I’ve only managed it accidentally.

But usually the pattern is contacting/getting contacted due to some matter that needs discussing. The source for that can be needing some pointers on using someone elses code, or providing feedback on how someone moderates, etc.

For me it’s usually come from participating more extensively. Engaging in community building such as my modding titanfall and hosting northstar servers. Starting a bunch of lemmy communities.

Then, once something gives you a reason to go a couple back and forths into an exchange, and there is mutual interest, it just keeps going. After that you can pick up with the same skillset you’d use with someone you’ve met in person.

The difference is that you’ll need to keep it going by deliberately reviving the conversation now and then. That won’t happen by running into the person face to face or some similar context the way it might work with a coworker or someone in your social circle.

Orange cat, only in the brain.

Yes.

Yes.

Absolut yes.

Completely immersive in-world UI with Alien Isolation vibes and WORLD TIER audio design. No hand-holding, almost no “yellow paint”, instead everything is done in realistic in-world context.

It’s about 5 hours to complete if you’re fast and care about that. But it’s a 5 hours with IMPECCABLE attention to detail.

Note, #4 will not apply to files MOVED into the directory. Only files that count as “created” there (like a copy operation).

No.

Assuming you’re accessing your files via network mounts, those file are being created by the relevant SMB/NFS/SSH daemon. ALL FILES are created by some process or other. If you click “new file” in Dolphin, that process is Dolphin.

Which means there is a umask you can configure.

My network SMB mount that leads to my jellyfin libraries is configured (on the server) to access the filesystem under my user and the shared media group. Hence files it “creates” are always done so with correct permissions. User, group and umask are configurable for each share.

This way, I can copy files from a share containing “my” files, into the media share, and have the owner, group and umask automatically “translate” into what they should be in either direction. Theres not really a way to do that directly on the system. It’s something you can only really set up when moving files around between the shares and hence their respective environments.

If I did the move in the terminal directly on the host system, shit would instantly break.

Files and directories can only have one owner for user and group, each. If files you create belonged to every group you are in, that’d be a permission nightmare, so by default, the group permission is “unused” (it just gets set to the same as your user).

You could add the application that needs to access files created by you to your user group, but that’s a lot more permissive than it needs to be (it gets to access all your files with group access granted).

If you’re managing files via network shares, the solution is in their settings.

If you’re managing files on the system directly, that’s trickier. New files can be made to inherit the group of the parent directory (edit: other commentor), which solves your problem only some of the time, as it won’t affect files you move into the folder. In those cases you still have to fix the group after moving. That, or copy instead of move.

Bottom line: You’re not really supposed to manage this stuff directly. When you go in and create files directly on the host system, you are bypassing the systems that would normally set permissions inside a share or directory that is being accessed by multiple applications, users, and services.

Making the default of files created in such a way non-permissive, is by design. That you have to manually correct the group permissions on a file you just moved in from another permission environment, is as it should be.

No, you are not expected to cronjob the chmod command.

You want “umask”. That’s the term, you can google it for more info.

Basically, it’s the setting that controls what the default permissions are for the files created by a process.

Jellyfins umask is configurable using the systemd unit file. Arr stack umasks are somewhere in their settings… Etc.

It is process specific, and something you configure for each application. Owner is whatever user runs the process, and same goes for group.

Group does not need to be same as user. You can run jellyfin with a “jellyfin” user and “media” group, so that some other app managing the media files can deal with jellyfins files by running under a different user but the same “media” group.

As long as you set their umasks to allow group access they will then always be able to pass files and directeries between them.

I’ve no experience with ACLs. I’ve used SMB, and SSHFS to a lesser extent, and always with very simple setups.

My user accessing my own users files via SMB, and ssh only via the admin user and as such with root level access.

But generally, unix-style permissions should allow you to be able to set things up in whatever way you might need. Getting there can be complex, but there is always a way to configure what permissions directories and files will have when created, and hence what users/groups can access them.

Also, sometimes the answer is to edit the groups. Such as adding a user being used to access a share to the relevant groups, instead of messing with owners or groups of existing files and directories.

Or, creating a whole new group (like “pictures” for your camera roll thing) which you add all the relevant users to (yourself, and whatever apps will be accessing the relevant directories). Plus then configure umasks so that files actually default to allowing group level access.

Some Lenovos only allow you to boot off USB via the “novo” button.

It’s a pin-tool reset hole with a “U” shaped icon next to it. I’ve seen it hiding among the cooling intake holes on some models.

Turn the laptop off, hit the novo button, and the laptop boots into a separate menu from the bios that among other things allows you to get to a boot device menu.

I’ve also come across a case where secure boot would just refuse to boot stuff that didn’t already have matching keys in the bios. Disabling secure boot, everything worked, but turning it back on shit immediately broke. In that case it had stopped booting its NORMAL install of W11 because it had become too new. I had to discover that windows ships a tiny EFI executable signed with the old keys whose sole purpose is to slip the new keys into the BIOS for exactly such a situation.

