Current Server Time: 19:45 (Central Europe)

#apertus IRC Channel Logs

2017/05/17

Timezone: UTC


01:26
Bertl
off to bed now ... have a good one everyone!
01:26
Bertl
changed nick to: Bertl_zZ
02:23
jucar
joined the channel
02:56
jucar
left the channel
03:22
slikdigit
joined the channel
04:59
slikdigit
left the channel
05:18
slikdigit
joined the channel
07:02
slikdigit
left the channel
08:19
Bertl_zZ
changed nick to: Bertl
08:19
Bertl
morning folks!
10:14
Rex0r
Is there the possibility of capturing the full sensor readout for larger than 4k video?
10:15
Rex0r
(From Twitter)
10:18
Bertl
the sensor is 4k i.e. 4096x3072 so there is nothing more to capture than that
10:19
Bertl
but if you assume that 4k (wide format) is 3840x2160 then yes, you can capture slightly more than that
10:27
max_bxl
joined the channel
10:27
max_bxl
hello everyone !
10:27
Bertl
hey max_bxl!
10:28
max_bxl
quick question : do we have a list of wifi dongle we already know that work with the beta ?
10:30
Bertl
[root@beta ~]# find /lib/modules/`uname -r`/kernel/drivers/net/wireless/ -iname '*.ko*'
10:30
Bertl
will give you a list of wifi drivers currently compiled in
10:31
Bertl
you can then check each of them with 'modinfo <module>' what hardware it supports
10:31
Bertl
e.g. modinfo rtl8192de
10:32
max_bxl
great !
10:32
Bertl
basically realtek variants rtl81* have been tested/used so far
10:34
max_bxl
ok
10:36
max_bxl
another question : I'm making the alpha-like plexiglass case and I was thinking to make (lots of small) holes just behind the fan ("rear" part of the case then). Should I consider making other holes on other sides (there will be holes for power, ethernet, hdmi anyway)?
10:38
Bertl
well, in a closed case, it is all about the air flow
10:38
Bertl
so you need to carefully consider where the air comes in and where it leaves the case
10:39
Bertl
obviously the main point to leave the case is where the fan is
10:39
Bertl
and I would make sure that there is near to nothing which keeps the air from flowing out there
10:40
Bertl
i.e. best make some kind of wind tunnel which directly connects to the fan top
10:40
Bertl
if you simply put a few 'holes' at the back and have a gap between fan and case, the hot air will mostly circulate inside the case, which is not what you want :)
10:41
max_bxl
ok
10:42
max_bxl
so more one big hole (the size of the fan) + something to make a tunnel from the exit of the fan to this hole
10:48
Bertl
yeah, something like that
10:49
Bertl
also measure the various temperatures on the camera and monitor them with and without case
10:50
Rex0r
And capturing the full sensor output is just a case of running some commands yeah?
10:50
Bertl
a script for collecting all the temp sensor data would be nice btw
10:50
Bertl
Rex0r: cmv_snap does capture the full sensor data
10:50
Bertl
i.e. the resulting image will be 4096x3072
10:51
Rex0r
When do you reckon you'll have done this by max_bxl? Will you take plenty of photos of the build process?
10:51
Rex0r
Thanks.
10:56
intracube_afk
changed nick to: intracube
10:58
max_bxl
hopefully will do it this afternoon ! Of course, plenty of pictures !
11:00
max_bxl
for the temperature, I guess I should use ./zynq_info.sh ? Any other command to monitor other parts ?
11:01
Bertl
that will only give you the zynq temp, there is a temp sensor on the power board as well and the sensor register 127 provides information as well
11:02
Bertl
cat /sys/class/hwmon/hwmon0/temp[0-9]_input
11:02
Bertl
cmv_reg 127
11:03
max_bxl
ok
11:04
max_bxl
around/above how much °C is temperature too high ?
11:04
Bertl
that really depends on the part of the camera
11:05
max_bxl
and for a script, how much mesure per minute should I take ?
11:05
Bertl
for the zynq, it will work relatively fine up to 90°C, although that is a little high for normal use
11:05
max_bxl
can you be more specific for each part of the camera ?
11:05
max_bxl
sorry too impatient !
11:06
Bertl
for the sensor, you want to keep that quite low, typically not more than a few degrees above room temperature
11:06
Bertl
(so adding slits or holes near the sensor for air intake is a good idea, but don't over-do it :)
11:07
Bertl
also, always keep in-take holes/slits at the bottom or at least bottom side, because of natural convection
11:08
max_bxl
ok ! two thin slits (3x20mm) on the front side for example ?
11:09
Bertl
might work, really depends on the placement
11:10
Bertl
as a rule of the thumb, you want to have about 1.5-2x the area you get for the exhaust
11:10
Bertl
(in total, which is the inside area of the fan)
11:11
max_bxl
"1.5-2x the area you get for the exhaust" for all the air intakes I'll make (also with the ethernet holes and so on), right ?
