For the past few weeks now we have had Peer Discovery, which was supposed to make IOTA more easy to use for humans. Now since then, we've collected enough data to be able to determine that Peer Discovery is causing more problems than it does good.
Now it should be noted that CfB was against PD since the very beginning, and the only reason why we decided to add it was to simply make the startup process easier with no hassle for people. But now exactly this is causing several problems.
Snapshotting: Peer Discovery basically makes it possible for previousu transactions to be rebroadcast and picked up by the majority of the network. Thus coming back to pre-snapshot state.
Difficulty Adjustment: Difficulty is supposed to self-adjust according to the network topology (tps, etc.), with PD this difficulty adjustment does not work as intended.
Bandwidth: The bandwidth consumption increases exponentially, so if we want to have a system fit for running in IoT where every tiny bit of resource is scarce, we have such a waste.
Syncing: Currently syncing is fairly slow - without PD it should speed up.
There are obviously more problems associated with additional attack vectors which are introduced through Peer Discovery (related to DOS and eclipse attacks, which Bitcoin has suffered from in the past).
Additionally, without Peer Discovery we have a system that mimics the IoT environment much more closely, thus making it more fit for applications at the edge of the network.
The thing is, this is a decision of either completely removing Peer Discovery, or keeping it. We cannot have a split network where one part is utilizing Peer Discovery and the other is not.
What we propose is to completely remove Peer Discovery and go back to the previous state, which was manually finding peers. This will help solve the above problems and make IOTA more fit for IoT. Obviously we understand that humans will largely be annoyed by this, so to help humans out in this situation we - the Foundation - will provide the following:
Peer Discovery Service: Automated web service for finding peers. This will of course have some sort of KYC measure (e.g. connect your twitter account), but it should help humans in retaining the same ease of use of IOTA as we currently enjoy
Web Wallet and Light Client: A web wallet as well as light-client that will connect to nodes operated by the foundation. This means that you as a user will not need to download the entire Tangle and can start utilizing IOTA from the getgo.
Sandbox: As discussed previously, we will provide the sandbox environment which consists of a set of trusted nodes. Main purpose is of course to easily interface with IOTA.
It should be noted that the entire vision and purpose of IOTA is the Internet of Things - therefore our main target users are machines and thus, large corporations. These services are provided to help kickstart the project and ensure that humans can benefit from it equally.
Voice your opinion below. We will make a decision as soon as possible (preferably this weekend), as this is holding back progress.