And I’m fine with the essential stance of Storj towards the matter as far I can deduce from reactions of Storj-representatives. But what turned me on in this discussion was the tone (e.g. calling it cheating) and also the reference to the outdated ToS pretending it says what they are saying (both by representatives of Storj, formal or informal; including you as assigned Leader). We had this discussion previously, for example here in which already all the same arguments used here passed by.
But most disappointing is the lacking reflection on it, again especially the tone of the discussion and unfounded references to the ToS.
And I agree, the best option would be a technical / software-implemented solution. But even then you need a ToS aligned with the spirit of Storj, in order to ban any attempts to spread any ideas on how to evade these technical solutions. Such like I did above, for which there is and would be no foundation in the ToS to ban or erase it as long as you don’t have it written down in ToS; even when implemented in the software (which is not normative). If you condemned for example VPNs to be used to connect to the network, the next could be VPS + iSCSI or any file transfer protocol over IP.
Essentially it means, that in the ToS something has to be stated like:
- As proxy for geographic location, segments of /24-IPv4 range of the world wide web are being used. Ingress and egress, are equally dispersed over these segments as one of the most determining factors; non-regarding the fact there can be multiple Storage Node Providers running the Storage Nodes in one segment.
- In order to avoid a single point of failure for multiple storage nodes of one storage node provider as much as possible - for example, but not not limited to, blewing of fuses, power outages, connection problem to the ISP-network or burning down of building in which the hardware had been located - all essential hardware of any Storage Node (including the physical machine on which the Storage Node Software is being executed and hard drive(s) used for it) shall be physically connected. None of these may be associated with another storage node connecting to the network, using an IP from another segment as meant by sub 1.
- If running multiple fully essential hardware-separated nodes in the same location with IPv4-addresses from different segments, each needs to have it’s own ISP-connection or there has to be at least one failover connection. Besides concerning power, every node needs to be on it’s own fuse or needs an UPS.