Quote from: jgarzik on March 12, 2013, 06:38:36 AM

It is always entertaining to watch non-contributors opine about completely obvious solutions that the devs are silly to have overlooked.

The interesting thing about bitcoin is its organic nature. The bitcoin codebase, warts and all, was dumped into the Internet's collective lap. Reality does not give anyone a chance to pause, wait for a specification to be polished, to wait for every single edge case to be tested (if that were even possible), etc.

Almost half a billion dollars in market cap, and the dev team is still largely unpaid volunteers, trailing behind events, cleaning up the messes reality leaves behind.

I am not sure how to read your post. Are you suggesting the current state of affairs is a good thing or a bad thing ? Obviously the original Satoshi client was a fact on the ground. And "reality" hasn't given anybody a chance to pause, but can we not take a small time out now and actually define a 1.0 spec and THEN implement a backwards compatible version from 0.8 today to that 1.0 ?

It's not like every pull request on GitHub is a do-it-now-or-bitcoin-dies urgent fix, actually most of them are trying to refactor, fix & alter the existing single implementation, and some like pruning and bloom filters add huge new features with *more* latent semantics sprinkled around the C++ instead of being in the spec.

I hope the gods on mount Bitcoin Foundation are listening once in a while, but Bitcoin's monoculture of a single "magical" C++ implementation whose code serves as the spec cannot go on forever.

I wanted to create an independent implementation from scratch in Scala, but soon realized that not only there wasn't enough documentation for protocol messages on the wiki, but worse, their semantics were constantly moving targets, potentially with every github pull request being committed. At that point I gave up.

I haven't heard of a from scratch C-based libbitcoin library much lately ... not sure if its author also had to throw in the towel after a while ... that would just be sad.

Guys, the time for heroics is great, but if you want Bitcoin to be Money of IP (as it's being advertised) then it has to be a network/message protocol spec that people can go away and implement in language X and interoperated with you.

In this Bitcoin will either follow in the successful tradition of TCP/IP, SMTP, HTTP, TLS, etc or it will be a series of heroic battles by "the community" to "save bitcoin" from future "genetic defects" found in its sole C++ implementation or one of third party libraries.

Cheers ...