If blocks will only contain an IBLT can the block time decrease?

[Robert Paulson]

since all blocks will have the same small size no matter how many transactions they represent would it not be a good idea to decrease the block time from 10min to something really low like 30 secs?
the block propagation time will be alot smaller than 30 secs and it would allow getting a confirmation alot faster.

[1714656338]

1714656338

[1714656338]

1714656338

[1714656338]

1714656338

[dabura667]

The block time is less about propagation (though low block times DO run into problems with propagation, yes) and more about the distribution of solutions over time.

Super fast propagation would help alleviate orphans in a low block-time environment, but would not eliminate it. The decrease from 10 min to 30 sec would increase the orphan rate FAR more than the IBLT would reduce the orphan rate.

[Billbags]

Just when I think I'm getting a hold of the current system, you have got me with this one. What is IBLT?

[teukon]

Just when I think I'm getting a hold of the current system, you have got me with this one. What is IBLT?

Invertible Bloom Lookup Table.

Most of the block-generating peers have most of the transaction data that you have and will be using the similar rules to determine which transactions to include in a block.  Given this, it seems wasteful to send the entire list of transactions of your block to each peer.  Instead you send an IBLT which contains enough data for your peers to, with high probability, construct the precise list of transactions in your block.

At least, this is my understanding.