Current Server Time: 04:23 (Central Europe)

#apertus IRC Channel Logs

2019/01/26

Timezone: UTC


07:53
se6astian|away
changed nick to: se6astian
09:09
BAndiT1983|away
changed nick to: BAndiT1983
09:16
se6astian
changed nick to: se6astian|away
09:20
BAndiT1983
se6astian|away, could you please try to build the firmware from command line? according to the mplab forum the command is: make -f nbproject/Makefile-default.mk SUBPROJECTS= .build-conf
09:21
BAndiT1983
tried to build in docker container, but it tries to execute the compiler from your machine, still investigating how to force it to use my local commands and not yours
09:21
BAndiT1983
you can find your absolute paths in a makefile
09:26
BAndiT1983
or we can try to move the build, like vup2 and anuejn suggested, to gcc or clang
09:28
BAndiT1983
changed nick to: BAndiT1983|away
09:31
BAndiT1983|away
changed nick to: BAndiT1983
11:25
Bertl
changed nick to: Bertl_zZ
11:37
se6astian|away
changed nick to: se6astian
11:38
se6astian
Will do BAndiT1983
11:38
se6astian
just arrived at office
11:38
BAndiT1983
take your time
12:24
RexOrCine|away
changed nick to: RexOrCine
13:57
sebastian_
joined the channel
13:58
sebastian_
BAndiT1983: mplabx seems to execute the following for building: make -f nbproject/Makefile-default.mk SUBPROJECTS= .build-conf
13:58
sebastian_
that is from the AXIOM-Remote/AXIOM_Remote_Prototype_V01.X/ directory
13:59
sebastian_
current errors: menu.c:27:16: error: static declaration of 'menu_offset' follows non-static declaration
13:59
sebastian_
static uint8_t menu_offset = 0; // when scrolling the menu this is the offset for the items
14:01
sebastian_
I see the same variable being declared in menu.c and globals.c
14:01
sebastian_
thats the source of the problem I assume
14:15
BAndiT1983
remove static in front
14:16
BAndiT1983
ah, right, then just remove it from menu, as i suppose that it will be used multiple times
14:16
BAndiT1983
between menus, i mean
14:23
BAndiT1983
sebastian_, tried the build command yesterday, after reading mplab forum about building from command line, but it quits as the paths to compiler are wrong, the cause is absolute paths from your computer
14:24
BAndiT1983
it's a shortcoming of mplab stuff, will check their forums again, but the docker container received latest version of their compiler
14:40
sebastian_
left the channel
14:41
sebastian_
joined the channel
14:56
sebastian_
wouldnt the same makefile with relative paths work as well?
14:57
sebastian_
needs to be tested I assume
14:57
BAndiT1983
of course, but we have to investigate, as CI should pull the repo and build it without much hassle, at the moment it's not possible
15:13
BAndiT1983
sebastian_, please read the comment in Makefile-local-default.mk file, do me a favour, remove/rename the file and open the project again, just as a test if the file can be removed for successful system independent build
15:13
sebastian_
checking
15:14
sebastian_
I got the code to almost compile again
15:14
sebastian_
now there is just some memory related error left...
15:16
BAndiT1983
which one?
15:17
BAndiT1983
also check please if the build folder of visualizer contains now the images folder with knob image, i'Ve supplied PNG and SVG, no license problems there, as they were created by me
15:17
sebastian_
"Finally, you can also # avoid using this file at all if you are only building from the command line with make."
15:17
sebastian_
is the next line related to that?
15:18
sebastian_
images are in https://github.com/apertus-open-source-cinema/AXIOM-Remote/tree/master/axiom_remote_visualizer/images
15:19
sebastian_
compile log: https://pastebin.com/X0gzNqvL
15:20
BAndiT1983
my comment was about checking if the images folder is residing next to the executable
15:20
BAndiT1983
will check with another xc32 docker container if i can build if the file is not there
15:20
BAndiT1983
checking log now
15:20
sebastian_
ok
15:21
BAndiT1983
oha, seems like there is something wrong, as the complaints are about exceeding the regions
15:21
BAndiT1983
what is .bss
15:21
BAndiT1983
?
