Using a Raspberry Pi to add a second HDMI port to a laptop

And I'd much, much rather read at my own pace than be forced to watch at someone else's pace.

Agreed. Building within constraints (even strict, self-imposed constraints) is more fun and interesting.

[1] https://gstreamer.freedesktop.org/

All Pis (except Pico series) support both accelerated encoding and decoding, and have done so since 2012. The Broadcom VideoCore IV in RPi 3 Model B does hw decode at 1080p30.

Can I suggest adding an RSS feed to your blog? I'd love to follow it, but realistically unless I can add it to a feed reader I'm going to forget it exists!

https://hackaday.com/2021/01/08/old-kindle-shows-hdmi-video-...

I may need to look again at the parameters you have used for ffpmeg, as it appear to be almost a sorcery to get the correct parameters.

Thank you for writing it and sharing it; I hope you write more technical stuff like this in the future.

To me this is much the same as writing "As someone who is deeply afraid of their computer, this could be the silver bullet." A Pi is just a credit card sized ARM computer - there is no magic, it can run many of the same OSes as any other desktop or laptop, it can run the exact same apps and software in many cases. Thinking of it as somehow different is already a bad way to begin.

What exactly is the silver bullet here unless you were planning on doing this exact project? Seeing someone else solve a problem with the Pi? There must be thousands of instances of people solving problems with a Pi on the internet before this post. And it sounds like you're familiar with them, since you cite statistics on the kinds of problems people run into when using the Pi.

Don't be afraid, I'll give you a few examples to cheer you up and motivate you... (just make sure to backup your working Pi setup: I just dd the entire microSD card once it's working).

I've got one of my Pi who literally reached more than one year uptime. It's gently sitting, since years, in another country, constantly and relentlessly establishing reverse-SSH tunnels. Zero issue.

I ran a RasPBX on a Raspberry Pi 1 (yup, 1, a decade ago or so) for years, hooked to a SIP trunk, to provide VoIP to four Cisco VoIP phones. Zero issue. Once it was setup, I did "dd" its microSD card and kept the backup safe and that was it. I never needed them. My wife's SME's employees used that daily, for years on. Zero problem.

Regarding TFA:

> Since I was not willing to spend more money on my setup, I used a Raspberry Pi 3 which was sitting unused in one of my drawers.

That's the spirit!

Recently I decided to use one of my idle Raspberry Pi 3 as a "firewall+router" between two private LANs. It's surprisingly capable at that even if it is definitely the bottleneck: I've got fiber to the home and the Pi 3 cannot do anywhere near Gigabit ethernet, especially seen that it's only got one ethernet port and I needed two, so I had to use a USB-to-ethernet adapter... But the Pi 3 only does USB 2.0. A few firewalling rules and it's happily firewalling and routing traffic.

After reading a few comments here I also decided to configure another Raspberry Pi as a DNS, running unbound (while my main "workstation" runs dnsmasq as a cache, but the DNS is unbound on the Pi).

Someone also mentioned using a Pi Zero to transform a non-networked printer into a networked one: I'm planning to do just that for my trusty B&W laser Brother printer (haven't done it yet).

I've got a Raspberry Pi with a Pi2JAMMA adapter in my vintage arcade cab. Zero issue. (Once again: I've got a "dd" of the full microSD card should some issue show up). Every single time I turn it on, it just works.

One of my Pi is my "airline / airgapped / booting and remounting the system read-only", with X all working in RAM.

Sure, these Pis aren't advanced servers full of redundancy and ECC RAM but they cost next to nothing and they're trusty little beasts. And on that subject of these Pis not being that powerful: there's no way I'd use a Pi as my main desktop, I've got a beefy desktop and I wouldn't want to replace it with a Pi.

P.S: I'll probably be only buying Beaglebones from now on because the Pis now seems to be unobtainium.

And the part of your comment about having fun hacking on a RPi is point on!

I may or may not take the time to polish this up in the future. In the meantime, I've licensed the repo under GPL so if someone feels like doing it, they're more than welcome to.

I wonder what OP uses as a DE. In sway (and presumably other wlroots-based compositors) it's very easy (one command) to add a headless output.

Edit, as I saw your username: sadly it's undocumented in the man page, IIRC :P

I think Apple has developed its own Airplay protocol, though. There are Airplay sinks for Linux, so if you can find a way to share a screen over Airplay the setup shouldn't be too bad, especially wired.

Makind this indeed felt like hammering on a screw at times.

    Upgrade the debian to bookworm (uxplay is in the repos for that one)
    apt install uxplay
    
    in in /boot/config.txt
    change doverlay to:
    dtoverlay=vc4-kms-v3d
    force hdmi: (i had some problems with kms)
    hdmi_force_hotplug=1
    add gpu mem:
    gpu_mem=256

    then run: (-a disables audio)
    uxplay -vs kmssink -a

I don't clearly remember, but I think ffplay detects there is no X server and automatically switches to using Direct Rendering Manager. For mpv, I added a specific option to use DRM output

Such a solution has nothing at all to do with virtual monitors or display protocols though. Firewire has a standard for transporting digital video, but it's not useful for much of anything except speaking with niche and legacy AV equipment.