Current Server Time: 20:33 (Central Europe)

#apertus IRC Channel Logs

2018/02/21

Timezone: UTC


00:06
TofuLynx
left the channel
00:06
rton
left the channel
02:26
slikdigit
joined the channel
03:12
futarisIRCcloud
left the channel
04:06
slikdigit_
joined the channel
04:07
slikdigit
left the channel
04:07
slikdigit_
changed nick to: slikdigit
05:10
davidak
i got some work done :)
05:10
davidak
https://lab.apertus.org/T827#13611
05:24
futarisIRCcloud
joined the channel
06:17
intrac
left the channel
06:19
intrac
joined the channel
06:36
slikdigit
left the channel
06:39
Bertl_zZ
changed nick to: Bertl
06:39
Bertl
morning folks!
06:46
davidak
good morning Bertl
06:51
Bertl
will upload the missing files soon
06:51
Bertl
git repo as working dir is not really feasable for several reasons
06:53
Bertl
*feasible
07:09
se6astian|away
changed nick to: se6astian
07:15
sebix
joined the channel
07:15
sebix
left the channel
07:15
sebix
joined the channel
07:15
davidak
good morning se6astian
07:16
davidak
do you know who initially wrote the octave scripts?
07:16
davidak
https://github.com/apertus-open-source-cinema/misc-tools-utilities/blob/master/color-calibration/main.m
07:24
davidak
was it also a1ex?
07:32
rton
joined the channel
07:56
anuejn
Bertl: why is using the git repo as working dir not an option for you? it would make it easier for others to understand your work too...
07:59
seku
left the channel
08:02
Bertl
well, one reason is that I have work I do for customers (including closed source) there as well ... another reason is the huge number of incomplete designs without proper copyright and/or license (i.e. internal ones)
08:03
Bertl
so once a design becomes 'usable' and is 'tested' it will be cleaned up, properly licensed and published
08:03
futarisIRCcloud
left the channel
08:08
anuejn
ok but the designs for the axiom project are all open source?
08:09
anuejn
and whats the problem with having untested stuff in a git repo too and just mege them into master, once they are tested?
08:10
anuejn
moreover, whats the problem with having pre versions licensed with gpl/ohl too and publish them on dev branches?
08:14
davidak
release early, release often ;)
08:17
Bertl
the problem is mainly that it adds a lot of additional work to my already huge workload ...
08:18
Bertl
if you feel like this is important for you, send me money on a regular basis to do the extra work and I'll gladly do it
08:18
davidak
sure, but it makes contributing easier and the published files are always up to date
08:21
Bertl
it might make it easier, I'm not really convinced there though
08:21
Bertl
having untested designs out in the wild can cause a lot of harm as well, especially with hardware
08:22
davidak
thats why it is important to mark working versions and have seperate development branch
08:23
davidak
currently we don't have a lot contributors, so it might not be worth the extra effort.
08:23
davidak
but maybe it's why we have not more contributors
08:25
Bertl
I think hardware is quite different from software regarding contributions ...
08:26
Bertl
for example, the plugin interface is set in stone since the beginning i.e. it has not changed the last two or three years
08:27
Bertl
it is simple and there is a lot of potential for creating your own
08:27
Bertl
(plugin that is)
08:28
davidak
i think a reason for that is that only very few people own an AXIOM Beta
08:28
Bertl
still we haven't had a single person (outside of apertus) creating a published plugin
08:29
Bertl
well, it might be related to that, but I also think it is realted to the fact that hardware needs to be designed, built and tested (similar to software) but the cycles there are much much larger and the cost is a lot higher
08:29
Bertl
s/larger/longer/
08:30
davidak
would it be possible for you to work in the git repo on AXIOM Beta PCBs and commit when a revision is ready
08:30
Bertl
so for example, if you had a new 'known not working' version of the power board every day, what help would it be?
08:31
davidak
and push to github. might be even faster than uploading to the fileserver?
08:31
davidak
and updating the wiki
08:31
Bertl
it would take you a day or two to figure out what is not working and why and at that point, the design already has changed
08:32
davidak
makes sense
08:32
Bertl
there is also not much point in having releases which get never built or tested online
08:33
Bertl
we had 36 major revisions on the main board now, but only built 30 cameras or so
08:33
Bertl
what would be the use of those revisions which were sorted out (because of flaws in the design)?
08:34
Bertl
so that somebody can build a non working main board?
