[17:14] <Randomskk> this being the weather station rather than a balloon
[17:14] <Randomskk> otoh bootloading a balloon payload would be even more exciting
[17:14] <DanielRichman> well ideally you'd be able to checksum and correct any mistakes before it restarts
[17:14] <Randomskk> yes :P
[17:15] <DanielRichman> but with limited flash...
[17:15] <DanielRichman> well ideally ideally you'd be able to upload it, check it, and if it times out and you don't manage to upload a proper firmware then it will revert
[17:15] <Randomskk> with two way comms you could checksum each packet and retry
[17:16] <DanielRichman> yup
[17:18] timbobel (~timboebl@5ED040EE.cable.ziggo.nl) left irc: Ping timeout: 240 seconds
[17:18] <DanielRichman> perhaps it'd be safest to make the bootloader massive
[17:19] <DanielRichman> ie. the bootloader will grab data from the gps, report its poistion etc.
[17:19] <DanielRichman> infact just program your mcu as you *would* and stick that in the bootloader section
[17:19] <DanielRichman> then have the ability to execute arbitrary uplinked stuff as a feature
[17:20] <Randomskk> uh
[17:20] <Randomskk> oh, right, because normal app code can't write to the flash and execute it
[17:20] <Randomskk> yes
[17:21] <DanielRichman> balloon OS.
[17:21] <Randomskk> :D
[17:22] <DanielRichman> btw on the stm32 chips, can you use spare dataflash - ie. stuff not used for the executable - as nonvolatile memory, a bit like the way the atmega chips all came with eeproms?
[17:23] <Randomskk> think so, not entirely sure
[17:24] <Randomskk> eurgh, brutal. ST release an app note/sample project for ethernet
[17:24] <Randomskk> including lwIP with relevent STM32 ports
[17:24] <Randomskk> but it's targetted at their example hardware board
[17:24] <Randomskk> and doesn't include makefiles
[17:25] <Randomskk> total and utter pain in the arse to try and build and make anything run
[17:25] <Randomskk> but writing it myself would be a humongous effort
[17:25] <DanielRichman> :(
[17:25] <Randomskk> their ethernet driver code alone is thousands of lines
[17:25] <Randomskk> god, ethernet is complicated
[17:26] <Randomskk> but it's possible it's because it has so many comments and can be configured to work in a number of ways as a demo
[17:26] <Randomskk> thing is I can't get anything working until I port it to my hardware
[17:26] <Randomskk> almost enough to make me wish I'd bought their board
[17:32] <DanielRichman> :X
[17:32] <Randomskk> not for the actual weather station, but to play with
[17:32] <Randomskk> but, no, allow that.
[17:32] <Randomskk> the thing that really starts to get annoying though is that I want to use libopenstm32
[17:32] <Randomskk> it's much much nicer than ST's library
[17:32] <Randomskk> but it's not complete and doesn't support the MAC yet
[17:33] <Randomskk> so either I use both, or just ST's
[17:33] <Randomskk> or neither and just write to registers directly but that starts to get infeasible quickly
[17:33] <DanielRichman> registers; what, like on the avr?
[17:33] <DanielRichman> that's not >too< bad
[17:33] <Randomskk> yea. that's how it all works deep down
[17:33] <Randomskk> it's not too bad on the AVR
[17:33] <Randomskk> on the ARM there are a lot more registers
[17:33] <Randomskk> even for basic things
[17:33] <Randomskk> GPIOs are much more complicated, for instance
[17:34] <Randomskk> it's not just DDR, PORT and PIN
[17:34] <DanielRichman> they've done something interesting for that on the xmega
[17:34] <Randomskk> you have things like what speed you get, whether it's pulled up or down, open drain vs push pull vs alt function outputs, floating or analogue inputs, other things
[17:34] <Randomskk> for other periphs it's far more complicated
[17:34] <Randomskk> for ethernet it's crazy
[17:34] <DanielRichman> they've created a PORT_T struct with IN OUT PIN1DIR (I think)
[17:35] <DanielRichman> then they just map the struct onto the first port register along with a volatile keyword
[17:35] <DanielRichman> and you can PORTD.OUT |= 0x02;
[17:35] <Randomskk> not bad
[17:35] <Randomskk> the ST library involves a load of longly named functions that do all the register writing for you
[17:36] <DanielRichman> hehe.
[17:36] <DanielRichman> I suppose you've also got more clock cycles to spare
[17:36] <DanielRichman> and ram
[17:36] <Randomskk> http://github.com/esden/libopencm3/blob/master/examples/stm32-h103/miniblink/miniblink.c
[17:37] <Randomskk> that's an example of registers vs using libopenstm32
[17:37] <Randomskk> (commented code is registers)
[17:37] <Randomskk> http://github.com/randomskk/STM32_SkeletonProject/blob/master/main.c uses the ST library
[17:38] <DanielRichman> I C
[17:38] <DanielRichman> the whole build-a-struct-with-your-config then pass it to a function seems a bit wasteful
[17:38] <Randomskk> yea.
[17:39] <Randomskk> also libopenstm32 comes with linker files which is always one of the more arcane bits of doing all this for me
[17:39] <DanielRichman> does hte library just then memcpy it into the register?
[17:39] <Randomskk> maybe, maybe not
[17:39] <Randomskk> I'd hope so I guess
[17:39] <Randomskk> at any rate libopenstm32 is a nicer interface to writing to the registers
[17:40] <Randomskk> I mean, I could just write to the registers, too, it's an option
[17:40] <Randomskk> so I'm not sure what to do for this
[17:40] <DanielRichman> Is the ST library not "open"?
