[cheezymoogle]

50kbp/s is equivalent to dial-up speed. 3G was only four times faster at 200kbp/s. I wouldn't say that LoraWAN's speed is the limiting factor.

A single second of transmission (i.e. ~6kb) is about 1,500 characters in UTF8, assuming no overhead. With an average of 6 characters per word, that's still 250 words per second, more than ample for human communication.

[vdqtp3]

Your comment alone, with associated HTML, is 1159 characters. The Manyverse logo on their homepage is 46K characters. Scuttlebutt's "large-hermies-dancing.gif" is 118KB.

Not including IP and TCP overhead, the Scuttlebutt handshake is 340 bytes, each message in a box stream is 35 to 4130 bytes, and the box stream finishes with another 34 bytes. There are two box streams for every Scuttlebutt communication. There are also RPC messages sent in the box stream that are a minimum of 9 byte header and 9 byte goodbye.

Transmitting a single post containing the text "Second post!" takes 563 characters, not including the handshake, header, footer or encryption overhead.

The web and social networks are far more data intensive today than low bandwidth modalities can support, especially long distance, low power RF links that would be rapidly saturated.

Restricted length text a la Twitter might be manageable. Encrypted high fidelity blockchain social networks? Not a snowball's chance in hell unless so few people are using it that nobody is saturating the link.

Edit: In North America for LoRa there are 64 125kHz uplink channels, 8 500kHz uplink channels, and 8 500kHz downlink channels. This should give you an idea of how little capacity there really is.

[cheezymoogle]

The scope to which I was responding was the assertion that 50kbp/s is not enough to transfer more than a handful of bytes.

You are correct that images, encryption, and blockchains are outside the scope of what a constrained connection can sustain, but that's a bit like saying "We can't power an industrial economy on consumer-owned 100 watt solar panels" which while true, misses the point.

Yes, if you're using court etiquette protocols to transmit data, you're going to incur very large amounts of overhead.

If you're using insecure low-sophistication protocols in sparsely populated areas (e.g. flyover America), then the possibilities are much more expansive. Yes, the latter by definition isn't commercially nor urbanely (viz. pertaining to densely populated cities) viable, but for certain demographics that's a feature, not a bug.

[vdqtp3]

It's not 50kbps per device, I would agree that could be manageable. It's 50kbps for the each channel (of which there are a limited number) that's serving a 5-20km radius.

[GekkePrutser]

In addition to this, in Europe LoRa (or more specifically the 868Mhz ISM band it uses) has a legally imposed duty cycle of 1% so it can only use 1% of the available airtime to transmit.

This makes it really impossible to send anything but a handful of bytes

[genewitch]

Matrix protocol has a extreme low bandwidth mode, and I know a couple of other things are aiming to be more "iot/off grid capable"

Even packet radio would allow some of this stuff to flourish in remote areas - not matrix, it's encrypted.

[moderatorium]

very clear illustration of bandwidth usage there, thanks for sharing this one.