So much more. It’s not even in the same ballpark.

What’s up with the abuse of the word open lately. I had a look at that project to see how they were doing the conversion, but I couldn’t find it. But I found this:

Short answer, yes! OpenScanCloud (OSC) is and will stay closed source…

Your data will be transferred through Dropbox and stored/processed on my local servers. I will use those image sets and resulting 3d models for further research, but none of your data will be published without your explicit consent!

I feel like I’d rather use Autodesk at that point. At least I know what I’m dealing with right out of the gate.

But check that it has all the features you need because it lags behind gitea in some aspects (like ci).

I mean I learned it in a few days and found it very intuitive as well. Far more intuitive than I found fusion when I tried that years later. Inventor and onshape also feel more pleasant to use.

The issue seems to be that the fusion interface is very non-standard when compared to other cad suites, so people that get used to it first find everything else unintuitive.

Podman quadlets have been a blessing. They basically let you manage containers as if they were simple services. You just plop a container unit file in /etc/containers/systemd/, daemon-reload and presto, you’ve got a service that other containers or services can depend on.

I’ve been in love with the concept of ansible since I discovered it almost a decade ago, but I still hate how verbose it is, and how cumbersome the yaml based DSL is. You can have a role that basically does the job of 3 lines of bash and it’ll need 3 yaml files in 4 directories.

About 3 years ago I wrote a big ansible playbook that would fully configure my home server, desktop and laptop from a minimal arch install. Then I used said playbook for my laptop and server.

I just got a new laptop and went to look at the playbook but realised it probably needs to be updated in a few places. I got feelings of dread thinking about reading all that yaml and updating it.

So instead I’m just gonna rewrite everything in simple python with a few helper functions. The few roles I rewrote are already so much cleaner and shorter. Should be way faster and more user friendly and maintainable.

I’ll keep ansible for actual deployments.

If this was done by multiple people, I’m sure the person that designed this delivery mechanism is really annoyed with the person that made the sloppy payload, since that made it all get detected right away.

TIL there are Linux people that don’t use OpenWRT. I always assumed everyone in the Linux community used it. It’s great.

Works great with mt7621 based routers if anyone ends up looking for something compatible.

I use gnome for the most part. I have been checking out kde recently to see how the newer versions stack up (gave up on it during the 4.0 days). As you mention kde supports dpms changes on wayland because they have their own protocol extension for that.

That’s actually my biggest gripe with wayland - the huge amount of fragmentation it has caused. I’m pretty confident that almost all the missing features I talked about are possible on one or two of the compositors, but not all of them. And definitely not on the one I use. I’m sure once some pragmatism takes hold that all the issues will be ironed out, but my plan for now is to stick to X11 until that happens.

For me it’s a million little details that just don’t work. Stuff like positioning windows, removing decorations from a window, remapping buttons on a trackball, setting a graphics output to tvrgb, disabling a display via ssh and enabling it again, etc.

It’s not just about hardware compatibility. It has to be compatible with existing workflows, and it’s currently very limiting.

set -euo pipefail at the top of every script makes stuff a lot safer. Explanation here.

Xfreerdp and gnome work really well together for me. Extremely reliable and very quick. My only complaint is lack of multi monitor support.

I remember having this realisation about Mir, but only after we collectively ran it off the cliff wall. The main reason everyone piled on Mir was that it was thought that Canonical would be priming Linux desktop for fragmentation with two competing standards.

But in fact, Mir was providing a solution to the fragmentation Wayland was bringing. Now we have 3, 4, 5 Mir-s, all with slight incompatibilities. Want a feature? Better hope all of them decide to implement the extension after someone proposes it. We know how well that worked in the past.

This is also ironic because the detractors of Xorg constantly talked about the issues with Xorg extensions and how many of them there were. But I never really had to look up which extensions Xorg supported, while I have had to do that with Wayland compositors.

I remember some 10-15 years ago when I’d look at the y windows website every couple of months hoping for some news of progress, simply because I was sick of x11 being so crappy. I hated it, it was so fiddly, it didn’t work right, I just wanted something that worked.
So you can imagine how happy I was when Wayland started taking off. Here was the promise of something better, something that just worked, it sounded amazing. And yet, today I’m still running xorg and I will be for the foreseeable future.

The reason is simply that in the time passed xorg just became usable, I don’t have to think about it, it works reliability, it has all the features I need and I hardly ever have to touch it. Meanwhile, I log into my Wayland session and instantly 3 or 4 of the applications I use daily either don’t work or act weird. I go and try and fix the issues and I’m told to just accept it, or that I actually don’t exist because Wayland works perfectly for everyone. And I’m not even using an Nvidia card, just plain Radeon.

So I quit and go back to what works. Maybe in a couple of years, until then: no thanks.

If you need this frequently, I really suggest you look into GPU forwarding. I have a Windows VM setup with a second card and it works perfectly, I use it for games and CAD all the time. Figure out your iommu groups, pop a second card in your computer (and optionally a second nvme drive if you want max performance), and use virt-manager and the arch wiki to set it up.

For accessing the machine you can use a second monitor input, or you can get a window to the machine with looking glass or moonlight. I use moonlight as it lets me play games from my laptop on the couch, and looking glass was causing windows to crash sometimes.

It’s a bit of work to set it all up but when you’re done it should just be one XML file and maybe one modprobe.d config file.

I think I’ve been using this for over a year now and the single pain point I encountered in all that time was maybe that usb input hotplug isn’t supported, though there’s ways to fix that, but I haven’t bothered.