Badabinski

Cities probably have a higher density of towers, or the towers in cities have more capable antennas. Point-to-point microwave links can be pretty damn fast and reliable. They have their limitations, but even low-end systems like some of Ubiquiti’s 60ghz stuff can form full duplex 5Gbps links at 10+ kilometers. Fiber is still king, but I’m guessing the backhaul isn’t the issue.

I’m guessing that the issue is congestion on the client radios. 5g is supposed to be much better at dealing with this thanks to time sharing improvements, but it seems likely that there just aren’t enough towers. One scenario that seems reasonable is that your telco (incorrectly) assumed that they wouldn’t need as many towers when upgrading, so they only upgraded a subset of their towers and removed old ones once 4g was deprecated.

edit: you might be able to get better information about wtf is going on by using a community-sourced site like https://cellmapper.net/

I believe you can use that site to get info about how many towers there are and what the client-side congestion is like.

EDIT: ew, cellmapper is closed source. OpenCellid or beaconDB seem to be open source equivalents.

Abso-fucking-lutely, amen and hallelujah. I want 6G to focus on improving range and performance in marginal conditions. When shit is good, 5g is fast enough for now. I don’t know how you improve range and penetration without going to lower frequencies, so maybe we should try to do that? Lower frequencies mean less bandwidth, but RF is black magic fuckery and there’s all kinds of crazy shit that can be done with time division, so maybe we can improve throughout in the sub-ghz regime. I dunno about that, I’m just an idiot software developer who is thankful that shit works without me having to sacrifice a goat.

Maybe there’s a way to broadcast at higher power levels, and maybe there are ways for base stations to be more sensitive or do filtering to increase SNR. I have no idea, but I think that should be what the telecos focus on. Better service over a wider area with the same number of towers would be huge.

Python is my primary language. For the way I write code and solve problems, it’s the language where I need the least help from an LLM. Python lets you write code that is incredibly concise while still being easy to read. There’s more of a case to be made for something like Go, since it seems like every single god damned function call ends up being variable, err := someFuckingShit() and then a if err!=nil and manually handling it instead of having nice exception handling. Even there, my IDE does that for me without requiring a computationally expensive LLM to do the work.

Like, some people have a more conversational development style and I guess LLMs work well for them. I end up constantly context switching between code review mode and writing code mode which is incredibly disruptive.

As a senior dev, I have no use for it in my workflow. The only purpose it would serve for me is to reduce the amount of typing I do. I spend about 5-10% of my time actually writing code. The rest of my dev time is spent in architecting, debugging, testing, or documenting. LLMs aren’t really good at most of those things once you move past the most superficial levels of complexity. Besides, I don’t actually want something to reduce the amount I’m typing. If I’m typing too much and I’m getting annoyed then it’s a sure sign that I’ve done something bad. If I’m writing boilerplate then it’s time to write an abstraction to eliminate that. If I’m writing repetitive tests then it’s a sign I need to move to a property based testing framework like Hypothesis. If the LLM spits all of this out for me, I will end up writing code that is harder to understand and maintain.

LLMs are fine for learning and junior positions where you’ll have more experienced folks reviewing code, but it just is not that helpful past a certain point.

Also, this is probably a small thing, but I have yet to find an LLM that writes anything other than shitty, terrible shell scripts. Please for the love of God don’t use an LLM to write shell scripts. If you must, then please pass the results through shellcheck and fix all of the issues there.

Onshape is an okay option for Linux (I’ve been able to do everything I used to to in Inventory), although I hate that it’s cloud based. I know that a rug pull is inevitable, but I figure I’ll stick with it until then.