[DuskStar]

Ah, yes. The good old "don't make that tool because bad people might use it" argument.

Similarly, we should ban sharing information on hacking/security because bad people might use it, and require that all communications be visible to the government to prevent abuse.

[JoshuaMulliken]

That is not the argument being made. It is an argument that having friction in the process to do something that may have dangerous outcomes is sometimes positive.

It is much the same as arguments that there should be gun training required to purchase a firearm. Adding friction to a process sometimes provides positive benefits.

[DuskStar]

> That is not the argument being made. It is an argument that having friction in the process to do something that may have dangerous outcomes is sometimes positive.

The argument being made, as far as I can tell, is that this tool will make doing bad things easier, and thus this tool should not be made. As any tool to resist censorship will also enable said bad things, I do not find this to be a valid concern. (And since some governments would list censorship resistance itself as one of those bad things, this is always going to be true)

> It is much the same as arguments that there should be gun training required to purchase a firearm.

I think the tech equivalent to this would be mandating a "don't sexually exploit children" class before allowing purchase of a computer. (As you might guess, I'm not a fan)

> Adding friction to a process sometimes provides positive benefits.

Sure, but you have to make sure the collateral damage is minimized AND that you're actually catching the offending segment of the population. Mandating gun training before purchase is actually a great comparison here - the vast majority of gun crime in the US is committed with illegally acquired weapons (which would be unaffected by the mandate), just like I imagine that the majority of CSE imagery shared on the internet uses something more robust than OnionShare. So in both cases you'd have high collateral damage with minimal impact to the population you actually care about affecting...

[JoeSmithson]

> I imagine that the majority of CSE imagery shared on the internet uses something more robust than OnionShare

The majority is shared on normal social media and Bittorrent, although the worst is shared on dark web sites of the same robustness as OnionShare.

[dependenttypes]

> and Bittorrent

What? No, far from it. Bittorrent is pretty clean regarding illegal material (excluding piracy ofc) - which makes sense, after all you leak you ip address and the torrent that you are downloading to the whole network (if you have DHT enabled) and/or to your tracker, plus your ISP can see what you are sending and receiving (while there is a standard for encrypted transmission in bittorrent I do not think that it is widely used). What do you consider as "CSE" anyway? Would a picture of a girl at the beach be considered a "CSE"? If so you might find such torrents (I wouldn't know), but I think that calling it "CSE" is dishonest.

[JoeSmithson]

I'm a detective that works exclusively on online child abuse, I regularly arrest people who have downloaded and/or distributed IIOC over Bittorrent.

The definition of IIOC is provided by the Home Office and split into three categories. Ultimately it is decided by a jury although the categorisation is rarely contested.

[vageli]

> > I imagine that the majority of CSE imagery shared on the internet uses something more robust than OnionShare

> The majority is shared on normal social media and Bittorrent, although the worst is shared on dark web sites of the same robustness as OnionShare.

By what measure?

[JoeSmithson]

Number of files shared

[DuskStar]

Number of files shared that you know about, which I think is a pretty important distinction. Bit harder to track sharing over Tor, after all. (Which is the point!)

[vageli]

I meant the distinction between majority and worst. What is "worst" in this context? And by what measure?