It’s also possible to circumvent the online requirement if you know how. The key combo is shift+f10 (or shift+fn+f10 depending on how the function keys are bound) to bring up cmd, after which you can run OOBE\BYPASSNRO to reboot into a mode that allows you to skip ms account setup.

If you create the bootable USB with Rufus (which you unfortunately need windows to run 🫤), it boots into that mode from the start.

You may want to try to deliberately get away from the “unreal fivey” look.

I like the underlying style but the strong reflections and resulting high contrast image really hits that “unreal five” look in a bad way, which I suspect will turn a lot of people off.

It’s unfortunate but “indie unreal five game” is the new “steam asset flip”, in that it is something that makes people skip a game. It’s easy to slap realistic lightning into games now, so it has stopped being an indicator of quality, and slowly become the opposite.

Try to find something that works with the games style that you can do to the shading or post processing that sets you apart.

I might mute reflections a bit and add at least some grime and imperfections to surfaces. The cleanliness of surfaces together with unreals realistic lighting is a big part of the “indie unreal five” look.

Yeah so that dude pretty clearly decided that driving made his right to the road absolute, and that cyclists encroaching on that right are deserving of death, or at the very least bodily harm and property damage.

It is terrifying how many drivers, when told they make pedestrians fear for their lives, react with “good, they should be careful”.

It does work.

If you have two EFI partitions, Windows will stop looking once it finds the first one.

Which is actually a problem sometimes, as it will stop looking and use the first one it finds, even if it’s on different drive than windows.

Imagine installing windows on a machine with two drives. Drive A already has an OS and an EFI partition, so you install Windows to drive B.

Except Windows will still go ahead and drop the bootloader onto drive A. Potentially breaking the OS that’s there, plus setting up Windows in such a way that it requires drive A to boot. EVEN THOUGH THAT ISN’T WHERE YOU INSTALLED IT.

This is why guides tell you to install windows with only one drive connected. It’s the only way to ensure it puts the bootloader and OS on the same drive.

That certainly helps.

In my case the solution is two EFI partitions.

Windows finds the first and doesn’t bother looking for a second one, which happens to contain grub and my efistubs.

Windows will go ahead and clear the UEFI menu sometimes, but manually booting an efi file and then re-adding the grub and stub boot entries is small compared to having my stuff actually deleted.

The water comes out at a single point at high pressure below the sphere.

It then travels out and exits all around the sphere.

The water can be pumped below the sphere very slowly, but at high pressure, and that way still lift the sphere.

The water doesn’t get squeezed and accelerate towards the outside edge, because as it flows outward, the circle gets bigger. The area expands.

Hence it slows down, drops in pressure, and comes out at a trickle.

It’s not like pinching the end of a hose. It’s like adding a kink in the hose before the end. There is huge pressure before the kink (the water that lifts the sphere), and after the kink the pressure is super low (the water exiting from below the sphere). It makes the water slow down and come out at a trickle. Even though the pressure before the kink is huge.

People who think awareness helps: “I know about this problem now” immediately stops thinking about it

The architecture is certainly a big issue.

Nothing else really maps 1:1 to the PS3 CPU. The main overhead is not from translating the libraries. It’s emulating the CPU/recompiling the code.

There would likely be no efficiency gain from running the games on software that more closely matches the PS3 OS.

It’d be like the difference of using windows instead of proton to run games. Not always better. Sometimes worse.

That’s because you can’t truly delete a comment in a chain without breaking it.

If others replied to your comment, you deleting your comment would break the thread.

So a placeholder gets left behind.

Makes sense to me, actually.

It’d be an easy way to get and also store the DRM-free offline installer, in where you don’t have to permanently allocate active storage to keep the installer around.

Cases like this is why I can never bring myself to let assholes be.

If someone says something rude in a game lobby, or lemmy comment…

I will tell them off. That they call me “soft” for caring is the pinnacle of madness. I can take it, the ones I’m worried about are the ones who shouldn’t have to.

If I can make even a fraction of those who engage in verbal abuse within range of me reconsider, I might save a life.

That some of the people I engage genuinely don’t care, makes them subhuman. Their tragic lack of empathy would be hilarious, if it weren’t lethal.

So what is it you actually wanted to add?

If you aren’t bringing up practical difficulty to argue in favor of theoretical impossibility, what are you adding?

How is difficulty relevant in the face of the importance of preservation?

As for python… What?

Why would you emulate for python, when what actually happens is that new interpreters get written? It already runs across several operating systems and processing architectures thanks to interpreters existing for each one.

Python isn’t some hobby language anymore. Far too much serious infrastructure uses it for future interpreters not to get made.

So?

Emulate.

As long as the application doesn’t rely on something external like a server that no longer exists, it can always be run.

This isn’t about a hardware system changing. If you can either find the original hardware or simulate it, it should run. Not just go “expiry date passed, fuck you”.

You’re saying it can sometimes be practically impossible. That doesn’t mean it has to be theoretically and actually impossible, too.

You’re right.

Burning the book in case it might offend some future reader is reasonable.

/S