15:22
BAndiT1983
nevermind -> .data for initialized data and .bss for uninitialized
15:24
sebastian_
possibly related to lots of vars being declared in main.h
15:24
sebastian_
that was already removed a few commits back but got in again through the merge
15:25
sebastian_
all moved to globals.c and globals.h
15:27
sebastian_
fixed
15:27
sebastian_
compiled
15:28
sebastian_
will push now
15:30
sebastian_
done
15:31
BAndiT1983
after removing the local default file, it complains about -g not found, investigating what's the cause
15:31
BAndiT1983
but at least it now tries to build
15:32
sebastian_
great
15:32
sebastian_
will try visualizer now
15:37
sebastian_
almost succeeds: https://pastebin.com/3KcautU1
15:37
sebastian_
lots of font related problems
15:38
BAndiT1983
let me check local state
15:38
BAndiT1983
do you have them in your globals.c?
15:39
BAndiT1983
there is still one with underscore
15:39
BAndiT1983
and in .h they all have them
15:39
sebastian_
yes just see it
15:40
sebastian_
yep that solves everything
15:40
sebastian_
will push
15:41
sebastian_
done
15:42
sebastian_
visualizer works
15:42
sebastian_
colors are off though interestingly in the menu
15:47
BAndiT1983
maybe the conversion is a bit off
15:47
BAndiT1983
but i have noticed also some strange line at the top of the menu
15:48
BAndiT1983
also some dots aobve top-left button of the home page
15:51
BAndiT1983
alright, have found a fix, now i need just to fix the errors, or to pull your fixes and maybe i can get the firmware binary built
15:53
BAndiT1983
sebastian_, has the merge changed transition_framebuffer to _transitionframebuffer?
15:57
BAndiT1983
ok, docker build is working now
15:57
BAndiT1983
some adjustments are still missing, but the compiler works
15:59
BAndiT1983
please commit latest makefiles, but without the local default one please, as the compiler complains about utility.cm which is now renamed
16:01
BAndiT1983
just as a reference for anyone who is interested ->
16:01
BAndiT1983
to successfully build in docker following commands are required (but also up to date nbproject files):
16:01
BAndiT1983
export MP_CC="/opt/microchip/xc32/v2.10/bin/xc32-gcc"
16:01
BAndiT1983
make -f nbproject/Makefile-default.mk SUBPROJECTS= .build-conf
16:02
BAndiT1983
tested with docker container: marlon365/mplab-xc32-2.10:sat-xc32-2.10-fw
16:10
BAndiT1983
also for reference (to retrieve the info for docs later):
16:10
BAndiT1983
docker pull marlon365/mplab-xc32-2.10:sat-xc32-2.10-fw
16:10
BAndiT1983
docker run -it -v /home/dev/Source/AXIOM-Remote/AXIOM_Remote_Prototype_V01.X:/tmp/project marlon365/mplab-xc32-2.10:sat-xc32-2.10-fw
16:34
BAndiT1983
changed nick to: BAndiT1983|away
16:34
se6astian
sebastian_, has the merge changed transition_framebuffer to _transitionframebuffer? <- should be the other way around actually
16:35
se6astian
what makefile do you need commited ?
16:36
illwieckz
left the channel
16:51
sebastian_
Makefile-local-default.mk ?
17:20
illwieckz
joined the channel
17:27
Ashu
joined the channel
17:28
Ashu
Hello everyone .
17:34
Ashu
I want to participate in this years Gsoc . I want to know if opencine is in preference for Gsoc19 projects ?
17:36
se6astian
hi Ashu
17:36
se6astian
it seems like we will focus tasks more in directly camera related things this year, less focus on image post processing like Opencine
17:40
Ashu
hi se6astian , so what are other specific areas this year related to c++
17:46
Ashu
what are directed camera related subprojects of axiom beta ?
