Recent searches
Search options
well that's a way to answer a question without answering a question https://docs.syncthing.net/users/faq.html#why-does-syncthing-connect-to-this-unknown-suspicious-address
so it turns out this open source dropbox-style thing has "relaying" on by default; it doesn't ask you, and instead assumes it's ok by you to just connect to random ips without asking first
I don't care that they "can't see my data", the assumption is a breach of trust
Of course, it took hunting down some forum thread to understand what was going on
open source projects really need to take this sort of user experience more seriously
@mattly@hachyderm.io Do they have any technical justification for the relay system? The only thing i can think of is that it might help transfers on networks with poor peering.
@puppygirlhornypost2 I didn't go looking for one; but the non-answer in their FAQ is enough to put me off from the whole thing
@mattly@hachyderm.io I mean yeah me too, I'm just curious at this point why they would think that's okay.
@puppygirlhornypost2 @mattly i think the main point is that most home- and almost all mobile networks use NAT. so, connecting from your (NATted) phone to your (NATted) homelab is going to be quite difficult without outside help - this is what the "relay" proto is for:
https://docs.syncthing.net/specs/index.html
wether or not you want your node to participate as a relay node ... now, this is something that maybe they should get consent for. but it does help everyone to turn it on.
@rm if it looks like malware and talks like malware…
@mattly it is a p2p protocol. so it does p2p protocol things. would you be suprised if your bittorrent client started to connect out to the world? of course you wouldnt!
@rm it’s not advertised as that, it’s advertised as a file sync program. I see nothing on Syncthing’s homepage indicating it’s P2P. My use case doesn’t require that, and the architecture actually works against my use case
@mattly yes, i agree that they probably should be a bit more upfront about this - but on the other hand - in all UIs etc, it never uses DNS names. and yet, clients do find each other. without a centralized server, and if not for some sort of p2p functionality - how do you think this would work?
i mild disapproval of the lack of "p2p" in the overview is understandable, but malware? that's quite a leap.