Comment by TeMPOraL

3 years ago

Neither. I find this a weird notion. The message belongs to both sender and recipients. All of them, in case of one-to-many communication.

If there's someone least entitled to message ownership, IMO it's the sender. To grant the sender the ability to take messages back is to allow them to encroach the "personal space" of the recipient - to make unilateral changes to the sphere of reality they considers their own.

Physical analogy: to "unsend" a letter, you'd have to break into my house and steal it.

Digital analogy: to "unsend" an e-mail, you'd have to break into my computer / mail server and delete it there.

"Unsending" e-mail exists in corporate (organizational in general) contexts, but this is tied to an artificial environment following a much different sense of rules - those e-mails aren't truly yours, they're the property of the company (org) as part of which you're communicating. In the same way, corporate might let you "unsend" a physical letter too, at least internally. But this is an exception, tied to acting as an agent of an organization; private communications have different defaults.

Also in general, as I mentioned in another branch of this subthread, I don't like solutions that let someone mess with someone else's perception of reality. "Unsending" is doing exactly that, so at the very least, it must not be silent - it must always leave a visible mark. "There was a message here. It was removed by the sender."

I agree with you. Analogously, I find the notion of forced read receipts to be weird. I can accept having a "delivered" status, to indicate that the message reached one of my devices, but a "read/seen" status is intrusive. The physical equivalent of this is sending an internet connected camera and let it send a message back if I opened your box.

This is very much a framing issue and trying to apply physical interpretations on non-physical things.

> Digital analogy: to "unsend" an e-mail, you'd have to break into my computer / mail server and delete it there.

They're not on your computer or mail server though. They're on discords servers, which grant you permission to view them upon you providing credentials to prove you're allowed to see them. Removing the message is removing your access to content I have authored and previously granted.

It wouldn't be weird for me to remove a blog or mastadon post of my own, yet those are very clearly one to many communications. If I hosted a page and gave you login details which I later rescinded, that wouldn't seem odd would it?

None of these are conceptually all that different though. I write a thing and let you read it. I later don't want you to allow you to read it any more. Having granted access once, must that always translate to permanent irrevocable permission? That seems like an extreme position - the most obvious place that comes up is with mistakes. I mistakenly "send" a message to you intended for my wife. Do you have a permanent and fundamental right to it? If you've not even seen a notification that it's arrived, is it encroaching on your personal space for you to not be able to read a message I don't want you to read?

  • >I write a thing and let you read it.

    By the why the physical universe you live in works the moment you let someone see an informational project (text for example) you no longer have a monopoly on that information. You can never remove access of that data from my mind, and with digital information under a end to end encrypted channel (and assuming you're not watching it with an OCR application from a monitor) you can't remove access to any other allowed person.

    You have an extreme position, yes. It is extreme in the sense your idea only works in highly controlled situations or where everybody agrees to play nice. These rules do not work on the internet and there is no means or way of enforcing it.

    Once you let photons go, you no longer have a permanent and fundamental right to recall it.

    • I think this has gotten wildly off track. It started with whether it's a feature that you can delete a discord message, there's a lot of steps from that point to wiping information from your mind.

  • > It wouldn't be weird for me to remove a blog or mastadon post of my own, yet those are very clearly one to many communications.

    But that's more of a forum or a publication than an interpersonal communication. I think when you and I talk, we both own the content. The conversation is two-way, and thus it is ours. You have no right to take those conversations from my email server any more than removing them from my mind.

    It's not about one-to-one or one-to-many, it's about is the conversation one-way or two-way. Are you talking to someone or with them?

    • I think the same applies to the Mastodon or blog example (or Twitter, or Facebook) - to me, these are closer to broadcast publishing than to producing creative works, so copyright issues notwithstanding, I feel GP doesn't have full ownership on what they published on the public web either.

      In this context, I feel it's OK for GP to remove their blog or Mastodon posts - not weird at all. However, I believe I am well within my rights to scrap GP's blog, screenshot all their Mastodon posts, and keep them for reference; I'd expect the blog to be indexed by the Internet Archive, and would find attempts at preventing the IA from making a copy as something between peculiar and antisocial.

      That is, there's a distinction between owning the message itself, vs. owning its physical form or means of publishing. The former, I believe, is always shared between sender and recipient or recipients (up to and including everyone, if we're talking about a regular website without any access control). The latter is handled through normal property law mechanisms, but owning the medium only means you can decide who can access the message on/through that medium, not that you own the message itself.

      2 replies →

> Neither. I find this a weird notion. The message belongs to both sender and recipients.

If only we could admit that about copyrighted work.