In fact the Linux date command even has an iso flag: date --iso-8601

There are also variants with single dots too. I sometimes do timestamps without any separator too, depending on the context (such as file backups): YYYYMMDD in example 20260805.

This is art.

You are so evil! :D

There is for everything a

YYYY-MM-DD

The problem is not the format itself, its just sometimes we don’t know which format is used: 07/07/26

Edit: This was probably a bad example, because it does not matter which format, as it is clear it is 7th of July 2026 in both possible formats. A better example: 01/02/26

This is actually one of the reasons why I don’t like the way some do their posts (I don’t mean cross-post, but normal post). Instead a link to the article or whatever they are linking, an image is uploaded, and they link to the main article in the body text. This will prevent the Lemmy system from automatically connecting all articles. So it does not matter if the post is shared with cross-post or created from scratch, if they link to an article with the same address, then the Lemmy system will recognize the same article and connect them in the outer list, across all Communities. I really love this feature!

I’m glad this helped in understanding. And also thanks for updating the post.

Lastly, the links inside the body show only on one of the cross-posts, but the ones outside the body show on both cross-posts.

The outer list is automatically generated and every post to that link will appear in this list, even years later. Its pretty cool, but nothing new, we had that back in Reddit too. When Reddit existed. The post content is just a link added once. When I do a crosspost, then the content of the original post is added automatically too. This is only one source where the original post was shared from, and will not automatically update (unlike the above outer list). I assume it depends on the software they are using, when generating the crosspost. The post body is managed by the user and the app they are using to generate the content, which is a normal link to the web like any other. The above list is managed by the Lemmy system and out of our control. So it can do the optimized linking with short style automatically.

I don’t use PipePipe, but FreeTube. Since YouTube updated their system to use SABR (I don’t even know what exactly that is, I just know it causes issues for us), I get sometimes a reload loop for some videos. In some cases my IP is blocked. Only a VPN helps in those cases. This plagues me / us for weeks and months. I’m surprised you didn’t encounter this issue earlier; probably because the developers of the app you are using did a good job.

I run Qwen 3.6 27B quantized down to Q4_K_M or even a variant Q5_K_S on my 8gb VRAM entry level AMD GPU, with support of my CPU and 32gb system RAM. Yes, I also limit the Context Length heavily to something like 18k. It’s slow. But the point is, you don’t need necessarily 16gb VRAM.

But it’s better to use faster models for this type of hardware anyway. The MoE type of models (such as 26B A4B, or 35B A3B) are vastly, vastly faster for normal usage, but they are a bit worse in some cases. Also Googles QAT trained model versions also have less RAM requirements without losing much quality.

I’m just saying that, so others are not discouraged too much. It’s not the same experience as the full version off course, but you can use them with some tricks on weak hardware too.

Just because you like children does not make you an evil person.

I am just wondering why would that be an option 😆?

I guess so that actual bots can be marked as bots, so they do not get deleted. There is also an option to hide bot posts (which I do), for anyone who is bothered by. Bots does not automatically mean spam. There are legit bots useful to grab news and auto-post them faster than humans could. I didn’t know they cannot vote, which makes totally sense. Any bots not marked as bots can be reported and they will get banned (presumably, I don’t know).

Sorry for the random late reply. I just created a post and was looking through other posts.

I just played The Stanley Parable: Ultra Deluxe on Linux.

Also noteworthy is that Cloudflare isn’t about Desktop operating systems only. I don’t know how StatCounter operates, but they probably ignore everything that is identified as mobile and Smart TV. Cloudflare is about ANY HTTP connection / request. So its natural that these numbers are quite different, as they are different in their scope anyway.

I don’t think one data point is better than the other. They are all a small portion of the entire base. Its like looking at different game reviews from different people to get an idea how good it is (I know little odd comparison :D). In example Cloudflare data is only relevant to those communities and sites that can pay for it and don’t have a different solution. Cloudflare is also a “recent” phenomenon; not sure how relevant that info is though…

It doesn’t necessary mean Linux. Given than Windows is going rapidly down and people get fed up with Windows a lot, but can’t or don’t want to switch to Linux, it is also probable that a lot of people go into stealth mode. Lot of YouTubers and media talk about the issues now. Then off course you have those niches grouped together like BSD and other systems, Tor browser and so on. It’s just a curious thing, because that makes a huge chunk of the entire user base.

My curiosity is, consists Unknown a collection of many small groups or is there a huge group responsible?

Arch is the endgame.

Windows:  56.61%
Unknown:  21.45%
OS X:     11.89%  |
macOS :    4.48%  | --- 21.94%
Linux:     4.36%  |
Chrome OS: 1.21%  |

I wonder what “Unknown: 21.45%” consists of. That is almost as much as OSX, macOS, Linux and ChromeOS combined “11.89+4.48+4.36+1.21 = 21,94”. This is a huge chunk. Also why is macOS and OS X listed separately, as they are basically the same group of operating systems. I mean we group all Windows into one, and even all Linux into one.