| 00:17 | aombk2 | left the channel |
| 00:19 | aombk2 | joined the channel |
| 00:42 | xthenode | joined the channel |
| 00:43 | xthenode | left the channel |
| 00:49 | aombk2 | left the channel |
| 00:51 | aombk2 | joined the channel |
| 01:23 | aombk2 | left the channel |
| 01:24 | Guest27507 | left the channel |
| 01:24 | felix_ | left the channel |
| 01:24 | apoorva_arora | left the channel |
| 01:24 | uberardy | left the channel |
| 01:25 | aombk | joined the channel |
| 01:30 | Guest27507 | joined the channel |
| 01:30 | felix_ | joined the channel |
| 01:30 | apoorva_arora | joined the channel |
| 01:30 | uberardy | joined the channel |
| 01:33 | aombk | left the channel |
| 01:34 | aombk | joined the channel |
| 02:09 | aombk | left the channel |
| 02:10 | aombk | joined the channel |
| 02:36 | aombk | left the channel |
| 02:36 | aombk | joined the channel |
| 02:50 | aombk | left the channel |
| 02:50 | aombk | joined the channel |
| 03:50 | aombk | left the channel |
| 03:51 | aombk | joined the channel |
| 04:34 | aombk | left the channel |
| 04:35 | aombk | joined the channel |
| 05:50 | aombk | left the channel |
| 05:50 | aombk | joined the channel |
| 06:11 | aombk | left the channel |
| 06:11 | aombk | joined the channel |
| 06:32 | BAndiT1983|away | changed nick to: BAndiT1983
|
| 07:26 | aombk | left the channel |
| 07:26 | aombk | joined the channel |
| 07:54 | aombk | left the channel |
| 07:55 | aombk | joined the channel |
| 08:29 | aombk | left the channel |
| 08:30 | aombk | joined the channel |
| 08:50 | aombk | left the channel |
| 08:51 | aombk | joined the channel |
| 09:05 | aombk | left the channel |
| 09:05 | aombk | joined the channel |
| 09:10 | RexOrCine | joined the channel |
| 09:57 | aombk | left the channel |
| 09:58 | aombk | joined the channel |
| 10:24 | aombk | left the channel |
| 10:25 | aombk | joined the channel |
| 10:47 | Bertl_zZ | changed nick to: Bertl
|
| 10:47 | Bertl | morning folks!
|
| 10:54 | aombk | left the channel |
| 10:54 | aombk | joined the channel |
| 11:20 | aniketonroll | joined the channel |
| 11:24 | RexOrCine | left the channel |
| 11:25 | RexOrCine | joined the channel |
| 11:25 | aniketonroll80 | joined the channel |
| 11:26 | aniketonroll80 | left the channel |
| 11:26 | aniketonroll | left the channel |
| 11:27 | aniketonroll | joined the channel |
| 11:27 | aniketonroll | hey everyone, I was curious if I can contribute with the knowledge of C++ ?
|
| 11:28 | BAndiT1983 | hi
|
| 11:28 | BAndiT1983 | aniketonroll: what is your area of interest?
|
| 11:36 | aniketonroll | I'm getting started with whole tech side. I know C++ so far and intrigued towards developing and contributing.
|
| 11:38 | Bertl | off for now ... bbl
|
| 11:38 | Bertl | changed nick to: Bertl_oO
|
| 11:39 | BAndiT1983 | currently we have some active C++ development for the AXIOM Remote -> https://wiki.apertus.org/index.php/AXIOM_Remote
|
| 11:39 | BAndiT1983 | have you also check lab.apertus.org, where we are tracking tasks and issues?
|
| 11:40 | BAndiT1983 | repo: https://github.com/apertus-open-source-cinema/AXIOM-Remote
|
| 11:40 | aniketonroll | Thank you so much. will look into it.
|
| 11:41 | BAndiT1983 | if you want to try it, then please use the vis_adjust branch, as it's newer at the moment, but dev branch will be updated soon, we just had no time to merge other changes form forks etc. yet
|
| 11:41 | BAndiT1983 | no problem, if you have questions, ask anytime here, as the channel is logged on irc.apertus.org, so the messages won't be lost
|
| 11:41 | aniketonroll | so are you the mentor of this organisation?
