@stupid_seb: Glad you are interested in the project! We will be in touch regarding French translation.
@eXpl0sive: Your comments are pretty much spot on. Thanks for taking the time to read in depth about the project! We know there is quite a lot of detail to get through. I will address each of your points in turn.
Yes. All a 'full' client or miner needs to do to join the network is download the last sync block, and the TX stubs from that block to the present block (the sync block 'period' will change over time, but the total download required will stay comparatively minute).
Yes. The only clarification I would make is that the more blocks a miner has, the more likely they are to be able to mine the next block. However, not all blocks will be equally stored in the network, so miners are incentivised to store blocks that few other miners are storing (as this lets them compete in the PoW part of the PoA system among a smaller group of miners).
The media can be included on the weave, but owing to the price, we don't expect people to do this often. The files associated with the hashes can be stored anywhere, and 're-united' with the data on the weave later. We anticipate that if the Archain is successful, 10 - 20 years from now sites will pop up where people can offer rewards for the original media that matches a hash stored on the weave, allowing them to completely re-construct pages they find particularly important.
It is possible, but the incentive for someone to find a block that appears 'lost' becomes extraordinary. It would grant the holder the soul ability to mine a new block that requires that old block, guaranteeing the associated rewards. Then, once the new block is found, in order to be accepted and verified by the network the miner needs to distribute the recall block as well as the new block. As the recall block is 'rare', many of these miners will store copies of it. Subsequently, over time, these storage imbalances will 'level out'.
This certainly covers a good part of the project's functionality. I would add that it also does decentralised verification of the content prior to archiving (unlike archive.org), as well as arbitrary permanent data storage for a fee ('Permanence as a Service').
Thanks again for your attention to the Archain project! I hope this helped clarify some aspects of the project. Let us know if you have any other thoughts/comments/questions/ideas!
Thanks,
Archain Team