[martin_a]

Yeah... But WiFi doesn't work when I'm sitting on the toilet two rooms down the corridor. Crazy times.

No, to be honest: Very impressive. It's kind of a "dream" of me, to bring Internet to the river/picknick area two kilometers away from me. LoRa would probably be the way to go, will have to look further into that.

[Springtime]

From the relatively little I understand the lower the radio frequency the lower the bandwidth but more range it has/can pass through, while conversely the higher frequency the more bandwidth but shorter range/harder to penetrate obstacles (see eg: 5G). LoRa is the former and used for very low bandwidth applications, so not really suitable for browsing websites but has been used for for remote IoT stats, text messaging, etc.

(Fwiw I wasn't the one who downvoted but thought it'd be worth mentioning)

[stavros]

It's not so much about the frequency (though what you say is true), as there's LoRa at 2.4 GHz as well. I'm guessing the one in the article was 868/915 MHz, though.

[wienke]

For sure would be great but LoRa has limitations as it is designed for small sensors that sent very little data: https://www.thethingsnetwork.org/docs/lorawan/limitations/

[martin_a]

Ah, I see. Seems like directional antennas for WiFi would be what I need then.

[rcarmo]

Yes. 2.4GHz ones will do fine (and better than 5GHz for that range).

[Svenstaro]

You're looking LoRaWan which is not the same as LoRa. In fact I believe web browsing to be quite feasible if you just use LoRa with your own network stack.

[goodpoint]

Not at all. The effective speed is extremely low due to the duty cycle.

[wienke]

How would that exactly work?

[Svenstaro]

I did this once as a small experiment. Basically use a small TCP stack with LoRa as layer 1 and then expose the whole thing as a Linux kernel module. It was a bit shitty but it seemed promising.

For some reason, I don't find LoRaWan at all alluring. As far as I can tell, in the end it's in the hands of one company and that really defeats the spirit for me. I mostly just use LoRa for point-to-point connections with my own shitty networking stack.

[dbrgn]

At least on 868 MHz, if you stick to the 1% duty cycle limitations that the ISM band allows, I doubt that you'll be happy with the web surfing performance.

(For those that don't know duty cycle limitations: https://www.thethingsnetwork.org/docs/lorawan/duty-cycle/)

[dgacmu]

This may be an EU duty cycle vs US time per channel thing? In the US you can frequency hop to get higher performance (the limit here is no more than 400ms per 20s per channel), which seems more aggressive than the 1% power sub-band EU rule (but ianal and I haven't read the EU rules carefully).

[wienke]

That is pretty cool. You might want to try again with 2.4Ghz LoRa. Range is shorter but still longer than 868mhz. If you search 2.4Ghz LoRa on our Youtube channel you'll find some interesting content on that.

[moffkalast]

Define "internet". LoRA is more like an occasional SMS service than a network connection, so I doubt you could get the average modern 100 MB website to render through it. Would be more practical to get a Starlink dish set up or a directional wifi antenna on both ends, those can do a few km on a good day.

Or just you know, use the 4G that is probably there already.

[martin_a]

> Or just you know, use the 4G that is probably there already.

Yeah... Projects like open WiFi are probably becoming less and less useful, as everybody is carrying 4G/5G modems in their pocket.

[dopidopHN]

At some point I was bullish on LoRA + Gemini or scruttlebutts.

A different, low resolution internet basically. But LoRA is too slow for a lot of things. Too.

Still nice.

[moffkalast]

In the end, our butts remained unscruttled.

[Tepix]

LoRa has pretty strict limits on how often you may use the frequency. That severely limits use cases. Browsing the internet is completely unfeasible.

[Saris]

The difference is the 160bps they have with this isn't going to be too much fun for web browsing!