Current Server Time: 06:56 (Central Europe)

#apertus IRC Channel Logs

2019/02/27

Timezone: UTC


05:28
shivang
joined the channel
05:29
shivang
left the channel
05:29
shivang
joined the channel
05:29
shivang
hello ??
05:30
shivang
\help
05:34
shivang
left the channel
06:38
BAndiT1983|away
changed nick to: BAndiT1983
06:53
Bertl_zZ
changed nick to: Bertl
06:53
Bertl
morning folks!
06:56
moadmmh
joined the channel
07:54
se6astian|away
changed nick to: se6astian
07:56
se6astian
good day
08:09
se6astian
now its official: GSoC 2019 with apertus!
08:09
se6astian
https://summerofcode.withgoogle.com/organizations/6567493276008448/
08:12
Bertl
you are not the first to report that :)
08:13
se6astian
correct, but its my duty as org admin :)
08:32
futarisIRCcloud
left the channel
09:26
se6astian
funny facebook notification: "Your post "Attention: this page is being deleted by ...." is performing better than 95% of posts on that Page. Boost it for €27 to reach up to 123 more people daily."
09:26
se6astian
read up here: https://www.apertus.org/facebook
09:32
Bertl
might even go viral! :)
09:33
BAndiT1983
maybe it will become a meme for everything social media stands for
09:43
se6astian
https://i.kym-cdn.com/photos/images/original/000/339/529/280.jpg
09:45
se6astian
facebook also provides a warning when I try to boost that post
09:45
se6astian
as the term "facebook" is used inside the text: "Your ad might not be approved We require ads featuring the Facebook or Instagram brand to follow specific brand guidelines because all Facebook and Instagram features and icons are legally protected brand assets. "
09:45
BAndiT1983
changed nick to: BAndiT1983|away
10:03
Bertl
off for now ... bbl
10:03
Bertl
changed nick to: Bertl_oO
10:07
se6astian
lets make it go viral: https://cloud.gerade.org/index.php/s/Ji73XbAE9pJ8tRZ
10:07
se6astian
https://cloud.gerade.org/index.php/apps/gallery/s/Ji73XbAE9pJ8tRZ#FB-fail-01.jpg actually
10:07
se6astian
I posted it also on our FB of course
10:08
Bertl_oO
you might want to add some kittens ...
10:08
se6astian
good idea
10:14
se6astian
done
10:14
se6astian
https://imgur.com/a/2AEKqOw
10:15
se6astian
with the added cat image this will be the FB bots trying to make sense of this posts content: https://media1.tenor.com/images/bbae33f21e2868d7ed8aca3d0df24ffe/tenor.gif?itemid=8143522
10:42
BAndiT1983|away
changed nick to: BAndiT1983
12:23
Dev_
joined the channel
12:25
Dev_
left the channel
12:45
rohan_
joined the channel
12:46
rohan_
hey! finally got the image : https://pasteboard.co/I36C04N.png
12:47
rohan_
BAndiT1983 : I swapped the bytes as suggested by you
12:47
BAndiT1983
hi rohan_, very good
12:48
rohan_
also, I spent time making my code much more readable and have already started with the bonus tasks
12:48
se6astian
hi rohan_, great!
12:49
rohan_
should I share the current progress in a repo or should I first do more work?
12:49
se6astian
apertus has officially been listed as org for gsoc 2019 yesterday
12:49
rohan_
thank you BAndiT1983 and se6astian
12:49
se6astian
repo is fine as you can also use it for the application later on
12:50
rohan_
okay
12:50
moadmmh
left the channel
12:51
rohan_
I'll add comments to my code to make it more understandable also
12:51
rohan_
will talk soon!
12:52
rohan_
left the channel
13:42
BAndiT1983
changed nick to: BAndiT1983|away
13:44
Y_G
joined the channel
13:46
BAndiT1983|away
changed nick to: BAndiT1983
13:52
RexOrCine|away
changed nick to: RexOrCine
14:10
BAndiT1983
changed nick to: BAndiT1983|away
14:47
aleb
se6astian: FOSDEM report finished https://blog.aleb.ro/2019/02/fosdem-impressions.html
14:48
Dev_
joined the channel
14:49
sushmita0810
joined the channel
14:51
aleb
BTW, congratulations for being accepted in GSoC! We're also in. :D
14:54
se6astian
great!
