Bitcoin as its own sidechain!

[aliashraf]

Hello everyone,
Historically, in the main and trending bitcoin literature, the idea of sidechains have been shifted from a scaling solution to a testbed recommendation, a sandbox for experimenting weird ideas. It happened mostly because, the same guys who pioneered the original idea, somehow abandoned it as a scaling solution and focused on LN for this purpose, which is not a blockchain technology, after all. They have stressed enough on testbed and sandbox properties of a sidechain as an advantage, from the first day, so they have their own excuse for the shift, I don't buy it, TBH, but it is ok, not here to put fingers and making accusations, I'm just exploring the field and sharing ideas with forum members, whenever there is something with a merit.

Just a few hours ago, this was posted in a bitcoin-dev discussion thread:

-snip-

A "reasonably standard" approach was pioneered in Elements Alpha, where an entire federated sidechain is created and then used as a testbed for new mechanisms, such as SegWit and `OP_CHECKSIGFROMSTACK`.
However, obviously the cost is fairly large, as you need an entire federated sidechain.
-snip-

which reminded me of this "sidechain as scaling solution vs testbed" controversy, encouraging me to change some plans, sharing this idea, as I'm a believer in literature and discourse and their impact on everything including technical developments.

As a matter of fact, sidechain related ideas and constructs are building blocks of a frameork that I'm working on (to be shared as well, very soon, I hope). The interesting point about this framework is that the sidechains under consideration are essentially nothing other than standalone instances of the original bitcoin! No disruptive new feature (not even the notorious covenants support ), nothing other than some glues, suture, cesors,etc. I mean few protocol stuff and features for interactions and relaxation of constraints suchas being backward compatible with old technologies and so fort, still, nothing other than our Plain Old Bitcoin, essentially. I'm not a genius, you know, just a regular person who is obsessed with integrating and things using glues.  

So, how is that impressive, fellas? The idea of using bitcoin as its own sidechain?

[garlonicon]

The idea of using bitcoin as its own sidechain?

Do you mean Layer Zero? Because normally, Bitcoin is the Layer One, LN is the Layer One and a Half, and typical sidechains or federations, are on the Layer Two. Because the Layer Zero would mean the layer above the mainchain, where the currently existing chain would contain more things (exactly in the same way as LN or typical sidechains contain more transactions, that are finalized on-chain in some batched form). The idea behind that is simple: if sidechains are good, and if they have some nice properties, then what would happen if the existing Bitcoin blockchain would be treated as the sidechain of some other chain?

I thought about something like that, based on block headers, as a way to decentralize mining, but the main problem is to get the right performance. Because obviously, receiving all blocks from all miners would work in theory, but may not turn out well in practice. On the other hand, lazy verification is also an option, and could be considered as a some kind of shield from the 51% attacks: if the honest miners own only 10% of the total mining power, then they should receive only 10% of those coins, and the rest should be treated as unspendable/frozen in such model, exactly in the same way as invalid signatures are rejected. In this way, the heaviest chain of Proof of Work is always traced, but only the heaviest valid chain is spendable, it is only about being aware of the potential attackers, and burning/locking some coins, to have rewards, proportional to the total work.

[aliashraf]

The idea of using bitcoin as its own sidechain?

Do you mean Layer Zero?  Because normally, Bitcoin is the Layer One, LN is the Layer One and a Half, and typical sidechains or federations, are on the Layer Two.

LN is not a sidechain, AFAICT, but, I'm ok with labeling bitcoin (its blockchain to be more specific) as  Layer 1 hence its immediate  sidechains which are going down as Layer 2, Layer 3, and it is good labeling practice to name the immediate above side chain as  Layer 0, even going higher to Layer -1, Layer -2, etc.
My universe, is two-dimensional, you are free to put blockchains above or under bitcoin, it is an advanced topic, tho.

For now, the question is about the specifications and nature of these blockchains: Is it feasible and (more importantly) good practice to put another bitcoin above or a number of other bitcoins below the current bitcoin we have?

Take RSK for instance, it iss a sidechain below current bitcoin blockchain, it is merged mined with zero inflation, good, good, but it is following Ethereum's stupidity in terms of concepts, i.e., (wreckless)Turing Completeness and (too)Smart Contracts, besides a majority of its codebase. I just can't help it not to see the founders as being distracted and de-railed by misconceptions surrounding sidechains. Presumably, they have felt an obligation to comply with the idea of sidechains as some experimental, free to-do-anything or to-be-anything, whatsoever and decided to follow the trend.

I'm asking this question to relax sidechain development from such a prerequisite to be (or not to be) very different than bitcoin, from there we may be able to proceed to the next step: recursion.

I thought about something like that, based on block headers, as a way to decentralize mining, but the main problem is to get the right performance. Because obviously, receiving all blocks from all miners would work in theory, but may not turn out well in practice. On the other hand, lazy verification is also an option, and could be considered as a some kind of shield from the 51% attacks: if the honest miners own only 10% of the total mining power, then they should receive only 10% of those coins, and the rest should be treated as unspendable/frozen in such model, exactly in the same way as invalid signatures are rejected. In this way, the heaviest chain of Proof of Work is always traced, but only the heaviest valid chain is spendable, it is only about being aware of the potential attackers, and burning/locking some coins, to have rewards, proportional to the total work.

Very thoughtful insights, as usual you bring meat to the table. Too advanced though, and I share two points, as a way of showing my appreciation.

1- Don't be obsessed with verification, blocks are not spontaneously mined by temporary miners. In practice there is a stream of blocks flawing from the same source, it suffices to have partial verifications constantly and full verifications, occasionally.

2- Using a recursive model realized in multi layers, grown up and down the reference blockchain, each chain is in charge of maintaining its own blockchain and (pruned versions) of the ancestors (not the descendants) it is how the overhead  can be afforded.

The key is recursion, where scaling and decentralization converge.