08:35
davidak
when you notice a problem and have it in the last 5 revisions, it's good to have version control to look what changes there
08:35
Bertl
did you try to check the changes between two minor revisions with git?
08:35
davidak
no :D
08:36
Bertl
please do that (no need to check major revisions to see my point there)
08:36
davidak
i saw that it is some text format
08:36
Bertl
it is pure XML
08:36
davidak
but don't know if the diffs are helpful
08:36
Bertl
so no problem to make a diff, but the result is typically useless
08:37
Bertl
which also doesn't make it very suitable for revision control
08:38
Bertl
as I already pointed out, I'm perfectly fine with having all the revisions we built an tested in a git repository, as long as somebody takes care of that
08:39
Bertl
I'm also happy to clean up, check and upload those files :)
08:39
davidak
it don't makes sense to maintain two places where the files are
08:40
davidak
so in your expereance just having a file for every revision on a fileserver works best for hardware development?
08:42
davidak
it might help to just clean up the fileserver, have a folder for every PCB and module
08:42
davidak
only the current version in it and a archive folder with the older versions
08:43
davidak
the current structure in the wiki makes it quiet easy to find the files
08:47
Bertl
I'm fine with having a 'cleaned up' file repository somewhere
08:48
Bertl
I don't have a problem putting files there as long as a simple rsync/scp can do that
08:49
Bertl
I do not suggest to move stuff around when a new version is released, because this tends to break internal and external links
08:49
se6astian
davidak: pretty sure all octave scripts were created by alex
08:49
Bertl
(which is why I do not 'reorganize' the stuff on vserver.13thfloor.at
08:49
Bertl
)
08:55
davidak
se6astian changed
08:56
se6astian
thanks
09:04
davidak
Bertl could we break links once to organize files on folders?
09:05
davidak
put every axiom_beta_main_board_* in /Stuff/AXIOM/BETA/MAIN_BOARD/
09:06
davidak
you argued that old versions are not useful when better versions are available, so broken links should not be critical
09:06
davidak
when it is critical, 301 redirects could be setup
09:07
davidak
then we can delete the beta-hardware repository from github
09:27
Bertl
how about this: http://vserver.13thfloor.at/AXIOM/
09:30
davidak
thats even better
09:31
davidak
could we use files.apertus.org?
09:31
Bertl
sure, no problem with that
09:31
davidak
ah, that already exists :D
09:32
davidak
fat-lady.jpg :D
09:32
Bertl
you know the saying: 'It ain't over till the fat lady sings'?
09:33
davidak
no
09:33
Bertl
https://en.wikipedia.org/wiki/It_ain%27t_over_till_the_fat_lady_sings
09:35
davidak
i see
09:41
Bertl
so, where would you like to have this structure once and for all :)
09:49
davidak
what about http://files.apertus.org/HARDWARE/AXIOM/BETA/MAIN_BOARD ? se6astian
09:56
Bertl
btw, just saw it on the wiki, the 'Dummy Interface Board' is called 'Interface Dummy Board' which might even be more correct but is probably a little confusing too
09:58
Bertl
off for now ... bbl
09:58
Bertl
changed nick to: Bertl_oO
10:17
TofuLynx
joined the channel
10:18
TofuLynx
Hey! Regarding GSoC, do you guys consider optimization important in the C/C++ challenge?
10:22
se6astian
davidak looks good
10:23
davidak
ok, then we will do it like that
10:23
davidak
after Bertl moved the files, i will update the links in the wiki
10:26
se6astian
thanks
10:42
TofuLynx
left the channel
10:45
TofuLynx
joined the channel
11:06
se6astian
TofuLynx: optimization is not important for the C/C++ challenge
12:33
TofuLynx
Okay :)
12:34
TofuLynx
Is it possible to know who are the potential mentors?
12:38
Bertl_oO
davidak: done
12:39
Bertl_oO
TofuLynx: the mentors are listed on each task
12:40
Bertl_oO
(not the challenges, the actual tasks)
12:41
TofuLynx
Ah!
12:42
davidak
great. thanks!
12:42
TofuLynx
who is alex?
12:43
Bertl_oO
https://www.apertus.org/magic-lantern-getting-to-grips-with-axiom-beta-image-sensor-article-feb-2016
12:43
Bertl_oO
davidak: you're welcome!
12:43
TofuLynx
Oh! I remember him!
