Camelia, the Perl 6 bug

IRC log for #crimsonfu, 2013-07-31

crimsonfu - sysadmins who code

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

All times shown according to UTC.

Time Nick Message
00:41 pdurbin_m joined #crimsonfu
00:42 pdurbin_m nice seats: https://plus.google.com/10994058​3958656614840/posts/UJiFX2ScvZQ
00:57 magoo joined #crimsonfu
01:29 mawuli joined #crimsonfu
01:30 mawuli joined #crimsonfu
02:42 magoo joined #crimsonfu
03:40 mawuli joined #crimsonfu
09:34 boegel joined #crimsonfu
09:57 boegel joined #crimsonfu
10:10 _boegel_ joined #crimsonfu
10:25 _boegel_ joined #crimsonfu
10:35 boegel joined #crimsonfu
10:41 _boegel_ joined #crimsonfu
10:46 pdurbin semiosis: heh. "no one understands upstart" :)
10:46 pdurbin (catching up on the backlog)
11:10 chasmo joined #crimsonfu
11:45 boegel joined #crimsonfu
11:51 joshu pdurbin hey ;)
11:52 pdurbin joshu: I'm glad your upstart script works
11:52 pdurbin joshu: I'm sure to be asking you for help some day
11:53 pdurbin right now I'm still on the latest CentOS, which doesn't use upstart
11:53 joshu thanks. hehe no worries I can be your upstart sifu :P
11:53 joshu I've never used CentOS. Only linux dist I've used is Ubuntu.
11:56 pdurbin ubuntu is a solid choice
12:06 joshu pdurbin do you just do linux server backend stuff as your primary work?
12:06 pdurbin no, but I used to
12:08 dotplus semiosis: you saw coresos it on hn? me too. It sounds quite interesting for certain kinds of purposes. I haven't thought hard yet, but perhaps coreos/docker is analagous to vagrant + (vbox|vmware|whatever), only more lightweight (and less mature?). I suspect that there is room for several different stacks of these kinds and the massive proliferation of choices at the moment will settle down to a smaller number
12:14 dotplus the ones that survive will be the ones that achieve a critical mass which is either provided by a small, determined community (think ericsson and erlang (before erlang became "popular"): not many need a telco language) behind a stack that is particularly good for a (less common) use case OR a larger community behind a stack that is suitable for a wider or more common set of use cases (think openbsd which has (had?) a larger community than ...
12:14 dotplus ... erlang because it's suitable for a common use case: everyone needs a router/firewall)
12:16 dotplus One thing is sure... if you can come up with a simple, reliable, flexible way to marshal all these different deployment/provisioning/cloudy systems (or at least most of the popular ones), you're going to have a huge success
12:22 pdurbin so many choices
12:27 * dotplus wants to not pick betamax
12:28 pdurbin dotplus: just pick VHS and you're fine for many years :)
12:36 _boegel_ joined #crimsonfu
13:43 ben_e we're not using docker... yet
13:44 ben_e but it seems to have a lot of momentum for a youngish project
13:48 ben_e i think it was agoddard who was tweeting about docker
13:48 ben_e but i've said something like that before...
13:49 ben_e if it's impossible to actually search twitter, then there's plausible deniability!
14:34 semiosis dotplus: no i dont read HN, i saw it on twitter.
14:34 semiosis i do use openbsd for my router/firewall
14:39 dotplus sure, netiher do I. many don't. but many do. it was an analogy, substitute puppet vs bcfg2, ubuntu vs arch, or even "linux" vs QNX for examples of the larger community:)
15:07 magoo joined #crimsonfu
15:09 melodie joined #crimsonfu
15:09 melodie hello !
15:09 melodie does someone know about  a specific technical criteria in USB sticks which could be isolated, in order to try to find out why one kernel version sees some mass storage peripherals and not others?
15:09 melodie I have 4 USB sticks, 2 of them are seen and 2 of them are not seen, with the session booted to the kernel  3.10.3-1, but all of the 4 are seen when booting to the kernel linux-lts 3.0.87-1
15:10 melodie I have questioned the forum of Archlinux.fr, and other members can only notice it happens for them.
15:10 melodie they can give the output of lsusb and or usb-devices but I don't know what to search for from there, this is why I thought I would come here
15:10 ben_e quality control on usb thumbdrives seems highly variable
15:11 ben_e the freebies that vendors/conferences give away are free for a reason
15:32 melodie joined #crimsonfu
15:32 melodie joined #crimsonfu
15:32 melodie hi ben_e
15:33 melodie I had not seen your answer
15:33 melodie ben_e the thumbdrives which are the subject of the story both work perfectly with some kernels, and not with the latest installed in arch