14:54
se6astian
thanks for the report, reading
15:03
sushmita0810
left the channel
15:08
Dev_
left the channel
15:59
BAndiT1983|away
changed nick to: BAndiT1983
16:05
BAndiT1983
vup2, anuejn, what is the current state of git submodules in the firmware?
16:06
vup2
under what aspect?
16:07
vup2
there are two submodules atm, one for the control daemon and one for the misc tools
16:07
vup2
i guess you are talking about the control daemon?
16:08
vup2
the commit used is currently 4ea5116cef from 19.12.18
16:08
BAndiT1983
yep, daemonm
16:08
BAndiT1983
it can be updated already, as we have some improvements in the webui and daemon itself
16:09
BAndiT1983
how could we integrate the webui for now, so new firmware has it available from the start on?
16:09
vup2
the repo for the webui is https://github.com/apertus-open-source-cinema/web-remote right?
16:09
BAndiT1983
yes
16:11
BAndiT1983
it uses npm to build final folder
16:11
vup2
then adding it as a submodule to software and installing node and running npm install when the image is assembled should be enough
16:12
vup2
a systemd unit to start it automatically would probably also be nedded, but that should be quite simple
16:13
BAndiT1983
ok, do you need special infos from me which could help?
16:14
BAndiT1983
lighttd would be good for now, also placing the content of dist folder there, to be able to call webui without further setup
16:14
BAndiT1983
*lighttpd
16:14
vup2
what role does the devserver have?
16:14
anuejn
webui just needs a static http server?
16:15
BAndiT1983
for now yes, later we will replace it with socat, when webui will be self-contained app
16:15
anuejn
probably only hot code replacement?
16:15
BAndiT1983
you mean in the node setup?
16:16
vup2
i guess the WSServer needs to be running for the webui?
16:16
BAndiT1983
"npm run build" should be sufficient to get "dist" folder, files there are final
16:16
BAndiT1983
yes, also wsserver, it does the communication between daemon and webui
16:17
vup2
ok so the devserver is not needed?
16:18
anuejn
BAndit1983 how do you want to use socat?
16:18
BAndiT1983
what is devserver?
16:18
anuejn
it probably builds and serves the app...
16:19
BAndiT1983
socat would listen to some port and send self-contained app package to browser
16:19
vup2
the stuff here https://github.com/apertus-open-source-cinema/web-remote/tree/master/devServer
16:19
vup2
why not use a real http server instead of socat?
16:19
BAndiT1983
ah, it's some test server from Francis, for tests without daemon, although i've suggested to use daemon in mock mode also for more real tests
16:19
vup2
ok
16:20
BAndiT1983
it was suggested by someone long time ago, this approach would reduce the calls, as the app doesn't need any more resources from server
16:20
BAndiT1983
if we have webui in browser cache, then only wsserver is used for queries
16:21
anuejn
i dont think, that this leads anywhere usefull
16:21
BAndiT1983
why?
16:22
anuejn
lets just use a simple http server and stay on mainstream grounds for serving a webapp
16:22
vup2
if the webui is in the browser cache it should also only use the wsserver for queries
16:22
vup2
also the performance of a real http server is probably good enough to not care about it
16:22
BAndiT1983
self-contained app doesn'T need a webserver, which would reduce cpu usage and so on
16:23
anuejn
lighttpd is pretty lightweight
16:23
anuejn
i wouldnt bother
16:23
BAndiT1983
do as you think, i don't care at the moment
16:23
vup2
also a self-contained app still needs a webserver for initial load
16:24
BAndiT1983
nope, socat is enough
16:24
BAndiT1983
was also thinking same before, then i'Ve read a bit on the net
16:24
anuejn
but that would then implement a webserver ;)
16:24
vup2
but socat is then doing the same work a normal http server would do
16:24
anuejn
so it IS one
16:24
BAndiT1983
we don't need a real webserver as no comms are done to it
16:25
BAndiT1983
we only need a provider for the app, afterwards there is no data exchange, only with wsserver
16:25
vup2
yes but any non shitty webserver should be at 0% cpu usage if it gets no data
16:26
anuejn
k lets justpone this to the day, when we need the ressources :)
16:26
vup2
does the WSServer have a systemd unit already somewhere?