12:44
Bertl_oO
davidak: let me know what you are missing there, I'll try to add it ... note that the BOM needs to be manually cleaned up so that will take some time
12:45
TofuLynx
also
12:45
TofuLynx
linearization is responsible for the image's true to life colors, right?
12:46
Bertl_oO
yes, but it is a little more complex than just that
12:46
TofuLynx
and makes the image overall brighter, right?
12:46
TofuLynx
so I don't need to worry about my image being a little dark?
12:46
Bertl_oO
yep, that's because you don't have a gamma applied
12:47
TofuLynx
Ok :)
12:49
sebix
left the channel
13:13
aombk
left the channel
13:34
TofuLynx
left the channel
13:40
aombk
joined the channel
14:02
davidak
left the channel
14:31
TofuLynx
joined the channel
14:32
davidak
joined the channel
14:52
slikdigit
joined the channel
14:52
slikdigit
left the channel
14:56
hozer
joined the channel
14:58
se6astian
changed nick to: se6astian|away
15:04
BAndiT1983|away
changed nick to: BAndiT1983
15:26
TofuLynx
Bandit, I have some questions. You are the mentor of the T722 task, I would like to get to know what you do as a mentor and stuff
15:27
TofuLynx
and how does the accepted candidate, which I hope to be, work, in his house? Does he have to travel somewhere sometimes?
15:27
BAndiT1983
i guide the student, give hints and check the work
15:28
BAndiT1983
gsoc does not require travel
15:28
BAndiT1983
every month there is a meeting between mentors to grade the progress
15:28
TofuLynx
and how does the grading works?
15:29
TofuLynx
is it like from 0 to 20
15:29
TofuLynx
or simply approved or not?
15:29
BAndiT1983
more like approved or not, but trying to give chances, if we think that the student could improve
15:30
TofuLynx
Ok :)
15:31
BAndiT1983
have you already completed the C++ challenge?
15:31
TofuLynx
and yeah
15:31
TofuLynx
yeah*
15:32
BAndiT1983
is the clipping problem gone?
15:32
TofuLynx
the two strange layers? Yep, gone
15:33
rton
left the channel
15:33
BAndiT1983
sounds good
15:34
TofuLynx
Where do I submit the application when it opens?
15:34
BAndiT1983
have to discuss with se6astian etc., when we would look through submitted stuff
15:34
TofuLynx
Ok :)
15:34
BAndiT1983
there is a gsoc page for that, you just have to sign in with your gmail account
15:35
rton
joined the channel
15:35
BAndiT1983
i suppose that it will be there ->
15:35
TofuLynx
hmm, and how do I submit the C/C++ challenge code? with the application?
15:35
BAndiT1983
summerofcode.withgoogle.com
15:36
BAndiT1983
you could upload it to some github repo
15:36
BAndiT1983
or in lab
15:36
TofuLynx
yeah, I discovered apertus from there
15:36
TofuLynx
Ok! :D
15:37
BAndiT1983
as we are looking through each submission manually, it is harder to cheat, if anyone tries to copy the code
15:38
BAndiT1983
every developer has his own style and it's recognizable, we will also ask questions to see if the application was really written by the same student
15:39
TofuLynx
here in the chat or by email?
15:41
BAndiT1983
depends on the situation, but it think some private chat room, it's not about embarassing the people
15:42
TofuLynx
Ok!
15:47
TofuLynx
Is it the first time Apertus is on GSoC?
15:47
BAndiT1983
no, second one
15:48
TofuLynx
And did some of the students still continue to contribute after it?
15:49
BAndiT1983
can't say for sure, but niculescu_vlad is still on the IRC here
15:49
BAndiT1983
haven't seen others since gsoc
15:50
Bertl_oO
well, the two other students didn't finish GSoC 2017
15:50
TofuLynx
ah
15:50
TofuLynx
they gave up?
15:50
Bertl_oO
one did kind of give up, the other was 'given up' :)
15:51
TofuLynx
oh Ok!
