chore(sdk): improve usability tx primitive traits

[emhane]

Implements Transaction for SignedTransaction types

[mattsse]

don't fully get the motivation here, doesn't this effectively just extend the transaction trait with an AT for txtype?

[mattsse]

I don't think we want a default here

[emhane]

it's derived for reth_primitives::TxType, so we probably wan't to remove that attribute next before integrating the generic tx type, in order to keep scope meaningful

on the other hand, if a tx type is unknown, don't we fallback to try to decode it as a legacy tx? hence it's kind of like a default? @mattsse

[mattsse]

do we still need this?
unclear why we need 3 traits now if this trait is now just a default

[emhane]

I thought about moving the rlp encoding traits to FullTransaction...to make primitive traits more lightweight. I'm also good with removing the Full-traits for the data primitive traits and adding Compact as a super trait always. wdyt? @mattsse

[emhane]

@joshieDo ?

[emhane]

In essence, the new trait AlloyTransactionExt can be removed with #12465 and moving is_dynamic_fee down a level into alloy_consensus::Transaction.

[emhane]

imo we can add SigneableTransaction as super trait like I did initially, since all the tx types we now import from alloy, already impl this trait @mattsse

[emhane]

don't fully get the motivation here, doesn't this effectively just extend the transaction trait with an AT for txtype?

it add blanket impl of Transaction for SignedTransaction types, as pr description says. since the type reth_primitives::TransactionSigned which generic T: SignedTransaction aims to replace, defes to reth_primitives::Transaction, which in turn is aimed to be replaced by generic T: Transaction - this facilitates integration, since we do not have to call .transaction() then, everywhere we have before just called reth_primitives::Transaction methods right away on the reth_primitives::TransactionSigned type.

[mattsse]

last nits

[mattsse]

lgtm