16:27
BAndiT1983
not yet
16:27
anuejn
k
16:27
BAndiT1983
have to review systemd stuff soon
16:28
vup2
under what aspect?
16:28
se6astian
will head off shortly, do you still need anything from me regarding this topci?
16:29
vup2
i don't think so
16:29
BAndiT1983
we had some discussion about log files before, which should run in parallel, also have to check if installing systemd descriptions still works, there were some shell scripts
16:30
vup2
regarding systemd, i think having the daemon also work without systemd would be quite nice
16:30
anuejn
agree
16:30
BAndiT1983
it is
16:30
vup2
at the moment that should be quite easy, but just something to keep in mind when further integrating with systemd
16:30
BAndiT1983
systemd is only starting it, but you can use it also without, that's how i test usually
16:30
vup2
well it does not compile without systemd :)
16:31
BAndiT1983
of course, as it uses sd-journal
16:31
anuejn
basically all the logging code could just log to stdout
16:31
BAndiT1983
we can also ditch systemd, then i would grab some small lib or create simple logger
16:31
anuejn
i would vote for stdout logging
16:31
vup2
on non systemd systems one could also just log to stdout and stderr and use rsyslog ore something
16:31
vup2
s/ore/or/
16:32
BAndiT1983
ok, will note it
16:33
vup2
i also have nothing againts supporting systemd and using it if it is present, i think having the parts where it is used easily replaceable would be enough at the moment
16:33
se6astian
all the best then, off for now
16:33
se6astian
changed nick to: se6astian|away
16:34
BAndiT1983
it's not a big problem to rework it, as the logging is not good yet and was on the list for review
16:34
anuejn
sytemd collects all output of all services it starts automatically
16:34
anuejn
nice
16:34
vup2
yes but you lose loglevels
16:35
vup2
(but they are just special character sequences anyway)
16:35
anuejn
do we want to reverse-proxy the websockets connetion?
16:35
BAndiT1983
in which case do we lose log levels?
16:35
vup2
when just logging to stdout and sterr
16:35
vup2
*stderr
16:35
BAndiT1983
anuejn, which effects does it have?
16:35
vup2
then there is only the error log and normal log output
16:35
BAndiT1983
vup2, it'S not a problem to create simple logger
16:35
vup2
yes sure
16:36
BAndiT1983
we already have similar one in OC
16:36
anuejn
we only need one port (80) on the camera
16:36
BAndiT1983
i could copy it to daemon also
16:36
vup2
but the a custom logger then would not have nice color highlights in the systemd log :)
16:36
BAndiT1983
if it is possible, then we could do it, but i have not much knowledge there yet
16:36
anuejn
and it gets a path assigned (ie. /backend or /api)
16:37
BAndiT1983
ah, we could try with paths
16:37
anuejn
what setup does the webui currently expect?
16:37
anuejn
is there a special port for the websocket?
16:37
BAndiT1983
npm for build
16:38
BAndiT1983
7070 for wsserver
16:39
anuejn
k
16:40
anuejn
then we probably leave it like this for now, as it eases development a bit
16:40
BAndiT1983
ok
16:42
anuejn
is there anything else to consider?
16:42
anuejn
do?
16:43
BAndiT1983
we have covered most stuff, i will let you do the adjustments to the build/setup process
16:43
BAndiT1983
if you have questions just ask
16:43
BAndiT1983
in the next time daemon and webui will be extended to usable state, so it is more or less important to get it into the firmware
16:44
anuejn
okay will try to find time this we
16:44
BAndiT1983
take your time, it's not urgent
16:44
vup2
regarding bumping the version of the control daemon and also the webui, i think it would be good in the future that whenever you have a stable / tested version of them to just open a pull request on axiom-beta-firmware which bumps the submodule commit
16:45
BAndiT1983
ok, will consider it
16:45
BAndiT1983
the process will contain PR from dev to master, afterwards PR from master to beta firmware repo
16:47
vup2
it would also be possible to put a big disclaimer on the daemon repo that the master is untested and that they should use the commit used in the axiom-beta-firmware repo if they want a stable version
16:47
vup2
then development of the daemon could happen on the master branch
16:48
anuejn
seems legit
16:48
vup2
but when we are ready to provide real stable version of the firmware images, we will also probably need something like a testing branch
16:48
anuejn
no really, i like thin aproach
16:49
anuejn
for the firmware?