15:51
BAndiT1983
one wasn't up to the task, don't know about the other one
15:52
Bertl_oO
he did some good work, but at the end, he had no time for GSoC because of his exams
15:53
BAndiT1983
yeah, gsoc requires a lot of time, from both sides
15:54
TofuLynx
Well, May is exam time for me too, but my university will host the Olympic University Games, so it can save me time
15:54
TofuLynx
as the exams are being done earlier
15:55
BAndiT1983
with right preparation and time management, gsoc is doable
15:55
BAndiT1983
but it's no walk in the park, had to learn and prepare a lot before being able to do mentoring
15:56
BAndiT1983
students have to do the same
15:56
TofuLynx
I see
15:58
TofuLynx
Everyone learns from the GSoC experience
15:58
TofuLynx
the students, mentors and the organization
15:59
BAndiT1983
there is always a hope that students would stay with the project after gsoc
15:59
BAndiT1983
but not many people want to work for free
16:01
TofuLynx
yeah I understand that
16:02
TofuLynx
but as GSoC says on its page, it's FOSS organizations
16:07
BAndiT1983
to be honest, a lot of students are doing gsoc because of money
16:09
TofuLynx
Yeah, I see
16:14
BAndiT1983
can't blame them, earning money as student is not easy, it depends on the country also, it's just a little bit annoying when someone tries to attend but has not much knowledge, which is a waste of time for both sides
16:18
TofuLynx
but weren't there a lot of applicants on apertus?
16:19
BAndiT1983
yes, we had a long selection process, but it was our first time
16:19
BAndiT1983
so the guys came up with challenges for this year to see if candidates are suitable
16:21
TofuLynx
Makes sense!
16:23
TofuLynx
was it frequent to appear applications from students that never talked with you before?
16:24
BAndiT1983
most students were present in the chat, from time to time or just few times
16:24
BAndiT1983
but if i'm not mistaken, then there were applications from unknown ones
16:25
BAndiT1983
but they had not the right volume of information, we had a template and questions, so most of them were not considered
16:26
TofuLynx
Ok! :)
16:26
TofuLynx
Thanks for answering me the questions!
16:27
BAndiT1983
no problem, maybe also others want to know that stuff, as it's FOSS organization ;)
16:28
TofuLynx
:D
16:59
TofuLynx
The GSoC student guide says that the application platform varies from organization to organization. Can I put images in the application platform of Apertus?
17:03
BAndiT1983
the application is usually a PDF file, so the images can be placed there
17:07
seku
joined the channel
17:15
se6astian|away
changed nick to: se6astian
17:29
TofuLynx
Bandit, did you try to reproduce the issue?
17:29
BAndiT1983
not yet, still not successful?
17:30
TofuLynx
yeah, havent tried again today though
17:30
BAndiT1983
alright will try it now
17:44
davidak
left the channel
17:45
sebix
joined the channel
17:45
sebix
left the channel
17:45
sebix
joined the channel
17:45
BAndiT1983
TofuLynx, if your feel ready, then you can upload your challenge code to github or similar
17:58
TofuLynx
Ok! Just making sure
17:58
TofuLynx
I don't need to worry about linearization neither gamma applying, right_
17:58
TofuLynx
?*
18:02
BAndiT1983
no, nothing of the advanced stuff, no performance optimizations
18:02
BAndiT1983
this things would be important for gsoc maybe, but not for the challenge
18:03
TofuLynx
Ok!
18:04
TofuLynx
I really liked learning while making the Challenge
18:07
micha__
joined the channel
18:26
TofuLynx
Bandit, were you successful with the beta firmware?
18:30
BAndiT1983
still testing, my VM is a bit broken, let me re-install it quick
18:49
anuejn
Bertl: davidak: thanks a lot :)
18:49
anuejn
this is a quite good organized form
18:50
anuejn
if this is easier than git for you bertl, than it ia equally fine
18:50
anuejn
much love <3
19:12
Bertl_oO
perfect!
19:17
micha__
left the channel
19:28
BAndiT1983
TofuLynx, have you also installed the packages? -> g++-arm-linux-gnueabi and gcc-arm-linux-gnueabi
19:33
BAndiT1983
also boxes is required in fresh linuxmint, but there is a general problem, that some sub-modules can't be pulled
19:54
se6astian
off for today
19:54
se6astian
good night
19:54
se6astian
changed nick to: se6astian|away
19:56
TofuLynx
yes I installed. But that step isn't where I couldnt overcome the issues
19:56
TofuLynx
It was at the step 4
19:57
BAndiT1983
seeing a lot of issues while trying to install, tried on linuxmint and manjaro
19:58
TofuLynx
yeah :/
19:58
seku
left the channel
19:58
BAndiT1983
now my manjaro is broken, probably modification of bashrc
19:59
TofuLynx
my fix for it was by using echo
19:59
TofuLynx
instead of copy and pasting it directly
20:00
BAndiT1983
what do you mean by that?