[17:40] <Randomskk> libopenstm32 isn't complete enough to use for everything
[17:40] <Randomskk> the ST library is half open, but mainly it's just a bit annoying
[17:40] <DanielRichman> ha
[17:40] <Randomskk> I think the issue is that it's not a community project at all and it's kinda very verbose and crap
[17:40] <Randomskk> but it's not that awful
[17:40] <Randomskk> and does at least cover all functionality
[17:41] <Randomskk> so maybe I'll just use that, certainly until libopenstm32 gets there?
[17:41] <Randomskk> I just want to get this ethernet crap working, :(
[17:41] <SpeedEvil> umm. I think you've forgotten something.
[17:41] <SpeedEvil> Unless you've got a 40km ethernet cable.
[17:41] <Randomskk> as it happens
[17:43] KingJ|Alt (~k@ joined #highaltitude.
[17:43] KingJ|Alt (~k@ left irc: Client Quit
[17:45] <DanielRichman> if the predictor is entirely written in C... then all you need to do is add one library and you've got a predictor-fcgi server
[17:45] <DanielRichman> you know you want to
[17:45] <Randomskk> ...:D
[17:46] <DanielRichman> then the predictor is just html, js, fcgi server
[17:46] <Randomskk> well
[17:47] <DanielRichman> if you embedded the js in the html and compiled it into .data... no... that'd just be silly :P
[17:47] <Randomskk> there is already a version that's just html, js, a python .cgi that runs the C predictor
[17:47] <DanielRichman> yeah I think I found that while looking around
[17:47] <Randomskk> besides which we want our pretty unified interface
[17:48] <Randomskk> otoh having the predictor be accessed via http might be interesting
[17:48] <Randomskk> a restful predictor
[17:48] <Randomskk> just hit up a url with posted parameters, get back a track
[17:49] <DanielRichman> well at this rate, if the tracker is just a couchdb server and the predictor is just a C binary, then the whole thing will have no php or python in it
[17:56] <Randomskk> that isn't always bad
[17:56] <Randomskk> to be honest, quite amusing
[17:56] <Randomskk> I suspect we'll have some glue logic in there somewhere though
[17:58] <Randomskk> hmm. there is alternatively a version of uIP for stm32, also ported by ST
[17:58] <Randomskk> uIP is smaller, more integrated, less featureful
[17:58] <Randomskk> can run on 8bit platforms and has been ported to AVR, too.
[17:58] <Randomskk> includes an http server though
[17:59] <Randomskk> though that isn't really any use, I want http client
[18:01] <DanielRichman> There's an AVR project that if you plug it into a pc it enumerates as a usb hdd; you put files on it. Then if you plug it into a USB ASDL (Ithink) modem it will become the host and start http-serving the data
[18:02] m1x10 (m1x10@ppp089210151148.dsl.hol.gr) joined #highaltitude.
[18:02] <Randomskk> neat.
[18:03] <Randomskk> hah
[18:03] <Randomskk> the same guy wrote uIP and lwIP
[18:03] <Randomskk> he says uIP fits into 4-5kb code and a few hundred bytes of ram, lwIP is a bit larger but has more functionality
[18:04] <Randomskk> he also write twIP, or tweetIP, which is an IP stack that fits into a twitter message - 128 bytes of source code and answers IPv4 pings
[18:04] <DanielRichman> :D
[18:04] <Randomskk> also a VNC server for 8bit micros
[18:04] <Randomskk> and a web server and TCP/IP stack in 30 bytes of ram... blimey
[18:04] <Randomskk> http://www.sics.se/~adam/software.html
[18:04] <Randomskk> https://twitter.com/adunk/status/3660934560
[18:05] <DanielRichman> So when the STM32 sheet says that it has an "Ethernet peripheral" what is actually included, and what do you have to do in software?
[18:05] <Randomskk> it includes the MAC, which is what is usualy included in devices saying they have ethernet
[18:05] <Randomskk> the MAC is the medium-independent controller for ethernet
[18:05] <SpeedEvil> the phy isthe physical layer
[18:06] <Randomskk> in hardware, you have to provide a PHY, which converts data from the MAC onto your medium (typically an ethernet cable)
[18:06] <Randomskk> (but could be ethernet over fibre optics)
[18:06] <Randomskk> (or anything else)
[18:06] <SpeedEvil> which may need transformers and stuff
[18:06] <Randomskk> so the ARM has the MAC, but connects to a PHY onboard, which connects to an RJ45 socket typically
[18:06] <Randomskk> the MAC will control the PHY and send it ethernet frames
[18:06] <Randomskk> many MACs have a few extra features, e.g. they will calculate checksums for you
[18:07] <Randomskk> in hardware, so fast and saving cpu time
[18:07] <Randomskk> it does not provide TCP/IP, so you have to do everything from that layer up, which includes reassembling packets in order, etc
[18:07] <Randomskk> but the STM32 MAC does calculate TCP and IP checksums
[18:08] <SpeedEvil> that's unusual
[18:08] <Randomskk> yea
[18:08] <Randomskk> provided as a utility feature to help keep cpu usage down etc
[18:08] <SpeedEvil> at least at the low end
[18:09] <Randomskk> however there are TCP/IP software stacks out there that work
[18:09] <Randomskk> namely uIP and lwIP
[18:09] <Randomskk> which provide the TCP/IP software stuff and often some app level stuff - e.g. am http server, a telnet client, dhcp, dns etc
[18:09] <Randomskk> but you need to write the code to connect the TCP/IP stack to your ethernet hardware
[18:10] <Randomskk> ST have done that for uIP and lwIP
[18:10] <Randomskk> uIP is device independent C, expects a small buffer that you place an incoming ethernet packet into, then it processes it, or it makes an ethernet packet, then you send it
[18:11] <DanielRichman> I'm not sure if this guy is awesome or crazy
[18:11] <Randomskk> haha yea
[18:11] <Randomskk> he likes his tcp/ip
[18:11] <Randomskk> I think I am going to use the ST library then
[18:11] <Randomskk> now the question is lwIP or uIP
[18:11] <Randomskk> to be honest, this is powerful enough for lwIP
[18:13] <SpeedEvil> stm32 is stupidly powerful.
