Only [ Confirm ] and [ Set later ] in the dialog? No way to never set/change/cancel? Rapist mentality?

Whenever i get the Crash report window I fill it as much as possible and send, I stopped manually going into bugs.kde.org and manually create bug after the response I got was “it’s probably nvidia issue”, but for last months I’m on new AMD gpu and those issues still persist and I already learned to just periodically run kwin_x11 --replace& disown; exit to fix them and go with my life.

And yes, the issues is probably not easy to track, as it only happen after at least few days or running it.

Just recently the switch from plasma 5 to plasma 6 introduced another bug with the windows - every few hours/days i get a window that is visible but all mouse clicks go thru it and actually the window behind it receive those clicks, to fix it I (once again) do kwin replace.

And yes, I’m heavy user with 2 workspaces * 2 desktops * a lot of open apps/windows running 24h/day, to the point just few days ago I got and error in console “maximum number of clients reached” when on X11 when trying to open any new window… (And I didn’t run out off RAM (48GB) or VRAM (16GB))

So as much as I’m full time on Linux desktop for 8 years, the experience did improve a lot, but the stability of long running KDE Plasma slowly degraded over last years, hence my response.

And I did have the same issues on my 8 year installation of Arch and on KDE Plasma.

I had only few crashes and only on plasma 6, yes I did switch to 6.0.1 which is very quite new, so few crashes could be expected, but I’m on 6.0.4 (will update to 6.0.5) and kwin is still buggy with window ordering, and compared to xorg, on wayland I can’t do kwin_wayland --replace because it closes all my open windows/apps which is annoying. Heck, even killing/restarting plasmashell also shutdown all apps that was started from menu…

Why did I switch? Because of hoping that issues from 5.x was fixed in 6.x but that was not the case :(

Same that not many issues will crash whole kwin so they will never trigger “crash report”. It would be nice to have some some way to easy “trigger” state dump and open issue. That I would like the ‘stability’ and those last 0.001% of stability polish to be the focus for KDE.

So as much as I love <3 KDE Plasma, and I yearly donate to KDE, I still can’t with clear mind suggest anyone to switch from MS Win to Linux (KDE) especially now with Wayland that made me lose all my unsaved work 3 weeks ago due to kwin crash.

Make the plasmashell especially kwin stable, as its crashing a lot, and with wayland it’s taking whole session down…

To the point I went back to Xorg

I use my phone as display to play videos of playera playing slow chess games that help me go to sleep, and not all phone screens can be set low without hackery (like using player that software video dimming filter) so I personally love this change!

I didn’t left it, I needed provide that “part” to it to get the correct answer.

Because like in the whole thread is mentioned over and over again, chatgpt doesn’t know the correct answer, it’s a mathematical model of “what looks ok” and “what should be the next word”, it looks ok to try to put --reset parameter to reset it, but because chatgpt can’t actually check documentation of podman stats if the param exists, it just generate it based on “common known text patterns”, and “common known text patterns” are written in a way suggesting that it is the truth.

So once again - do your own research if following the results it could cause breaking both in tech and especially in life. And that is true for both chatgpt and random pages on internet.

In this case I did exactly follow chatgpt answer without doing fact checking - I asked chatgpt, I copied the command and pasted it into terminal, because I know that if it didn’t work the worse that could happen it would fail and do nothing. But It’s bad for new people that will not know what the result could be if it’s wrong!

@z00s Don’t take me wrong. I’m not telling not to use it, on the contrary.

You should use any tool that helps you do your job/task. But you should try to understand how to use those tools wisely.

Telling someone never to use ChatGPT is like telling someone to never use excavator. That is wrong, you should use excavator but you should know what is an excavator, and what harm it could do by for example accidentally destroy a building or even hurt someone (or youself) if not use wisely.

Ktoś mógłby się przyczepić do “or search history after you’ve closed all of your incognito tabs” Wskazując na to że nadal google bazując na twoim IP (albo nawet powiązując z twoim kontem powiązanym z chrome) zapamiętuje co wpisywałeś w wyszukiwarce mimo że cytat sugeruje że nie będzie tego robił.