20:00
TofuLynx
by using the echo command, it changes the path
20:01
BAndiT1983
you mean temporarily?
20:01
TofuLynx
no no
20:01
TofuLynx
wait
20:01
TofuLynx
I will explain it
20:01
TofuLynx
I have a example
20:02
TofuLynx
Ok
20:02
TofuLynx
the ./install.sh
20:03
TofuLynx
said at the end to copy and paste this on the bashrc
20:03
TofuLynx
export PATH=$PATH:/home/claudio/Desktop/ApertusTesting/axiom-beta-qemu/external/gcc-linaro-4.9-gnueabi/bin
20:03
TofuLynx
I copy pasted it on the bashrc and didnt work
20:03
TofuLynx
then I used echo, and it wrote this:
20:03
TofuLynx
export PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games:/$
20:04
TofuLynx
which is a lot different
20:04
TofuLynx
but fixed.
20:04
BAndiT1983
usually you should prepend it with $PATH: so the previous one is stored
20:05
TofuLynx
well
20:05
TofuLynx
it was my fix
20:06
TofuLynx
and the issue I cant overcome in the step 4 is related with the patch file
20:07
TofuLynx
and I have no knowledge on how patch files work
20:07
BAndiT1983
the process really broke my linux and keyboard does not work in emergency shell
20:08
TofuLynx
wow :O
20:08
BAndiT1983
the patch files are just applying diffs of the code
20:08
BAndiT1983
luckily i use a VM, so not a big problem, but the issue is also possible on physical machines
20:09
TofuLynx
Hmm
20:09
TofuLynx
it may be the reason it broke my linux too
20:09
TofuLynx
but not sure
20:09
BAndiT1983
have you tried my suggestion with deactivating patch files?
20:09
TofuLynx
Hmm I dont think so
20:10
TofuLynx
Also, by the time I was trying to install the firmware, I had to repartition the disk so linux could have more space (was dual booting with windows), after it, it broke linux and I had to reinstall it
20:10
TofuLynx
now you said it, not sure if I blame windows
20:10
BAndiT1983
the patch files could be obsolete, because they were added 3 months ago and xilinx maybe changed the related code already
20:11
TofuLynx
I will check it!
20:12
BAndiT1983
that's why i use virtualbox, also because of manufacturer recovery partition, don't want to mess it up at the moment
20:12
TofuLynx
I understand
20:20
TofuLynx
Bandit
20:21
BAndiT1983
?
20:21
TofuLynx
I have uploaded my code
20:21
TofuLynx
for the challende
20:21
TofuLynx
to the repository
20:21
TofuLynx
I am sending the link
20:21
BAndiT1983
alright
20:24
TofuLynx
https://github.com/TofuLynx/apertus-CppChallenge
20:25
BAndiT1983
will store it in the lab, so it doesn't get lost
20:25
TofuLynx
Thanks :)
20:25
BAndiT1983
have you registered there already?
20:25
TofuLynx
yeah
20:25
TofuLynx
TofuLynx too
20:26
TofuLynx
Do you have any idea when I will get feedback about it?
20:26
BAndiT1983
i hope soon, have to sit down and take a look but together with some others
20:27
TofuLynx
Ok! :)
20:27
futarisIRCcloud
joined the channel
20:28
BAndiT1983
so, added your link to the desccription, you should get a notification from the system, because of name linking
20:28
TofuLynx
Got it!
20:41
sebix
left the channel
21:15
BAndiT1983
TofuLynx, have you looked at https://lab.apertus.org/T737 already?
21:15
BAndiT1983
it was the beginning of QEMU prototype
21:18
BAndiT1983
here is the SD image ->
21:18
BAndiT1983
http://vserver.13thfloor.at/Stuff/AXIOM/BETA/
21:18
BAndiT1983
just sort for size, it's one with 3.6G
21:25
BAndiT1983
just be aware that it is rather old one, but shouldn't be a problem to try the firmware out
21:26
BAndiT1983
will try to discuss with se6astian and Bertl_oO further approach tomorrow, as we need to correct this scripts
21:26
BAndiT1983
changed nick to: BAndiT1983|away
22:17
TofuLynx
I will check it
22:26
TofuLynx
left the channel
22:37
futarisIRCcloud
left the channel
22:51
Bertl_oO
off to bed now ... have a good one everyone!
22:51
Bertl_oO
changed nick to: Bertl_zZ