Filecoin February 2020: tipsets, proof system, roadmap

Periodically, I follow updates on blockchain projects that I believe have still unexpressed potential. Among these is Filecoin, which is a project built on IPFS (which I have already discussed extensively).

The updates for February are as follows.

Tipsets

Filecoin explains how tipsets work, a module of their Expected Consensus Protocol. Tipsets are a variant of child blocks in a DAG, a directed acyclic graph, used on Ethereum. In short, while block consensus on the bitcoin blockchain happens on the longest chain, on Ethereum each block can have ‘uncle’ blocks (Uncle Blocks), which are mined in the same Epoch as a parent block, but did not reach the majority consensus.

In Ethereum, Uncle Blocks can give more relevance to a new child block that was mined from the parent block, effectively entering the main block chain. In Filecoin, however, the approach is slightly different: each tipset (set of latest blocks) has a weight, given by the sum of the parents it is connected to. In the image below there is a schematic representation, which is certainly easier to understand.

Since, moreover, validated blocks are only those that come from the same parent and are mined in the same epoch, a potential divergence of blocks (a temporary fork between one epoch and another) is quickly resolved and at the same time more miners are rewarded, who may have, together, produced more blocks. Thinking about this, a doubt comes to me: how to avoid mining then being centralized in the hands of a few? How to ensure that at each epoch there is even a partial change of miners elected for the production of the next block?

(Source)

The Proof System

Filecoin announces details of its Proof System, a mix of proofs to demonstrate authority, deserve it again and receive reward from stored data space:

• Proof-of-Space: to demonstrate authority in obtaining data to save on one’s hard disk, the miner must demonstrate having the required space on the storage medium;
• Proof-of-Storage: is the system that serves to demonstrate having a specific piece of data in one’s storage medium;
• Proof-of-Spacetime: is the system that asks a random number of miners to publicly prove that the data they received are still in their storage medium.

Proof-of-Space and Proof-of-Storage were combined by Protocol Labs (the team developing Filecoin) into Proof-of-Replication.

Along with this explanation (which is actually much more complex than this), Protocol Labs explains the reasons why it took so long to develop Filecoin: to optimize these configurations (finding the right mix between incentives, processing time and security), they had to rely on software that solves optimization problems with constraints, and find the right constraints that guaranteed the aforementioned mix.

(Source)

Roadmap

• After launching Phase 1 of the Testnet on December 11, 2019, over 3.6 PiB of proven storage were accumulated;
• Developers released Filecoin implementations in C++ and in Rust and are working on interoperability between these implementations;
• Proof-of-Replication was optimized (as described above), optimizing the SNARK proof system (2x increase);
• Textile announced they are working on tools to easily build applications for the Filecoin network.
• The platform Terminal.co was launched (not by the Filecoin team), which promises to easily distribute your site with IPFS starting from a Github repository. The site is simultaneously distributed both on IPFS and on http(s) protocol via a global CDN;
• Zondax published the repository related to the development of an app to view the Filecoin Ledger, while Open Work Labs is working on a wallet.
• Trustwallet announced support for Filecoin tokens

Source

Other resources: How Filecoin works, in 5 minutes