17:48
se6astian
the camera control daemon for example
17:48
se6astian
or the AXIOM Remote
17:52
sebastian_
left the channel
17:52
_Dev_
joined the channel
17:54
_Dev_
Hello sebastien ! , Is frameserver of opencine project is also not in preference which was incomplete last year , i was looking for that in upcoming gsoc 2k19
17:57
se6astian
frameserver might still be on the list yes
17:58
se6astian
we are still in the process of discussing it currently in the course of our own gsoc application
18:01
_Dev_
okay thanks , i was trying to figure out its action plan under supragya ,
18:03
se6astian
we should have more details soon, feb 6th at the latest as that is the org application deadline IIRC
18:05
se6astian
BAndiT1983|away: Makefile-local-default.mk is up to date, no commit required
18:05
se6astian
leaving axiom office now
18:05
se6astian
bbl
18:06
se6astian
changed nick to: se6astian|away
18:09
_Dev_
left the channel
18:28
Ashu
okay thanks sebastian , i will read about it and get back soon .
18:32
Ashu
and i was also studying and reading codebase of what TofuLynx implemented last year and thinking to continue from where he left , are there still chances that org would include that project in preference ?
18:53
se6astian|away
changed nick to: se6astian
18:59
se6astian
Ashu: if you make a good case for its inclusion very soon the chances would definitely increase :)
19:01
se6astian
what I mean with that is that you write an email to team (https://www.apertus.org/contact) outlining your conclusions from studying the code and plans for building on top of that
19:02
se6astian
if that impresses mentors its more likely to become a task for this year
19:05
Bertl_zZ
changed nick to: Bertl
19:05
Bertl
morning folks!
19:08
se6astian
hi there
19:14
_Dev_
joined the channel
19:22
_Dev_
left the channel
19:23
BAndiT1983|away
changed nick to: BAndiT1983
19:24
BAndiT1983
se6astian, about makefiles, in fact it should be the total opposite, local default makefile is specific to your mplab install and should be removed from repo, but other ones are required, as they contain the current structure of the project
19:25
BAndiT1983
that's why i've suggested to add another build platform, to be independent of makefiles from mplab, which is limiting us to the requirement that we need mplab ide on the local machine when changing the structure
19:26
BAndiT1983
my docker build fails with undefined references, as the files are up to date in the repo, but makefiles still contain old names or even miss files like globals.h and globals.c
19:29
se6astian
BAndiT1983: understood
19:32
Ashu
left the channel
19:33
se6astian
there are several other makefile related files I think
19:33
se6astian
but will remove nbproject folder from repo
19:40
se6astian
done
19:43
BAndiT1983
the files in nbproject are ok, just the local one shouldn't be there, i could build under docker, so at the moment they are required, till we can create independent makefile or cmake script
19:45
se6astian
ah, reverting
19:46
se6astian
and removed local file
19:46
BAndiT1983
thank you
19:55
BAndiT1983
will check master repo out and try to build
19:57
BAndiT1983
not important currently, but submodule is not properly connected yet, it's missing the commit hash, which points to original repo of it, will fix it later
19:57
BAndiT1983
just compare our other repos, like axiom daemon in beta firmware
19:59
se6astian
yeah something was fishy when the submodule init didnt work
20:00
BAndiT1983
please update the makefiles in the repo, they contain the old structure
20:01
BAndiT1983
you could add local default to gitignore
20:09
se6astian
left my dev machine at the hub unfortunately
20:09
se6astian
made a note so I will do it when I am there again
20:11
BAndiT1983
no problem, is not urgent, the docker container works fine, so we just need to test it at some point
20:12
BAndiT1983
but the dependency on mplab should be loosen
20:16
Spirit532
left the channel
20:19
Spirit532
joined the channel
20:30
se6astian
agreed
20:30
se6astian
btw regarding visualizer
20:30
se6astian
the knob can also be pushed (button)
21:19
illwieckz
left the channel
21:51
BAndiT1983
se6astian, noted, maybe i will add just additional button for this in the middle of the knob, to simplify the task
21:57
se6astian
that would be fine yes
22:02
BAndiT1983
so, now the knob rotation is on the list, do we need additional controls, like buttons, LEDs or similar?
22:09
se6astian
well 2 status leds are planned current for the next revision, one is installed in current hardware
22:09
se6astian
but thats probably not that critical for the visualizer
22:11
BAndiT1983
ok, adding new things is not problematic
22:31
se6astian
off to bed
22:31
se6astian
good ngith
22:32
se6astian
changed nick to: se6astian|away
22:55
BAndiT1983
changed nick to: BAndiT1983|away
22:55
BAndiT1983|away
changed nick to: BAndiT1983
22:56
illwieckz
joined the channel