the entire codebase is written in python yeah https://github.com/markqvist/Reticulum regarding packet loss, if your packet takes 2 hours to go through it may not "technically" be packet loss, but it's practically packet loss for the average person, especially if the goal is the one I want: replacing all ISPs

packet loss is high, latency is high, no QoS, also python implementation

My thoughts is that if we can get an improved version of reticulum up Abd running, and widespread in the next decade, ISPs are gonna have trouble... But that's a huge assumption of course

Just a PSA: fuck starlink. I can't wait till mesh networks replace ISPs for the average person. --- got this email: oh nice, my plan is getting cheaper right? Wrong, this is my current plan: and after this oh so generous offer (where I save 1 entire euro, so generous) it's gonna become this Wait, isn't that half the speed I'm currently getting (on a good day by the way)? Yes, yes it is. So what if I want to keep the 400mbps? Easy, I should upgrade to Residential Max! What's that? 59 euros/month. So 59 euros is cheaper than 40, huh? Must be some math I'm not familiar with, perhaps math has changed recently, and we live in the age of ai... soo let me ask claude me: claude, is 59 more than 40? claude: yes So fuck you starlink!

poor doggo, how dare this human cut his nail, this is an insult to his whole family! 🤣

because this is hard to read (I know, problem of my own making)

biggest miss in ditto is that users should be allowed to have different fonts per section, not just one global font. CSS in general would be nice, but different fonts for header, name and content would be nice

well, there's definitely some code that wasn't human written, but the majority is good old traditional elbow grease

Yeah, I call it fastr. And the only goal I have is to be at twice better than strfry in every bench and so far I've succeeded. It's buggy tho, and to fix certain bugs I will have to drop performance a little bit, so take those numbers with a grain of salt. They're representative, but expect a drop of 10-20%.

I may have a solution for those issues in the coming weeks. Lower disk space (about 1/5th), significantly lower RSS (9x, close to 10x), significantly lower write strain on disk (over 500 times in fact during writes, haven't benchmarked read yet), lower CPU usage (currently 10-25x, might change), faster queries (8x), significantly faster ingest (200x). To be clear the software exists already so I'm not speculating, but it needs a lot of work and improvements.

Was playing the new slay the spire 2, when I saw this card