[18:13] <SpeedEvil> Compared to only a few years ago.
[18:13] <Randomskk> yea, I know, it's awesome
[18:13] <SpeedEvil> I have sketches of doing GPS in software.
[18:13] <SpeedEvil> Though I haven't got far enough witht hat
[18:13] <Randomskk> though that said lwIP needs ~40kB code
[18:13] <Randomskk> it has 128kB flash
[18:14] <Randomskk> well, mine does
[18:14] <Randomskk> so that is starting to use significant amounts
[18:14] <Randomskk> but to be honest, not unaffordable amounts
[18:16] <Randomskk> on the other hand, if uIP does everything I need....
[18:16] SpeedEvil (1000@tor/regular/SpeedEvil) left irc: Quit: Leaving.
[18:17] SpeedEvil (~user@tor/regular/SpeedEvil) joined #highaltitude.
[18:19] <Randomskk> thankfully it looks like ST's demo board uses the same PHY I do
[18:19] <Randomskk> maybe there just aren't many PHYs. it was the only model that came up on farnell
[18:20] <Randomskk> looks like mbed uses the same, too
[18:23] SpeedEvil (~user@tor/regular/SpeedEvil) left irc: Quit: Leaving.
[18:24] SpeedEvil (1000@mauve.plus.com) joined #highaltitude.
[18:24] SpeedEvil (1000@mauve.plus.com) left irc: Changing host
[18:24] SpeedEvil (1000@tor/regular/SpeedEvil) joined #highaltitude.
[18:34] natrium42 (~natrium@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) joined #highaltitude.
[18:42] earthshine (~earthshin@cpc3-orpi1-0-0-cust867.bmly.cable.ntl.com) left irc: Read error: Connection reset by peer
[18:44] earthshine (~earthshin@cpc3-orpi1-0-0-cust867.bmly.cable.ntl.com) joined #highaltitude.
[18:45] <DanielRichman> hey there natrium42
[18:45] <natrium42> hi
[18:45] <DanielRichman> I believe you don't yet have an account on nessie.habhub.org
[18:46] <natrium42> not that i know
[18:46] <DanielRichman> got an ssh key to hand?
[18:47] <DanielRichman> and what username would you like?
[18:48] <natrium42> natrium is fine
[18:50] <DanielRichman> ok natrium42 I'll need your ssh public key
[18:50] <natrium42> i usually use passwords
[18:51] <DanielRichman> we've kinda disabled password authentication
[18:51] <natrium42> ...
[18:51] <DanielRichman> +secure
[18:51] <DanielRichman> passwords are so last century
[18:53] <natrium42> pastie.org down, grr
[18:53] <natrium42> http://pastebin.com/aRw6zvQd
[18:54] <natrium42> bbl
[18:55] natrium42 (~natrium@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) left irc: Quit: Leaving
[19:03] Jasperw (~jasperw@212-98-44-230.static.adslpremium.ch) left irc: Ping timeout: 258 seconds
[19:03] LazyLeopard (~irc-clien@chocky.demon.co.uk) joined #highaltitude.
[19:07] Jasperw (~jasperw@212-98-44-230.static.adslpremium.ch) joined #highaltitude.
[19:21] <Upu> i remember passwords, SSH keys are so much more preferable stops me posting my password once a week to IRC
[19:21] <Randomskk> :P
[19:22] <Randomskk> /msg nickserv identify randomskk1!
[19:22] <Randomskk> amirite
[19:22] <Upu> yeah that's the one
[19:22] <Upu> dual monitors doesn't help
[19:22] <Randomskk> :P
[19:27] <SpeedEvil> Upu: letter substitution in the client is good for that
[19:27] <Randomskk> or auto identification on connect to be honest
[19:27] <SpeedEvil> skk -> 5ap
[19:28] <SpeedEvil> that too
[19:28] <SpeedEvil> but I've accidentally typed most secret pw into IRC before
[19:28] <SpeedEvil> when an xterm died
[19:28] <Randomskk> ouch
[19:28] <Randomskk> that's unfortunate
[19:29] <SpeedEvil> unplug incoming net, say bugger, repassword.
[19:30] <Randomskk> hmm. I wonder if it's worth running an OS on this
[19:30] <Randomskk> freertos works on cm3
[19:54] <Upu> Noob question which patch lead do I need to plug this Yagi into a 817 ?
[19:54] <Upu> http://www.radioworld.co.uk/~radio/catalog/a430s10r-diamond-430440mhz-70cm-yagi-p-3962.html?osCsid=64f1baa3b074168a3500e80dac6c693a
[19:54] <Upu> there are a few suggestions at the bottom
[19:55] <LA3QMA1> ft817 can use both BNC and PL-259
[19:55] <Randomskk> that yagi is pl-259
[19:56] <Randomskk> so just pl-259 to pl-259 patch
[19:56] <Randomskk> get a few meters of coax, two pl-259 connectors and grab your soldering iron to be honest
[19:56] <LA3QMA1> Randomskk: it says SO-239
[19:56] <Randomskk> SO-239 is the female socket for PL-259 cables
[19:56] <LA3QMA1> so two PL259 and your done
[19:57] <Upu> super thx
[19:59] <LA3QMA1> 11dB is usable :o)
[19:59] <Upu> they aren't too expensive what length should I aim for ?