11:13
Bertl
yep, but don't count the holes for plugs in full, they will be mostly obscured by connectors (I hope)
11:13
max_bxl
sure ! I thought so ^^
11:14
Bertl
if you have too little intake, the fan gets slowed down, if you have to much, cooling will be ineffective
11:18
max_bxl
noted !
11:18
max_bxl
gotta go to join Nikos, be back online in one hour, hopefully making some holes !
11:19
Bertl
cya
11:32
max_bxl
left the channel
13:38
intracube
left the channel
14:06
intracube
joined the channel
14:16
max_bxl
joined the channel
14:37
jucar
joined the channel
14:39
max_bxl
left the channel
15:02
max_bxl_
joined the channel
15:02
max_bxl_
hello again !
15:04
max_bxl_
new question : where would be the best place to write startup scripts (ex: ./set_gain.sh 2 ; ./gamma_conf.sh 0.3 ; and so on...) ?
15:08
slikdigit
joined the channel
15:31
jucar
left the channel
16:02
Bertl
depends on when you want them to run :)
16:02
Bertl
kick.sh and halt.sh are normally executed on start up and shut down when the cmv12k service is enabled
16:14
anuditverma
joined the channel
16:34
max_bxl_
left the channel
16:53
BAndiT1983|away
changed nick to: BAndiT1983
16:54
BAndiT1983
Bertl, i suppose we need some mechanism which lets daemon execute scripts at startup, as REST gsoc falls into the same category of controls
16:55
Bertl
at what startup?
16:56
BAndiT1983
of ArchLinux -> (05:53:20 PM) max_bxl_: 14:04:02> new question : where would be the best place to write startup scripts (ex: ./set_gain.sh 2 ; ./gamma_conf.sh 0.3 ; and so on...) ?
16:56
BAndiT1983
if we have a daemon, then the scripts can be reworked and also reduced
16:56
anuditverma
Hello everyone
16:57
BAndiT1983
further, some simple script parser for settings is required, but it's not a big deal
16:57
BAndiT1983
hi Anudit
16:57
BAndiT1983
as replacement for .sh ones
16:59
Bertl
well, systemd is running on the Beta anyway, so systemd can trigger whatever setup necessary via the control daemon
16:59
BAndiT1983
that's what i've meant
16:59
BAndiT1983
we need just a small mechanism for that, but this can be elaborated while gsoc is ongoing
17:02
__m__
Hello
17:02
BAndiT1983
hi
17:02
anuditverma
I tried ssh -ing into AXIOM beta setup, it seems the response time on it is very long, may be because of remote access I guess, or is it normal ?
17:06
Bertl
hello __m__!
17:06
__m__
Hej!
17:08
BAndiT1983
should we wait for se6astian?
17:09
__m__
Yes, let's wait
17:10
__m__
I will check if he has time today
17:11
BAndiT1983
by the way, i've tried a bit golang, it was scary how big the compiled executable of "hello world!" was
17:11
__m__
Sebastian won't be here today
17:11
BAndiT1983
golang 1.6 -> 2.3mb without optimization, with 1.6
17:11
BAndiT1983
with 1.8 it was reduced to 1mb, but i don't remember if optimized or not
17:12
BAndiT1983
then we should do a quick meeting
17:12
BAndiT1983
Anudit, you can start
17:13
__m__
The cam seems to be offline atm
17:13
__m__
We will fix that asap, hopefully this week
17:13
anuditverma
So I tried accessing the beta, it seems un responsive
17:13
anuditverma
ohh okay
17:14
__m__
When did you try to access the cam?
17:14
__m__
When did you try to access the cam first?
17:16
anuditverma
It was Saturday I think
17:16
__m__
(so that we know since when it's not accessible)
17:16
__m__
Alright, next time, please mail me or PM Sebastian/Bandit if something like this occurs so we can get rid of stopping tasks right away
17:17
anuditverma
Ok cool
17:17
__m__
Alright, so, what's the status of your work?
17:18
BAndiT1983
i can just forward it, as i'M sitting a little bit far away from vienna ;)
17:18
__m__
Oh yeah, that's right, sorry. So please @anuditverma, mail me or PM sebastian if something with the cam is wrong
17:19
anuditverma
I have started with re writing https://github.com/apertus-open-source-cinema/beta-software/blob/master/axiom_beta_control_daemon/API/Client.h into GO lang
17:19
anuditverma
well it's just a start
17:19
__m__
wrong = not working, not accessible or just very slow
17:19
anuditverma
@__m__ alright I will do it surely
17:19
BAndiT1983
have you committed some code? i need something to test with
17:19
__m__
Could you periodically push your work to the repo?
17:19
__m__
https://github.com/apertus-open-source-cinema/axiom-beta-rest-interface
17:20
BAndiT1983
wanted to write the access in go, but remembered that you are on the task, so i stopped further investigation for now
17:20
__m__
Be so kind to stick to the git-flow scheme of commiting when pushing your work: http://nvie.com/posts/a-successful-git-branching-model/
17:20
anuditverma
well not yet
17:21
aombk
left the channel
17:21
BAndiT1983
is there a "dev" branch yet?