15:35 melodie some other people at the forum noticed the same behavior : one machine up to date, the usb stick is not detected anymore, and in a machine where the update has not been done, it is detected. In my machine I have the latest regular kernel and the latest LTS kernel : the sticks are detected correctly in the session with the LTS kernel only
15:35 melodie this is why I try to get answers at the chan incase someone has an answer on which criteria should be checked on?
16:32 mawuli joined #crimsonfu
17:14 magoo joined #crimsonfu
17:23 magoo joined #crimsonfu
17:41 Hien joined #crimsonfu
17:45 magoo joined #crimsonfu
17:58 semiosis yo dawg i heard you like virtualization... https://www.dropbox.com/s/vr48ny7p3id37te/​1._corelocalhost_bash_20130731_101616.png
19:58 magoo joined #crimsonfu
20:03 joshu hi semiosis
20:03 * semiosis waves
20:04 joshu hehe how did you do that big dot?
20:04 joshu some irc command?
20:05 semiosis /me waves
20:05 * joshu waves back
20:06 joshu hehe
20:06 joshu new upstart script feedback please https://gist.github.com/ano​nymous/a238133933700bc267df
20:06 semiosis you'll be operating bots in no time!
20:06 joshu the top one is mine and the bottom one is from ubuntu
20:06 * joshu rofl
20:15 semiosis no feedback on that upstart job
20:15 semiosis does it work?
20:15 semiosis "If it compiles, it is good; if it boots up, it is perfect." -Linus
20:15 joshu that upstart script to cart brscan-skey isn't working. When I run it manually user test, e.g. brscan-skey it start the binary in the background until logoff and works correctly. The idea with the upstart is to have it start on boot and respawn if it fails
20:15 joshu semiosis hehe
20:16 semiosis well not sure how that program will handle multiple users... maybe you need to use sudo or su in upstart to run it as your user
20:17 semiosis another thing to note is that the environment in upstart isn't the same as the environment when you run commands in a shell
20:17 semiosis you may need to set env vars in the upstart job for the program to work
20:17 semiosis bbiab
20:17 joshu just for single user.
20:17 joshu https://gist.github.com/ano​nymous/1ea5ecb81ac5221d8e8c
20:17 joshu ok
21:35 melodie joined #crimsonfu
21:35 melodie joined #crimsonfu
21:42 joshu semiosis i have been able to make the upstart script work, but when I shutdown or reboot the OS hangs and I have to force shutdown or force reboot.
21:50 joshu I've confirmed that when I remove the upstart script ubuntu shuts down properly
22:01 melodie joshu there are new commands for clean shutdowns
22:01 melodie you might want to check the forums, or askubuntu.com
22:02 melodie hello :)
22:02 joshu melodie I have googled not like I just ask for help without trying ;)
22:11 semiosis joshu: i would try without any stop-on statement
22:12 joshu semiosis is the rest ok the expect daemon and respawn?
22:13 semiosis no way for me to know
22:13 semiosis the syntax appears to be valid, but how should i know if they're correct for your program
22:13 semiosis do you get any messages about this in syslog?
22:14 joshu semiosis sure how should I know..hehe I didn't make it ;) I'm trying now without the stop i'll let you know in 1min if it hangs again
22:21 joshu semiosis that fixed it thank you!! Don't get why it hung with the stop on, but without it it doesn't hang some I'm happy :D
22:22 semiosis verify that process starts ok with upstart using ps, then check 'service $whatever status' to see if upstart is still tracking the process
22:22 semiosis just a wild guess though
22:22 semiosis glad it works
22:23 joshu semiosis done and done. test       585  0.1  0.2 169816  1324 ?        Sl   00:20   0:00 /opt/brother/scanner/brsca​n-skey/brscan-skey-0.2.4-0
22:23 joshu brscan-skey-daemon start/running, process 583
22:23 semiosis well then idk why the stop fails
22:23 joshu I even tried killing it with sudo kill 583 and after that sudo status…. and it still said process 583
22:24 joshu good or bad?
22:24 joshu if it respawns when killed shouldn't it have a different process number?
22:26 semiosis yes
22:29 joshu so something is wrong with my script?
22:35 semiosis maybe?
22:35 semiosis i gotta run, sorry
22:37 * joshu waves bye
23:24 joshu it's working now semiosis killing it and it respawns a new process id ;)
23:28 joshu semiosis it don't work properly with exec su -c, only with setuid…but with setuid it want to write to /brscan and not /home/test/brscan. I tried using chdir /home/test but that doesn't help.
23:44 mawuli joined #crimsonfu

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

crimsonfu - sysadmins who code