[19:59] <Randomskk> depends. handhelding it?
[19:59] <Randomskk> probably like 2m or so
[19:59] <Randomskk> get decent coax
[19:59] <Randomskk> or at least, half decent coax
[19:59] <Randomskk> or at least, not the cheapest coax. :P
[20:00] <Upu> RG-8X or RG58A ?
[20:00] <Upu> I assume thats the coax grade ?
[20:00] <Randomskk> yea
[20:00] <Randomskk> well coax name/type/grade
[20:00] <Randomskk> RG58A specifies coax parameters
[20:00] <Randomskk> RG213 is the good stuff, RG58 is thinner but not awful
[20:00] <Upu> RG-213
[20:00] <Upu> ok
[20:01] <Upu> amusingly it seems i've picked someone elses shopping basket up
[20:02] <JackNorris> Or you posted your session to the entire channel ;)
[20:02] <Upu> well yeah
[20:02] <fsphil> d'oh
[20:02] <DanielRichman> I once picked up someones session at the checkout stage
[20:02] <Randomskk> haha
[20:02] <DanielRichman> after entering payment details
[20:02] <Randomskk> why the hell don't they use cookies
[20:02] <Upu> someone is about to buy a 817 they didn't want :)
[20:03] <Randomskk> Upu: well
[20:03] <DanielRichman> Randomskk: it's super annoying, if you open bookmarks to 2 products on radioworld and try to add them both to your cart
[20:03] <Randomskk> DanielRichman: hah I can imagine
[20:03] <Randomskk> stupid design
[20:03] <JackNorris> Upu its your basket lol :P
[20:03] <Randomskk> Upu: the URL you pasted to the yagi included your session ID
[20:03] <Randomskk> so anyone clicking it could then add products to "your" bascket
[20:03] <Randomskk> basket* even
[20:03] <Randomskk> including JackNorris it would appear
[20:04] <LA3QMA1> if you are going to use this portable i would just use 1-2m RG58
[20:04] <DanielRichman> basically if you click checkout we're going to deliver the warehouse to your doorstep
[20:04] <LA3QMA1> anything longer i would use RG213 or ecofelx10-15
[20:05] <Upu> hmm contact information "Call sign"
[20:05] <Upu> M00M00 ?
[20:05] <Randomskk> they verify your order over the air
[20:06] <Upu> that will be difficult :)
[20:07] <Randomskk> they start to regret using long session IDs
[20:07] <Upu> did you really just type all that ? :)
[20:08] <Randomskk> no
[20:08] <Randomskk> python
[20:08] <Upu> ah thats ok then :)
[20:09] <Randomskk> which I wrote just now, that is
[20:09] <Upu> I haz radio
[20:10] <Randomskk> huzzah
[20:10] <fsphil> yay
[20:10] <Randomskk> I want me an ft817 for this home radio station thing
[20:10] <Randomskk> oh well
[20:10] <Randomskk> later
[20:10] <Randomskk> there are other things on my list
[20:11] <Upu> Right JackNorris you better be ready for a scale up test :)
[20:11] <Randomskk> Upu: do you have a license yet?
[20:11] <Randomskk> (/are you getting one?)
[20:11] <Upu> no I might do the basic one
[20:11] <Upu> well I will do it
[20:11] <Randomskk> it's worth doing for funsies, it's pretty straightforward
[20:11] <fsphil> definitely, it's deadly simple
[20:12] <Upu> Been a while but I did alot of the theory at university
[20:12] <JackNorris> say what... :P
[20:12] <Upu> so jack here
[20:12] GW8RAK (~chatzilla@client-86-23-9-91.brhm.adsl.virginmedia.com) joined #highaltitude.
[20:12] <Upu> is going to put one of these on my car http://www.servocity.com/html/tube_servo_power_gearboxes.html
[20:12] <Upu> with a Yagi on it and something that makes it point to what ever is up there
[20:12] <Upu> no matter what the orientation of the car
[20:13] <Randomskk> awesome
[20:13] <Randomskk> gyroscopic gimbal mount with servos inside?
[20:13] <Randomskk> or just active monitoring of the car and servos inside?
[20:13] <JackNorris> Well, actually if the car somehow ends up... upside down that could be an issue
[20:13] <Randomskk> :P
[20:13] <Upu> Be the worlds first Quattro Yagi
[20:14] <LazyLeopard> Fascinating site. Anyone know of a UK-based equivalent?
[20:14] <Upu> I think all the bits are inside the card sensors and the like
[20:14] <Upu> If you find one let me know Lazy
[20:14] <Upu> that setup is about £200 with metal gears from the US
[20:15] <LazyLeopard> Not found one yet... and importing from the US can get expensive when HMRC get involved...
[20:16] <Upu> yeah
[20:16] <Upu> will put it through the company as an engineering sample or something
[20:36] <Randomskk> hmm well I have everything set up again... and my main.bin is size zero
[20:36] <Randomskk> I blame my .ld
[20:36] <Randomskk> but it works fine for the libopenstm32
[20:36] <Randomskk> sigh
[20:38] GW8RAK (~chatzilla@client-86-23-9-91.brhm.adsl.virginmedia.com) left irc: Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716]
[20:40] natrium42 (~natrium42@ joined #highaltitude.
[20:41] wonderchook (~wondercho@pool-173-79-184-194.washdc.fios.verizon.net) joined #highaltitude.
[20:41] <natrium42> DanielRichman: so, did you create my user?:P
[20:42] <natrium42> and am i a sudoer?