|
| 11:43 | BAndiT1983 | i'm one of active contributors, if you mean project lead, then it would be se6ast1an
|
| 11:55 | aombk | left the channel |
| 11:55 | aombk | joined the channel |
| 12:06 | aombk | left the channel |
| 12:06 | aombk | joined the channel |
| 12:33 | aombk | left the channel |
| 12:34 | aombk | joined the channel |
| 12:42 | aombk | left the channel |
| 12:42 | aombk | joined the channel |
| 12:51 | aombk | left the channel |
| 12:51 | aombk | joined the channel |
| 13:03 | RexOrCine1 | joined the channel |
| 13:03 | RexOrCine | left the channel |
| 13:07 | xthenode | joined the channel |
| 13:07 | RexOrCine1 | left the channel |
| 13:11 | RexOrCine | joined the channel |
| 13:17 | xthenode | is there or, is there any interest in a dev board that includes a desktop dev environment?
|
| 13:20 | aombk | left the channel |
| 13:21 | aombk | joined the channel |
| 13:28 | BAndiT1983 | hi xthenode
|
| 13:29 | BAndiT1983 | could you elaborate on this, please?
|
| 13:33 | aombk | left the channel |
| 13:34 | aombk | joined the channel |
| 13:36 | xthenode | use the apertus hardware and os to build the code
|
| 13:36 | RexOrCine | left the channel |
| 13:36 | RexOrCine | joined the channel |
| 13:36 | danieeel | xthenode: how do you run vivado on arm?
|
| 13:37 | danieeel | changed nick to: danieel
|
| 13:37 | BAndiT1983 | currently the firmware is built with QEMU and infrastructure, FPGA code is additional complexity and is done separately
|
| 13:37 | xthenode | vivado?
|
| 13:39 | xthenode | this would be a special board that devs could use to build natively
|
| 13:39 | danieel | xthenode: what cpu / other spec of the board ?
|
| 13:40 | BAndiT1983 | problem with embedded boards is the performance, which is mostly low in comparison to a desktop PC or laptop
|
| 13:40 | Bertl_oO | xthenode: you can build kernel and userspace on the MicroZed, although it takes quite a while ...
|
| 13:40 | Bertl_oO | Vivado you do not want to run on a MicroZed or similar system though
|
| 13:40 | xthenode | i currently build on the pi which is arm
|
| 13:41 | Bertl_oO | (and it wouldn't make sense to be on a 'native' platform there either)
|
| 13:41 | xthenode | dont agree
|
| 13:41 | vup | BAndiT1983: most of the "firmware" (kernel, additional software) is actually cross compiled, QEMU ist mostly used for setting up the system
|
| 13:41 | danieel | for a dev/user separation, it makes no sense. But if you think of all users are devs / skilled tinkerers, i appreciate the Gentoo approach
|
| 13:42 | xthenode | cross-compiling complicates and slows dev time
|
| 13:42 | vup | (using the still very wip xilinx nextpnr support one could theoretically build some bitstreams on ARM, not sure how far one would get with our gateware though)
|
| 13:42 | BAndiT1983 | xthenode: not that much, as the real hardware is often slower
|
| 13:43 | danieel | cross-compiling is outputting a binary for other arch, it IS NOT a performance hit in any sense
|
| 13:43 | xthenode | so i guess thts a NO
|
| 13:43 | xthenode | just a suggestion
|
| 13:44 | danieel | what cpu have you suggested?
|
| 13:44 | xthenode | arm
|
| 13:44 | danieel | which ?
|
| 13:44 | xthenode | dont care
|
| 13:44 | danieel | can you do thunderx2 / centriq?
|
| 13:45 | RexOrCine | left the channel |
| 13:45 | RexOrCine1 | joined the channel |
| 13:45 | xthenode | as long as a linux dev env that generates code that con be run on the camera..
|
| 13:45 | BAndiT1983 | have used compiling on the real PI through the big visual studio via remote feature, but it was often slower than local cross-compiling
|
| 13:46 | xthenode | i use make and qt creator on pi 4. plenty fast.
|
| 13:47 | vup | well, how long does it take to compile the linux kernel on a pi4?
|
| 13:47 | xthenode | providing a cheap dev board would encourage dev for the camera
|
| 13:47 | danieel | xthenode: devs dont already have computers? or what?
|
| 13:48 | vup | I don't see how providing a dev board would make thing easier than we already have here: https://github.com/apertus-open-source-cinema/axiom-firmware
|
| 13:48 | vup | are you running into any problems with compiling / building the firmware?