16:49
BAndiT1983
dev -> master is what i'm used to, also at work
16:49
anuejn
maybe, but not now
16:50
vup2
sure if this suits you better then thats also okay for me, i was just trying to decrease the work you need to do
16:50
BAndiT1983
not much work there, just a PR from dev to master
16:50
vup2
anuejn: yes for the firmware and yes not now, but something to keep in mind before getting some fixed workflow
16:51
vup2
BAndiT1983: and a PR to the firmware whenever a version is tested / considerd stable
16:51
BAndiT1983
yep, that's the plan
16:52
BAndiT1983
next update will follow soon
16:52
vup2
ok cool
16:54
vup2
what happend to the problems with ninja?
16:55
BAndiT1983
it works on the beta (without) i have, also using it locally, but the beta at the hub has clock skew, which also freaks out make
16:55
BAndiT1983
*(without sensor)
16:56
BAndiT1983
you can try to execute it for the build with cmake -GNinja ..
16:57
vup2
ah okay so the problems are only with the beta at the hub?
16:57
BAndiT1983
seems so for now
16:57
vup2
ok
17:24
devar
joined the channel
17:25
devar
Hello Everyone !
17:25
BAndiT1983
hi devar
17:26
devar
I am computer science enthusiast , looking forward to contribute this organisation through GSoC
17:26
devar
what are the current project ideas !
17:27
BAndiT1983
very good, which area are you generally interested in?
17:27
BAndiT1983
please take a look at this page -> https://lab.apertus.org/project/view/20/
17:27
devar
thank you !
17:28
BAndiT1983
no problem
17:31
devar
Is there something important already discussed here ? because previous chats are not visible!
17:31
BAndiT1983
you can see previous chats on irc.apertus.org
17:32
BAndiT1983
you should specify the area, as we have discussed a lot of different topics already
17:33
devar
I will checkout resources and report back asap.
17:39
devar
left the channel
17:47
vup2
left the channel
17:47
vup2
joined the channel
17:48
Y_G
To run apertus beta on qemu I tried to clone "https://github.com/apertus-open-source-cinema/beta-software" and run "makefiles/docker_make.sh qemu-shell" ,but i continuously get the error "cpufreq: __target_index: Failed to change cpu frequency: -16"
17:48
Y_G
Any fix for it?
17:49
anuejn
huh sounds bad
17:50
vup2
Y_G: could you upload the whole log (can be found under build/build.log) somewhere?
17:50
anuejn
can you post a full log somewhere?
17:50
anuejn
ah same xD
17:59
Y_G
It's too big ,pastebin and others are cropping it out :|
18:03
BAndiT1983
changed nick to: BAndiT1983|away
18:16
Y_G
https://drive.google.com/file/d/1den0dQ1Z1s8a55p_OZalkNToJ1DbJyWO/view
18:16
Y_G
the build logs^
18:22
vup2
thanks, i will take a look at it
18:22
Y_G
Thank you :)
18:29
illwieckz
left the channel
18:37
Y_G
left the channel
18:38
sebix
joined the channel
18:50
se6astian|away
changed nick to: se6astian
18:55
BAndiT1983|away
changed nick to: BAndiT1983
18:56
se6astian
vup anuejn please take a look at this PR: https://github.com/apertus-open-source-cinema/axiom-beta-firmware/pull/61
19:33
illwieckz
joined the channel
19:40
sebix
left the channel
19:41
vup2
se6astian: will do
19:46
anuejn
done
19:46
anuejn
merged
19:47
se6astian
great, thanks
19:48
anuejn
the cirrus ci dudes are indeed very cool :)
19:50
anuejn
(the pr was a response to https://github.com/cirruslabs/cirrus-ci-docs/issues/172)
20:15
anuejn
Y_G: you should be able to login anyways. it is just a warning due to incomplete simulation
20:37
Spirit532
left the channel
20:37
Spirit532
joined the channel
20:53
BAndiT1983
changed nick to: BAndiT1983|away
21:19
se6astian
good night
21:19
se6astian
changed nick to: se6astian|away
22:41
RexOrCine
changed nick to: RexOrCine|away