17:22
__m__
I suggest you open a develop branch and work in feature branches
17:22
anuditverma
So I keep a separate branch for my commits ?
17:22
__m__
no, just a master branch with a readme atm
17:22
anuditverma
okay got it
17:22
BAndiT1983
yes, we should merge only stale things into master
17:22
BAndiT1983
*stable
17:22
BAndiT1983
it's a good practice to work on "dev" branch and merge from time to time
17:23
__m__
I would suggest to only work in feature branches and merge the develop when the feature is donw
17:23
aombk
joined the channel
17:23
BAndiT1983
if there are some totally different tests, then one can temporarily create additional branches, which should be removed later, feature-based development
17:23
__m__
And I think it would be great if we only merge proof checked code to master
17:24
BAndiT1983
__m__: we should involve CI early enough
17:24
BAndiT1983
just to have some sort of good maintenance
17:25
__m__
@BAndiT1983, yes, you are right, we should do that. I suggest we do this End of May, when at least some code is in the repo
17:25
BAndiT1983
if it's possible to create a job which would build, test and then merge from dev to master, that would be perfect
17:25
anuditverma
Sorry, what's CI ?
17:25
__m__
I can do it end of may
17:25
__m__
Continous integration
17:25
anuditverma
ok thanks
17:25
BAndiT1983
i mean a separate job, which is executed manually, so we control what is supplied to master
17:25
__m__
*continuous
17:26
__m__
https://travis-ci.org/
17:26
__m__
Committed code will be built automatically and build/test results will be posted on irc/github/,...
17:26
BAndiT1983
so, the feature branches should be merged into "dev", afterwards "dev" to "master" from time to time
17:27
BAndiT1983
but dev to master only manually if possible
17:28
BAndiT1983
Anudit, when do you think some code is there? i could help you to finish the basic implementation, as i really need some base to see what is still missing from the daemon
17:29
__m__
@anuditverma, camera should work, please try to connect
17:30
BAndiT1983
at some point in time i would also need such access, but it's not mandatory currently
17:30
anuditverma
@BAndiT1983 I can push some code this weekend, I have just started
17:31
anuditverma
@__m__ ok trying
17:31
BAndiT1983
alright, as end of may is approaching we should hurry a bit, but the client is not complex at the moment, i think we could complete it this weekend, so it resembles C/C++ one
17:32
anuditverma
okay I will try to finish it fast
17:32
__m__
@anuditverma, great, so you will push this weekend to develop?
17:33
anuditverma
yes, the Go code for the client
17:34
BAndiT1983
i've added a branch "develop", just clone it with your git client and then you can add code
17:35
anuditverma
thanks
17:35
BAndiT1983
don't worry if the code is not finished
17:36
__m__
Alright, so, we meet again next Wednesday 4pm UTC?
17:37
__m__
anuditverma and Betl
17:37
__m__
anuditverma and Bertl are talking about how to access the camera atm in private, so I guess that's settled
17:38
__m__
Do you have any open questions atm @anuditverma?
17:38
aombk
left the channel
17:41
BAndiT1983
anuditverma: have you succeeded in starting the daemon? this would help you to test your client, no witchcraft is required as it's an ordinary executable at the moment
17:42
anuditverma
The remote setup is working fine :) I am testing it bit further with help from Bertl
17:44
__m__
great
17:44
anuditverma
@__m__ I think I am sorted for now, will ping if I need help for anything and yes remote testing is still in process
17:46
__m__
Alright
17:47
__m__
So, next meeting in one week ok for everyone?
17:47
__m__
May 24, 4pm UTC?
17:50
Bertl
it's very likely that I'm around, although I'm not really required I guess :)
17:53
anuditverma
I can come for the meeting, my exam will be over by the day, I can join in the evening.
18:06
__m__
@BAndiT1983, do you have time?
18:08
BAndiT1983
yes, i have time
18:09
BAndiT1983
sorry, was not here for a moment
18:13
__m__
Great, see you all next week 4pm UTC
18:13
__m__
Have a nice day
18:15
Rex0r
left the channel
18:15
Bertl
you too!
18:15
BAndiT1983
bye
18:59
Rex0r
joined the channel
19:11
anuditverma
left the channel
19:11
maxfuh
joined the channel
19:14
max_bxl
joined the channel
20:12
RexOrCine
joined the channel
20:26
se6astian|away
changed nick to: se6astian
20:26
Bertl
off for a nap ... bbl
20:26
Bertl
changed nick to: Bertl_zZ
20:43
max_bxl
left the channel
20:43
maxfuh
left the channel
20:45
champika
joined the channel
20:53
champika
left the channel
21:14
se6astian
changed nick to: se6astian|away
21:15
se6astian|away
changed nick to: se6astian
21:15
se6astian
changed nick to: se6astian|away
21:53
aombk
joined the channel
21:58
Rex0r
left the channel
22:17
BAndiT1983
changed nick to: BAndiT1983|away
23:32
Rex0r
joined the channel
00:13
dimaursu16
left the channel
00:37
intracube
changed nick to: intracube_afk