[chrismorgan]

> ignoring aggressive non-sharing downloaders of course

It has not been reasonable to ignore such users for the last five years or so. Like it or not, the computing landscape has shifted to mobile devices which aren’t on all the time and have limited power and bandwidth. Perhaps BitTorrent is OK now, in its niche, but if it were to go mainstream for downloads you’d go from a small fraction of non-sharing downloaders to a high fraction.

Skype’s architecture is an interesting one in that space: it used to be distributed, with many computers all over the place nominated supernodes; but the shift to mobile made that architecture untenable, and so they had to shift to a centralised model, which generally performs worse, to keep it working at all—not for surveillance, but for scalability!

[asdkhadsj]

> It has not been reasonable to ignore such users for the last five years or so.

You're taking me the wrong way. I wrote that because I didn't feel like writing paragraphs going into explicit detail over the pros and cons of seed avoiders and how one might handle them, etc. It wasn't the point of the conversation I was trying to make, informing about the general design. BitTorrent works without you hosting the file, so does IPFS, that's the point. Nothing more, nothing less.

Which, I should have known better from HN, but /shrug. I guess in the future I need to be more explicit when I ignore a topic.