[20:46] SpikeUK (~chatzilla@host86-161-154-254.range86-161.btcentralplus.com) left irc: Quit: ChatZilla 0.9.86 [Firefox 3.5.11/20100701023340]
[20:50] <Upu> You have a user natrium42
[20:50] <Upu> did you supply a ssh key ?
[20:51] <natrium42> yes
[20:51] <Upu> ah you did
[20:51] <Upu> just login with that
[20:51] <natrium42> but i left the private key at home
[20:52] <Upu> ah :)
[20:53] <Upu> want me to reset you password ?
[20:53] <Randomskk> password auth is off either way
[20:53] <natrium42> it's ok
[20:53] <Upu> oh unlucky :)
[20:53] <Randomskk> natrium42: might be worth generating a key you carry around, if you normally have a usb stick on you
[20:53] <natrium42> i usually just carry around my brain with passwords
[20:53] <natrium42> but it's ok :P
[20:54] <Upu> keys are more secure :)
[20:54] <natrium42> meh
[20:54] <natrium42> nobody is going to hack it
[20:54] <Upu> hah
[20:54] <Randomskk> except bots
[20:54] <DanielRichman> natrium42: yes you have sudo access. Can I ask that if you do decide to install anything, use aptitude (so that anything not explicitly installed is marked "automatically installed") rather than apt-get which doesn't have t hat feature
[20:54] <Upu> don't you believe it
[20:54] <DanielRichman> will help to keep track of cruft on a system with many admins
[20:54] <Upu> I'll put money it gets hit every hour with root login attempts
[20:54] <Randomskk> DanielRichman: why don't we symlink apt-get to aptitude? or put it in everyone's bash_aliases
[20:54] <DanielRichman> Upu: yeah I've already noticed tonnes of hits in the logs. Usually I move ssh to a different port
[20:55] <DanielRichman> but we've already turned off pwauth so that would make it even more inconvenient
[20:55] natrium42 (~natrium42@ left irc: Read error: Connection reset by peer
[20:55] <Upu> I wouldn't worry Daniel
[20:55] <Upu> they'll still find it
[20:55] natrium42 (~natrium42@ joined #highaltitude.
[20:55] <Randomskk> once pw auth is off it's really not a concern
[20:55] <Upu> security through obscurity isn't
[20:55] <Randomskk> no one's going to brute force a SSH key
[20:55] <DanielRichman> I know that's security through obscurity
[20:55] <DanielRichman> it just saves bytes on the logs >.>
[20:55] Action: Upu checks the disk space
[20:55] <DanielRichman> anyone wiedling nmap would find it
[20:55] <DanielRichman> haha
[20:55] <Upu> I think you're ok ....
[20:55] <DanielRichman> :D
[20:56] <DanielRichman> besides the root pw is empty and root logins are disabled
[20:56] <natrium42> DanielRichman: aptitude, got it
[20:56] <DanielRichman> oh an natrium42 are you familiar with etckeeper?
[20:56] <DanielRichman> finally natrium42 you might want to add yourself to /etc/aliases
[20:57] <natrium42> never heard of etckeeper
[20:57] <DanielRichman> (that box doesn't receive mail; habhub.org is handled by google apps. However it means that you'll gget root@localhost's mail; mainly spam from monit. All good stuff)
[20:58] <DanielRichman> natrium42: basically /etc/ is a git repo. etckeeper autocommits it everytime somethign is installed to separate changes from chagnes made by apt
[20:58] <Randomskk> but also, if you edit anything manually it's good practice to commit yourself
[20:58] <natrium42> cool
[20:58] <DanielRichman> sudo etckeeper commit; if you see yourself changing anything in /etc/
[20:58] <Randomskk> sudo etckeeper commit -m "I made some changes"
[21:00] <DanielRichman> natrium42: everything's setup. RJH suggested that we install a copy of the current trackerpredictor setup on that box and start to move people over to it; right now there's a holding page. I'm looking at making the predictor work a bit faster.
[21:00] <DanielRichman> well, the download
[21:00] <Randomskk> oh shit bbl
[21:00] <Randomskk> sherlock is on :o
[21:00] <DanielRichman> after that I think we're going to be looking at this whole couchdb tracker thing
[21:01] <natrium42> DanielRichman: sounds good
[21:01] <natrium42> what should the front page be?
[21:02] <natrium42> calendar of upcoming launches?
[21:02] <Randomskk> aww meh, I missed the start
[21:02] <Randomskk> will watch it on iplayer later tonight
[21:03] <Randomskk> natrium42: the plan at the moment is a kind of unified interface
[21:03] <Randomskk> the page is a full screen map
[21:03] <Randomskk> if there's a current flight, that's on the map
[21:03] <Randomskk> but you can bring up the predictor box and it'l add a track to the map
[21:03] <natrium42> hrm
[21:03] <Randomskk> and bring up the archive box and add an archived track to the map
[21:03] <natrium42> so where does calendar go then?
[21:03] <DanielRichman> in its own box ;)
[21:03] <Randomskk> sounds reasonable
[21:03] <DanielRichman> the default behaviour of the homepage changes depending on if there's a flight
[21:03] <Randomskk> it's all very work in progress at the moment, evidently
[21:04] <DanielRichman> if there's something in the air then I can see it focusing straight on that
[21:04] <natrium42> hmm, might work
[21:04] <DanielRichman> otherwise I guess by default it would want to present the user with the archive, predictor and "next flight coming up" boxes
[21:05] <natrium42> but for now, old tracker will live in /tracker
[21:05] <natrium42> ?
[21:07] <natrium42> could somebody replace my public ssh key with http://pastebin.com/h82CGBQx
[21:07] <DanielRichman> rjh suggested installing a copy of teh current tracker; which sounds like a good idea. Only arguemnts are against are that it might be effort (?)
