#apertus IRC Channel Logs

off to bed now ... have a good one everyone!

morning folks!

meeting!

vup anuejn?

am also here

yes, nothing big, but still

one thing which is not discussed, how we would provide access to the hardware for students and mentors

that's it for now

there is the visualizer...

so the axiom remote could be connected via usb to zbox and controlled through UART

or UART

random buttons of course :)

but how should we trigger them?

BAndiT1983: electro shocks!

should we modify pic16 firmware and trigger it through UART to get the "full path"?

with THT buttons this is relatively trivial, with SMD buttons a little trickier though :)

of course, the second best solution is to have a side channel for the PIC16s and some 'special' firmware to generate button presses on request

easiest solution is to manipulate the array with presses/rotations

but rigging up existing hardware, e.g. building another remote without buttons for example and wiring those up to a relay card is probably feasible

all simply connected to a raspberry pi for remote control

otherwise vup is in line to report next

I worked on several things,

I tired of the open source toolchain for the zynq 7010/7020 (https://github.com/daveshah1/nextpnr-xilinx/tree/xilinx-upstream) and a simple blinky test worked, haven't tried anything more complex yet.

Finally anuejn and me worked a bit on the gateware we are experimenting with that is supposed to be more modular / easier to adapt to multiple devices compared to the current gateware and got a basic axi lite peripheral working (just some simple register). That can be found here: https://github.com/axiom-micro/nGateware/tree/axi

maybe anuejn wants to add something

nope

but that would certainly be something interesting to test

but that turned out to be rather complex (a minlp problem) and needs more time

I can imagine ... and I'm pretty sure it doesn't even cover the PLL/MMCM oddities yet

well... currently it is a rather abstract thing, where you stuff in a theoretical description of the pll

are there any paticular footguns?

Hi!

hm... yeah, no

that needs another brainstorm

i guess any integer multiple should work?

ah i see

e.g. DDR or supersampling

DDR is probably the classic example

(DDR as in double data rate not as in DDR memory :)

anything else to add to this matter?

sure

great! that is on zynq 7010, yes?

yes thats on zynq 7010

yes

(nextpnr, not vpr)

sounds great! very interested in trying ...

Ah and as I now have the beta with me, testing HDMI / verifying the HDMI plugin module is working correctly with the beta should be easy aswell

sure

anything else to add?

thanks!

sure

as you all know, we got the new power board PCBs from OSHpark, but unfortunately the oversized drill hit again ... i.e. vias are drilled with 14mil (supped to be 8/10mil) so there are problems

currently I'm working on the firmware for the power controller and testing the interfaces to the rest of the power board

on the other end, my USB3 test PC rig stopped working (doesn't boot anymore) so I have to disassemble and repair it soon

and during tests I discovered that our 'simple debayer' with imagemagick doesn't work anymore with the newer imagemagick versions

otherwise I've mostly recovered from the flu, so things should be moving again in the next weeks

https://lab.apertus.org/T1164

yep

will add lab task as well

https://lab.apertus.org/T1165

quick update from me then

we discovered minor issues at fosdem

-) optical shim pieces production method was discussed

prototype will probably be cut with scissors

there is a hackathon planned in vienna 6.4 - 12.4 roughly

gsoc next date coming up is 20.2: mentoring orgs anounced

yes

a concept for this has been written and circulated to a few people so far, will involve more people over the course of the next week or so

anyone else want to share/report?

right then I guess we can conclude todays meeting and meet again same time next week

See you next week!

i am off to the superrmarket

In the readme, the cmake .. step is giving an error :

It can be fixed by adding quotes around the STRIP line in the CMakeLists.txt like this :

* In the readme, the `cmake ..` step is giving an error :

* In the readme, the `cmake ..` step is giving an error :string sub-command STRIP requires two arguments."

`string sub-command STRIP requires two arguments`

dev

the problem is known it's due the changes in the sdl2 lib

make -j4 just tells the compiler to use 4 threads, race conditions occur very seldom, if at all, at least for me

undefined reference?

i suppose you use archlinux or derivative, there is this line for the fix -> SET(SDL2_LIBRARIES SDL2::SDL2)

> have you gat any errors while "cmake .."?

but double quotes are already there in the dev branch

sorry, formatting got lost, will do pastebin

> but double quotes are already there in the dev branch

https://github.com/apertus-open-source-cinema/AXIOM-Remote/blob/dev/AXIOM_Remote_Firmware_Visualizer/CMakeLists.txt

first argument has double quotes, second does not require any usually, also manjaro here and it runs without a problem since the fixes were done

oh, sorry I just checked. I was setting up the remote visualizer 😅

okay, cool. I'll now do this one then :)

also FirmwareTest for catch2 unit tests

no problem