|
| 13:49 | xthenode | its linux. rebuilding the os is not necessary for most system and application software
|
| 13:50 | vup | so what software are you exactly talking about, which would be compiled on the dev board?
|
| 13:50 | xthenode | it would only be for development
|
| 13:51 | xthenode | standard linux dev env. pick one.
|
| 13:52 | vup | (also the firmware is a lot more than just linux, it also compiles a lot of other software for the camera, like the control daemon, various tools used to control the camera, development tools, etc)
|
| 13:52 | xthenode | apps, daemons and drivers
|
| 13:53 | xthenode | if set-up properly, modules can be used for drivers
|
| 13:53 | vup | sure, I don't see how a dedicated dev board would help there. Either you need a full camera to do actual driver work etc, or you can just test the software on your normal computer if you don't interface with any hardware specific stuf anyways
|
| 13:53 | danieel | i would suggest to xthenode to make a benchmark, and show the improvement factor
|
| 13:53 | vup | if you already have a full camera you can just compile on the camera itself (atleast small stuff)
|
| 13:54 | vup | now if you are talking about a cheaper version of the camera that could be used for development, this is what the https://wiki.apertus.org/index.php/AXIOM_Micro is about
|
| 13:56 | xthenode | ok. looks good.
|
| 13:57 | xthenode | if it can be like a pc with desktop env. awsome.
|
| 13:57 | xthenode | price?
|
| 13:58 | vup | Well software wise it runs arch, so you can do what ever you want, no gpu though, so if you mean a graphical desktop by "desktop env", it will be painful (also no native video output, but we got a linux framebuffer via HDMI working with a HDMI plugin module)
|
| 13:58 | vup | xthenode: you can build it yourself for ~150€
|
| 14:00 | xthenode | well i would say upping this product would be a great way to promote apertus
|
| 14:01 | xthenode | all i know is the pi 4 has changed everything.
|
| 14:01 | xthenode | kinda like musk landing that rocket on its butt.
|
| 14:03 | vup | well it needs polishing work before we want to push it more
|
| 14:04 | BAndiT1983 | pi4 is good, but it's still a small-scale thing
|
| 14:04 | xthenode | big things start small
|
| 14:04 | BAndiT1983 | although the 3d stuff on it was impressive, have used it a couple of months ago for another open source project
|
| 14:05 | vup | also if you expect anything near pi 4 performance, you will get severly disappointed (as I said, no gpu and the cpu is more pi 1/2 level I think)
|
| 14:05 | xthenode | i could work on it.
|
| 14:07 | xthenode | i started in the 4mhz era
|
| 14:07 | xthenode | i know slow
|
| 14:08 | xthenode | thats why in design, less is more.
|
| 14:08 | BAndiT1983 | it's always relative, back then it was all fast
|
| 14:08 | BAndiT1983 | yes, i remember turbo button, config.sys nad autoexec.bat, himem.sys etc. ;)
|
| 14:10 | xthenode | will ever need more than 640k...
|
| 14:11 | xthenode | i appreciate the feedback
|
| 14:11 | vup | xthenode: so if you want to work on the micro, what are your areas of expertise? The gateware is written is a python DSL named nmigen, software is currently python + rust mostly, hardware is done in kicad)
|
| 14:14 | xthenode | c/c++, in Linux, MacOSX and windows, linux kernel work, lowl-level firmware using JTAG
|
| 14:14 | xthenode | SolidWorks, Autodesk and Orcad
|
| 14:15 | xthenode | microcoding of massively parallel hardware
|
| 14:16 | xthenode | cpu design using bit-slices
|
| 14:16 | xthenode | ...
|
| 14:18 | xthenode | i started just wanting to do CAD to design motorcycle parts which ended up being an odyssey
|
| 14:18 | vup | c/c++ seems like a good fit for the Remote maybe? <- BAndiT1983
|
| 14:18 | xthenode | did you see my email?
|
| 14:19 | aombk | left the channel |
| 14:20 | aombk | joined the channel |
| 14:20 | BAndiT1983 | vup: yep, i know xthenode from some years ago, but OC is dormant at the moment and the qt got crappy, regarding the license and the approach of the company
|
| 14:20 | BAndiT1983 | xthenode: yes, email is there, am just working so can only react sometimes, have to finish some stuff before the year ends
|
| 14:20 | xthenode | i just use creator for dev
|
| 14:21 | BAndiT1983 | qt company wants to slow down the releases for community, so looking for alternatives or maybe even a webapp, which uses the core
|
| 14:22 | xthenode | the important thing is the server and client libraries
|
| 14:23 | xthenode | did you like the server name idea?