[21:07] <natrium42> nah, it's easy
[21:07] <DanielRichman> natrium42: add to or replace?
[21:07] <natrium42> hmm, or add
[21:08] <natrium42> i just didn't take my private key with me
[21:08] Action: natrium42 is at starbucks
[21:08] <DanielRichman> :O
[21:08] <DanielRichman> natrium42: go ahead and connect
[21:09] <DanielRichman> for the record, I did check that he was authed to nickserv before doing that
[21:09] <natrium42> :D
[21:09] <natrium42> security geek :P
[21:09] <Randomskk> before adding a random public key you found on pastebin to a sudo account? :P
[21:10] <natrium42> XD
[21:10] <natrium42> maybe we should all meet in person and setup ssh keys
[21:10] <Randomskk> hah
[21:10] <Randomskk> generate an emergency key and split it into pieces
[21:10] <Randomskk> secret sharing
[21:10] <DanielRichman> emergency?
[21:10] <DanielRichman> to do what?
[21:11] <Randomskk> do what the dnssec guys did
[21:11] <Randomskk> who knows?
[21:11] <Randomskk> though it is potentially a neat idea for backup restoration, that said
[21:11] <DanielRichman> but ... but... there's no point having an emergency key if someone can just root the box and take full control
[21:12] <Randomskk> e.g. ssss-split and ssss-combine on nessie
[21:12] <Randomskk> if I said DanielRichman, here's your part, 1-ddd1c6e70149d78e
[21:12] <Randomskk> and let's say natrium42 got 2-a7b9ef37cb00bca4
[21:12] <DanielRichman> awesome thx
[21:12] <Randomskk> and let's say I used a (2,3) scheme
[21:12] <Randomskk> three parts generated, two parts required for recombination
[21:13] <Randomskk> if you possess less than 2 parts you have absolutely no knowledge of the secret
[21:13] <Randomskk> with any two of the three generated parts, you can get the secret
[21:13] <Randomskk> so e.g. ssss-combine -t2 and input yours and, say, natrium42's
[21:14] <DanielRichman> does posessing 2 of the 3 secrets in addition allow me to calculate the third secret?
[21:14] <Randomskk> no
[21:14] <Randomskk> well, maybe?
[21:15] <Randomskk> I don't think so, but I don't know for sure.
[21:15] <Randomskk> different shares are generated each time you run, even with the same secret
[21:15] <DanielRichman> hmm.
[21:15] <Randomskk> the main point is that you can use this to generate shares that mean you can safely give out these parts, and unless the critical mass of people get together, no one gets the secret
[21:16] <DanielRichman> um
[21:16] <DanielRichman> the guy who generated it
[21:16] <Randomskk> obviously there's that
[21:16] <natrium42> i am in
[21:16] <DanielRichman> :)
[21:16] <natrium42> lusers
[21:16] <natrium42> huhu
[21:17] <DanielRichman> can we purge ssss now?
[21:17] <Randomskk> aww I uninstalled both talk and finger
[21:17] <DanielRichman> :D
[21:17] <Randomskk> DanielRichman: yea, done
[21:17] <Randomskk> talk was great fun
[21:17] <Randomskk> but wall is okay for annoying people
[21:18] dcrand (~d_crand@68-117-98-7.dhcp.eucl.wi.charter.com) joined #highaltitude.
[21:18] <DanielRichman> thank god there's a dotfile to disable receipt of talk stuff
[21:18] dcrand (d_crand@68-117-98-7.dhcp.eucl.wi.charter.com) left #highaltitude.
[21:18] <DanielRichman> it got quite annoying when I was actually using vim
[21:18] <Randomskk> yea
[21:19] <m1x10> hi all
[21:19] <Randomskk> oho, there is also `write`
[21:19] <DanielRichman> don't go there
[21:20] <Randomskk> I can't tell if this is more or less annoying
[21:20] <Randomskk> perhaps more
[21:21] <Randomskk> DanielRichman: mesg n
[21:21] <natrium42> so, i guess i will just copy tracker over
[21:21] <natrium42> no need to link it to git, as we are going to start from scratch
[21:21] <DanielRichman> natrium42: yeah aslong as you leave /var/www/default/index.html alone for now
[21:22] <DanielRichman> perhaps into /var/www/default/tracker ?
[21:22] <DanielRichman> natrium42: there's a mysql install and the root password is chmod 400 in /root
[21:23] <natrium42> k
[21:28] RocketBoy (~steverand@ joined #highaltitude.
[21:29] <natrium42> is php not added to httpd?
[21:29] <Randomskk> it.. is
[21:29] <DanielRichman> oh right which directory do you want it on?
[21:30] <natrium42> /var/www/default
[21:30] <natrium42> /tracker
[21:30] <DanielRichman> do you need it on just /tracker?
[21:30] <natrium42> yeah
[21:31] RocketBoy (steverand@ left #highaltitude.
[21:35] <natrium42> bbl
[21:36] natrium42 (~natrium42@ left irc: Quit: My other car is a cdr.
[21:38] <DanielRichman> (done)
[21:40] <Randomskk> ugh. I can't even tell how much porting the ST guys have done to the uip library
[21:40] <Randomskk> diff seems to think the files are quite significantly different
[21:41] <DanielRichman> hmm quite a lot of sql injection holes here
[21:41] <Randomskk> delish
[21:41] <Randomskk> this is why we use frameworks and ORMd
[21:41] <Randomskk> ORMs*
[21:42] RocketBoy (~steverand@ joined #highaltitude.
[21:42] <DanielRichman> heh
[21:42] <DanielRichman> atleast there's no use of eval
[21:57] Laurenceb (~laurence@host86-152-35-235.range86-152.btcentralplus.com) joined #highaltitude.