Ale bardziej kwestia że przeglądarka i wyszukiwarka (i inne) pochodzi od tej samej firmy, a więc ktoś mógłby myśleć że jeśli w przeglądarce “google (chrome)” mówisz że chcesz być “incognito” to “(wyszukiwarka) google” powinna nie pozwiązywać twoich akcji z tobą bo w końcu powiedziałeś to wprost “googlowi”. Więc jest to kwestia “przetwarzania danych” przy umyślnym wskazaniu że nie chcesz by je przetwarzano.

Example that confirms that “Chatgpt does not know truth. It does not know if the info it provides is true.” or more like “It will spell answer that match your inquiry that sound correct even if it’s totally made up.”

https://chat.openai.com/share/206fd8e9-600c-43f8-95be-cb2888ccd259

Summary:

User
in `podman stats` you see BLOCK IO as a summary of hard drive activity.
how to reset the 

ChatGPT
To reset the block I/O statistics displayed by podman stats, you can use the podman stats --reset command.

User
Error: unknown flag: --reset

ChatGPT
Apologies for the confusion. It seems I provided incorrect information. The podman stats command does not have a built-in option to reset the statistics.

So once again, don’t be afraid to use it, but do your own research especially if following LLM could result in something breaking both in tech or in life.

don’t run any commands that you don’t understand. Ask it to break down any commands it tells you to run if you don’t understand them.

You need to pay extra attention to this, as ML models will spit out commands and parameters that doesn’t exists if there was not enough examples in training dataset for that action. Especially with explain as it could just spit out totally wrong but “sounding good” explanation for parameter etc as it not always will tell the magic keywords like “typically” that indicate that it doesn’t have confidence as it’s “based on other similar command/knowledge”.

In your example it spit out:

 -m: Prune empty directory chains from the file-list.
 --prune-empty-dirs: Exclude empty directories that result from the inclusion/exclusion pattern.

which is actually exactly the same parameter with 2 different explanations, you can confirm this with man rsync

 --prune-empty-dirs, -m   prune empty directory chains from file-list

So the more edge case you have the bigger chance it will spill out bad results, but those new models are shockingly good especially for very common use cases.

ALVR works on Linux with Quest 2, BeatSaber works fine, dont know much about other games.

You just posted, so first step achived.

cfdisk only changes the partition table, this table like a small paper that you store at the front (or back) of drive where you put information, it’s just a list of coordinates like from this point to this point is your home, from this to this is your yard, from this to this is your neighbor. Just because you changed the values on your paper doesn’t actually make your neighbor closer or further.

System read this list to figure out where are the “borders” between different sections that you defined to load and use them logically for multiple file systems.

and must delete the browser signals that indicate when private browsing mode is active, to prevent future tracking.

Thinking how will they do that? Is private browsing sending some additional headers?

Connect, disagree, return?

Just by looking at the biggest instances someone could think that they cover most of the users but I’m positively shocked.

Kudos for the work!

But the summary is missleading.

If it’s Germany … My best bet would be at the TV kable/streaming companies, but probably behind additional paywall.

AFAIK Germany regulations are strange, the stream copyright owner can broadcast video and later based on your IP send you a Bill to pay claiming unauthorized access.

That’s why I stated that it should be illegal to promise product while selling a undefined time limited license, there should be a clear minimum time stated when you “buy a subscription” for (single player?) games.

I concur Buyer should not gain rights to product, so they should not be allowed to profit from it, but they should be able to preserve it, unless the license that you actually buy had a time limitation, but that should be clearly stated when you buy it that you only buy access to it to (at least) X amount of time like you have with online subscriptions.

https://www.youtube.com/watch?v=PRGszw0KubQ

I would suggest to add those additional AP as soon as possible if going UniFi way, as adding more later might require to replace all units as we already saw multiple times with UniFi devices, as they are quick to deprecate old ones and make them not work with new ones.

I like the concept, you could also detect the compose file name automatically.

Or write down hashes for tag even if compose don’t use your variables so reading from compose, so people would not need to relay on the script but could use it as additional backup