|
| 14:24 | BAndiT1983 | :D after looking the meaning up again, hope it won't be a vaporware
|
| 14:24 | BAndiT1983 | just kidding, vapoursynth is also a good product out there
|
| 14:25 | xthenode | it kinda goes with apertus
|
| 14:25 | xthenode | in a whole theory of opposites way...
|
| 14:28 | aombk | left the channel |
| 14:28 | xthenode | xthenode is a web-based ui used to generate code and project files
|
| 14:28 | vup | xthenode: so there are two control daemons currently. There is https://github.com/apertus-open-source-cinema/nctrl written in rust. This is the one used on the current firmware and it works together with https://github.com/apertus-open-source-cinema/webui.
|
| 14:28 | vup | Then there is https://github.com/apertus-open-source-cinema/axiom-control-daemon, not sure what the plan is with that going forward
|
| 14:28 | aombk | joined the channel |
| 14:30 | xthenode | the actual daemon is less important than the apis and transports used to talk to it
|
| 14:31 | xthenode | the beauty of microservices
|
| 14:32 | BAndiT1983 | if there is already one which is actively developed, have just put the old one into hibernation mode for now
|
| 14:32 | BAndiT1983 | but for development we will also use simpler implementation in python, to replicate the protocols and to test the visualiser for remote quicker
|
| 14:34 | vup | We have been thinking about using flatbuffers for serialization over network, API wise its not too fleshed out, the current state of the flatbuffers API is documented here: https://wiki.apertus.org/index.php/Control_Daemon
|
| 14:35 | BAndiT1983 | flatbuffers was also used in the C++ version
|
| 14:35 | vup | The rust control daemon just provides a fuse filesystem, which the webui looks at using basic programs like `find` :)
|
| 14:35 | xthenode | ive used googles version...
|
| 14:36 | BAndiT1983 | protobuf
|
| 14:36 | BAndiT1983 | it's good, but flatbuffers is slim
|
| 14:36 | xthenode | protobuf
|
| 14:36 | vup | p sure flatbuffers is also a google thing :)
|
| 14:36 | BAndiT1983 | have also create a fix/improvement for the Go version some years ago, when we were the first time joining gsoc, but heard nothing from the maintainers
|
| 14:38 | xthenode | c++ is my forte. since the kernel is written in C.
|
| 14:39 | xthenode | node.js rocks though...
|
| 14:39 | BAndiT1983 | yep, but with SWIG you can serve the interfaces pretty good
|
| 14:39 | BAndiT1983 | kicad does it for the C++ to Python, so one can use their api from python, works rather good, docs are meh, so mostly googling for examples or reading source code
|
| 14:42 | xthenode | to make things simple where speed is not a problem, i use cgi.
|
| 14:44 | xthenode | so is json to slow for transport?
|
| 14:45 | xthenode | sony uses it for their apis
|
| 14:45 | vup | I would like a transport with a strong schema, which json doesn't really provide on its own.
|
| 14:46 | BAndiT1983 | JSON can have a string schema, using it at work a lot, but flatbuffers is much smaller, which saves bandwidth
|
| 14:46 | BAndiT1983 | also deserializing there is not that CPU-heavy
|
| 14:46 | BAndiT1983 | *strong
|
| 14:46 | vup | I guess if you do JSON + a schema I don't see why one would not do something like flatbuffers
|
| 14:47 | BAndiT1983 | exactly
|
| 14:47 | xthenode | json = xml kinda
|
| 14:47 | BAndiT1983 | we have to do quite some validation, as the JSON data is deserialized to Java objects etc., so in the background Spring Boot is doing quite some schema checking
|
| 14:47 | BAndiT1983 | JSON is much smaller, as it's doesn't repeat the tags, like XML does
|
| 14:47 | xthenode | libjson...
|
| 14:48 | _bluez | joined the channel |
| 14:48 | xthenode | json IS better for transport and encoding. i agree.