[21:57] <Laurenceb> hi
[21:57] <Randomskk> yo
[22:01] <SpeedEvil> hi
[22:02] natrium42 (~natrium@CPE000625d867e2-CM0014045885be.cpe.net.cable.rogers.com) joined #highaltitude.
[22:03] <DanielRichman> natrium42: will the mysql user for the tracker require any more than select update delete insert permissions?
[22:03] <natrium42> nope
[22:03] <DanielRichman> good.
[22:04] <Randomskk> does it req delete?
[22:04] <DanielRichman> it uses replace into; that might require delete?
[22:04] <DanielRichman> or maybe just update
[22:04] <Randomskk> dunno
[22:04] <DanielRichman> To use REPLACE, you must have both the INSERT and DELETE privileges for the table.
[22:05] RocketBoy (~steverand@ left irc: Quit: RocketBoy
[22:08] <DanielRichman> natrium42: php works btw
[22:09] <natrium42> great
[22:09] <natrium42> do we have a logo?
[22:09] <natrium42> or should i just modify spacenear.us logo?
[22:10] <DanielRichman> I don't know. It would work for something temporary, at the very least, but we don't have an official logo for habhub yet
[22:13] timbobel (~timboebl@5ED040EE.cable.ziggo.nl) joined #highaltitude.
[22:25] <natrium42> DanielRichman, did you create db user?
[22:25] <DanielRichman> no, I haven't yet
[22:25] <DanielRichman> mysql password is chmod 400 in /root/
[22:26] <DanielRichman> I can if you want
[22:28] m1x10 (m1x10@ppp089210151148.dsl.hol.gr) left irc: Quit: I eat electrons for breakfast and I vomit thunders
[22:29] <DanielRichman> ping natrium42
[22:29] <natrium42> yo
[22:29] <natrium42> it's ok, i will create it
[22:30] <DanielRichman> sudo su is evil ; sorry to nag
[22:30] <DanielRichman> if you use standard sudo - as annoying as it is - then command executed as root go into the log
[22:30] SpeedEvil (1000@tor/regular/SpeedEvil) left irc: Quit: Leaving.
[22:30] <Randomskk> also sudo su should never be used, if you really want a root session sudo -i
[22:30] SpeedEvil (~user@tor/regular/SpeedEvil) joined #highaltitude.
[22:30] <DanielRichman> which is useful on a server with multiple roots so that we know what each other has done
[22:31] <natrium42> meh
[22:31] <DanielRichman> sudo -s lets you keep your env
[22:44] Jasperw (~jasperw@212-98-44-230.static.adslpremium.ch) left irc: Ping timeout: 276 seconds
[23:11] <natrium42> DanielRichman, something weird is going on
[23:12] <natrium42> mysql_connect() never returns in PHP
[23:12] <natrium42> see http://habhub.org/tracker/test.php
[23:12] SpeedEvil (~user@tor/regular/SpeedEvil) left irc: Ping timeout: 260 seconds
[23:13] <DanielRichman> natrium42: try now
[23:13] <natrium42> better, what was the problem?
[23:13] <DanielRichman> sudo aptitude install php5-mysql
[23:14] <natrium42> aah
[23:14] <DanielRichman> my fault; I installed it but forgot that it doesn't come as standard
[23:14] <natrium42> http://habhub.org/tracker is online now
[23:14] <natrium42> no prob
[23:14] <DanielRichman> awesome
[23:15] <DanielRichman> natrium42: where did that data come from?
[23:15] <Upu> nice we even have a logo :)
[23:15] <DanielRichman> or did you just copy the whole db over?
[23:15] <natrium42> yah
[23:16] <DanielRichman> cool. I never asked; how does rjh's server post info to the tracker?
[23:16] <natrium42> is rob planning to copy the dlserver too?
[23:16] <DanielRichman> yeah he claimed he could get it set up in a day
[23:16] <DanielRichman> if I see him tomorrow I'll ask him to :)
[23:16] <natrium42> it just posts the data to spacenear.us
[23:16] <natrium42> it's easy to repoint it to habhub, though
[23:16] Daviey (~Daviey@ubuntu/member/daviey) left irc: Excess Flood
[23:17] Daviey (~Daviey@bootie.daviey.com) joined #highaltitude.
[23:17] <DanielRichman> well if we install the dlserver itself (obviously will need a backup for those using older versions) then we can just have it put it straight into the mysql database
[23:17] <natrium42> yep
[23:17] <DanielRichman> natrium42: if any php is involved in a rewrite of this thing we need to be more careful about SQL injection
[23:17] <Upu> anyone mind if I install SNMP ?
[23:17] <natrium42> agreed
[23:18] <DanielRichman> Upu: I was hesitant to 'cause I have no idea how to configure it securely; it looks like the kinda software where you can give anyone on the network full root access very easily
[23:18] <DanielRichman> but go ahead
[23:18] Daviey (~Daviey@bootie.daviey.com) left irc: Changing host
[23:18] Daviey (~Daviey@ubuntu/member/daviey) joined #highaltitude.
[23:18] <natrium42> we could switch to python
[23:18] <DanielRichman> Upu: is there any way to make SNMP read-only?
[23:18] <DanielRichman> natrium42: well I *reckon* that it mighte ven be possible to have no php or python involved at all
[23:18] <Upu> Yeah thats all I was going to do anyway I'll make sure it's secure
[23:18] <Upu> do we have any sort of firewall script ?