|
| 14:49 | BAndiT1983 | and i just remember my current struggle with SOAP (XML), then i would prefer flatbuffers, then JSON and XML at last
|
| 14:49 | xthenode | my point was if you can have an xml schema then also a json one
|
| 14:49 | vup | xthenode: json and xml are very different, xml has builtin support for for schemas / validation, json does not have that
|
| 14:49 | danieel | or you can go ASN :)
|
| 14:50 | vup | :)
|
| 14:50 | xthenode | thats why node is so cool.
|
| 14:52 | xthenode | i would say. focus on the api
|
| 14:54 | BAndiT1983 | xthenode: then you should look at OpenAPI/Swagger ;)
|
| 14:57 | xthenode | similar to the rest api generator i did for verizon.
|
| 14:58 | xthenode | i started working on a json based one for the sony api
|
| 15:00 | xthenode | might want to consider a web-based one for apertus
|
| 15:03 | xthenode | left the channel |
| 15:05 | aombk | left the channel |
| 15:06 | aombk | joined the channel |
| 15:06 | BAndiT1983 | what do you mean by "web-based one"?
|
| 15:14 | aombk | left the channel |
| 15:14 | futarisIRCcloud | left the channel |
| 15:15 | aombk | joined the channel |
| 15:52 | _bluez | left the channel |
| 16:08 | aombk | left the channel |
| 16:08 | aombk | joined the channel |
| 16:16 | RexOrCine1 | left the channel |
| 16:17 | RexOrCine | joined the channel |
| 16:24 | RexOrCine | left the channel |
| 16:27 | RexOrCine | joined the channel |
| 16:30 | aombk | left the channel |
| 16:30 | aombk | joined the channel |
| 16:36 | RexOrCine | left the channel |
| 16:43 | RexOrCine | joined the channel |
| 16:47 | RexOrCine | left the channel |
| 16:55 | aombk | left the channel |
| 16:56 | aombk | joined the channel |
| 16:59 | RexOrCine | joined the channel |
| 17:06 | aombk | left the channel |
| 17:07 | aombk | joined the channel |
| 17:14 | RexOrCine | left the channel |
| 17:16 | _bluez | joined the channel |
| 17:19 | _bluez | left the channel |
| 17:19 | _bluez | joined the channel |
| 17:24 | aombk | left the channel |
| 17:25 | aombk | joined the channel |
| 17:36 | RexOrCine | joined the channel |
| 17:36 | aombk | left the channel |
| 17:37 | aombk | joined the channel |
| 17:48 | RexOrCine1 | joined the channel |
| 17:48 | RexOrCine | left the channel |
| 18:00 | RexOrCine1 | left the channel |
| 18:01 | RexOrCine | joined the channel |
| 18:10 | _bluez | left the channel |
| 18:16 | aombk | left the channel |
| 18:17 | aombk | joined the channel |
| 18:36 | aombk | left the channel |
| 18:36 | aombk | joined the channel |
| 18:47 | aombk | left the channel |
| 18:47 | aombk | joined the channel |
| 18:54 | aniketonroll | left the channel |
| 19:03 | aombk | left the channel |
| 19:04 | aombk | joined the channel |
| 19:11 | xthenode | joined the channel |
| 19:12 | xthenode | left the channel |
| 19:28 | BAndiT1983 | changed nick to: BAndiT1983|away
|
| 19:31 | BAndiT1983|away | changed nick to: BAndiT1983
|
| 19:46 | illwieckz | left the channel |
| 19:51 | illwieckz | joined the channel |
| 19:58 | BAndiT1983 | changed nick to: BAndiT1983|away
|
| 20:19 | aombk | left the channel |
| 20:20 | aombk | joined the channel |
| 20:29 | aombk | left the channel |
| 20:30 | aombk | joined the channel |
| 21:11 | aombk | left the channel |
| 21:11 | aombk | joined the channel |
| 21:34 | xthenode | joined the channel |
| 21:35 | aombk | left the channel |
| 21:35 | xthenode | left the channel |
| 21:35 | aombk | joined the channel |
| 22:20 | aombk | left the channel |
| 22:20 | aombk | joined the channel |
| 22:37 | aombk | left the channel |
| 22:37 | aombk | joined the channel |
| 22:55 | se6ast1an | off to bed
|
| 22:55 | se6ast1an | good night
|
| 22:58 | Bertl_oO | nn
|
| 23:31 | aombk | left the channel |
| 23:40 | aombk | joined the channel |