[23:18] <DanielRichman> if the predictor is just a C binary (cgi) and the tracker is couchdb
[23:19] <natrium42> and use sqlalchemy or something
[23:19] <natrium42> the predictor is in python
[23:19] <DanielRichman> http://groups.google.com/group/ukhas-tools/browse_thread/thread/584fdaa0e5229ab9
[23:19] <DanielRichman> natrium42: I'm looking at moving the bits that are written in python to native C
[23:19] <natrium42> why???
[23:19] <DanielRichman> and adam is investigating moving the bits that are in C to python
[23:20] <natrium42> lol
[23:20] <Randomskk> Upu: there is a firewall script, but I'm not sure where DanielRichman put it
[23:20] <Randomskk> natrium42: atm, the predictor is entirely C
[23:20] <DanielRichman> Upu: /etc/iptables.rules
[23:20] <Randomskk> but there is a python script to download the GFS data and export it to a CSV format the predictor can read
[23:20] <natrium42> so just keep it in C
[23:20] <natrium42> Randomskk, oh rite
[23:20] <Randomskk> DanielRichman is investigating having the C download the data
[23:20] <DanielRichman> natrium42: having the predictor in two halves like it is means you can't download depending on where the prediction goes etc
[23:20] <Randomskk> I'm investigating having Python run the prediction, but not as seriously
[23:21] <DanielRichman> Python is slow, though not as slow as th e gfs servers
[23:21] <Randomskk> either way, moving away from PHP is excellent
[23:21] <Randomskk> I'm kind of here and kind of paying attention to Sherlock
[23:21] <DanielRichman> Upu: however in order to update the iptables rules you have to flush out the current ones, I believe
[23:21] <natrium42> yeah, ok, so predictor can be entirely in C
[23:21] <natrium42> but the backend should use a scripting language
[23:21] <DanielRichman> Upu: give me a shout, or I can carry on explaining
[23:21] <DanielRichman> upstart script; $ cat /etc/init/iptables.conf
[23:21] <DanielRichman> pre-start exec /sbin/iptables-restore < /etc/iptables.rules
[23:21] <DanielRichman> post-stop exec /sbin/iptables -F
[23:22] <Upu> ok and are you then adding rules in on the command line ?
[23:22] <DanielRichman> wait... hold on a second, where did all the rules go?
[23:22] <DanielRichman> Upu: did you flush them?
[23:23] <Randomskk> natrium42: there might not be a backend, DanielRichman thinks
[23:23] <Randomskk> couchdb is a web-accessed DB that returns results in json
[23:23] <Randomskk> you could have the entire thing in JS client side
[23:23] <Randomskk> but if there was a backend, I'd definitely say python
[23:23] <Upu> haven't touched them :)
[23:23] <natrium42> so how do you admin users/groups?
[23:23] <DanielRichman> couchdb does that
[23:24] <DanielRichman> it's very cool
[23:24] <Upu> don't worry about it can wait :)
[23:24] <Upu> its getting late and I have work in the morning
[23:24] <DanielRichman> Upu: oook... I think it's open, so you go ahead and set up SNMP while I look at what on earth ahppened here ;P
[23:24] <Upu> lol
[23:24] Action: Upu books in the server for weekly nmap pen tests
[23:25] <DanielRichman> too slow, I'm already nmapping it
[23:25] <Upu> will look at it tommorrow night all
[23:25] Upu (~Upu@ubn.upuaut.net) left irc: Quit: zzz
[23:26] <DanielRichman> it looks like I was testing the upstart script and forgot to bring it back again.
[23:26] <natrium42> we could use ExtJS
[23:26] <natrium42> for the GUI
[23:28] <natrium42> well, boo, couchdb is not relational
[23:28] SpeedEvil (~user@tor/regular/SpeedEvil) joined #highaltitude.
[23:29] <Randomskk> hehe no
[23:29] <Randomskk> edmoore is a fan
[23:29] <Randomskk> it is excellent at replication
[23:29] <natrium42> looks fine for tracking data
[23:29] <Randomskk> yea
[23:29] <Randomskk> the advantage being that trackers in the field store to their own db
[23:29] <natrium42> but i don't know about all the other data
[23:29] <Randomskk> and replicates when they get connection
[23:30] <Randomskk> question is, at the moment, what other data
[23:30] <natrium42> the other data is very OO
[23:30] <natrium42> object is balloon, chasecar, receiver
[23:31] <natrium42> they have many common properties
[23:31] <Randomskk> yes
[23:31] <natrium42> and some unique ones
[23:31] <Randomskk> so then
[23:31] <Randomskk> I think potentially we want two systems, one of which is for received data so that it's easily replicated and offline chasers can get maps
[23:32] <Randomskk> but equally server-side we get a full on web app that has a proper relational database?
[23:32] <natrium42> sounds good to me
[23:32] <natrium42> each object has a track that is in the couchdb
[23:33] <DanielRichman> allright we have a firewall again...
[23:33] <natrium42> besides objects, there are also users and groups that are kept in a relational db
[23:33] <DanielRichman> just remind me to check if it's still there after a reboot
[23:34] <natrium42> yay firewall :)
[23:34] <DanielRichman> I don't really plan on rebooting the box just to see if it works
[23:35] <natrium42> bbl food
[23:35] <Randomskk> there's definitely a lot of discussion to have wrt what we want to make
[23:43] SpeedEvil (~user@tor/regular/SpeedEvil) left irc: Ping timeout: 240 seconds
[23:45] LazyLeopard (~irc-clien@chocky.demon.co.uk) left irc: Quit: Bye
[23:57] timbobel (~timboebl@5ED040EE.cable.ziggo.nl) left irc:
[23:58] sandra_f (~sandra_f@ANantes-552-1-20-225.w86-203.abo.wanadoo.fr) joined #highaltitude.
[00:00] --- Mon Aug 9 2010