[2015-12-07] Gavin Andresen Explains Why He Prefers BIP 101 Over BIP 100

[ezak]

Gavin Andresen Explains Why He Prefers BIP 101 Over BIP 100

There have been a few proposals to increase the maximum blocksize in Bitcoin over the past year, but the two plans that are garnering the most attention right now are Jeff Garzik’s BIP 100 and Gavin Andresen’s BIP 101. While Andresen has already implemented his plan in BitcoinXT, Garzik is still working on a formal write-up and the code for BIP 100.

Andresen shared his thoughts on BIP 100 via an interview on Epicenter Bitcoin. Although he did not seem completely dismissive of the proposal, it’s clear that he would prefer to go with his own idea for increasing the blocksize, which would offer more predictability in regard to where the blocksize limit will be in the future.

How BIP 100 Works

When asked directly about BIP 100 by Epicenter Bitcoin co-host Sébastien Couture, Andresen was quick to describe the basic way in which the concept is intended to work:

“Jeff [Garzik]’s proposal is that basically the miners get to decide what the next maximum blocksize should be. There’s kind of a regular vote every -- on the order of months.”

Andresen then discussed both BIP 100 and BIP 101 in terms of which proposal would lead to a larger increase in the blocksize limit. The Bitcoin Foundation chief scientist noted that Garzik’s proposal does not come with an immediate increase after it has been implemented, but this does not necessarily mean that it would limit the size of blocks over the short term:

https://bitcoinmagazine.com/articles/gavin-andresen-explains-why-he-prefers-bip-over-bip-1449506700

[1715004011]

1715004011

[1715004011]

1715004011

[1715004011]

1715004011

[1715004011]

1715004011

[1715004011]

1715004011

[Denker]

Gavin can prefer BIP 101 as much as he wants.
I'm sure the we will go a completely different way.
With segregated witness there seems to be some great oportunities ahead of us.This is combination with an conservative increase in Blocksize should give us more than enough time to develop future solutions.

[coinyard]

Whatever the argument about the future blockchain size and how to implement, we need at least 2MB by the end of next year.

[Carlton Banks]

Whatever the argument about the future blockchain size and how to implement, we need at least 2MB by the end of next year.

Why 2MB, and not 4MB? (my prediction is that your reason is pretty arbitrary). What if we could increase the amount of transactions per megabyte? What if that meant only 1.75 MB blocks were needed to handle the load?


Bigger blocks is not the only way to increase the transaction rate, which is entirely the point. If bigger blocks couldn't fit any extra transactions in for some reason, presumably you'd still be really keen to double the resource requirements for no reason at all?

[TraderTimm]

Gavin's "arguement" is based on consensus not being "predictable".

Uh, what?

A dynamic environment like fees/blocksize was never meant to be static and predictable. If he wants some over-arching master plan of order, he's better off doing something else than working on Bitcoin. Honestly, I don't know where his head is at times...

[virtualx]

Gavin's "arguement" is based on consensus not being "predictable".

Uh, what?

A dynamic environment like fees/blocksize was never meant to be static and predictable. If he wants some over-arching master plan of order, he's better off doing something else than working on Bitcoin. Honestly, I don't know where his head is at times...

Wasn't Bitcoin XT an experment with Gavins ideas? Why is he pushing for a protocol change of the bitcoin core chain?

[The00Dustin]

Wasn't Bitcoin XT an experment with Gavins ideas? Why is he pushing for a protocol change of the bitcoin core chain?

The blockchain is neither owned nor controlled by bitcoin core and should not be referred to as the "bitcoin core chain".  The bitcoinXT client runs as a full node on the same chain and follows the same rules.  It includes BIP101 support, but does nothing different and is not "an altcoin" as so many people like to argue.  In order for BIP101 to become a reality, it would need to be supported by a majority, and if this happened, the same chain would suddenly be running BIP101 and bitcoin core would almost certainly be updated to support this before that happened in spite of vocal opposition coming from some core devs.  To be fair, the bitcoinXT client contains some other changes that some people don't like.  That along with lots of fear mongering and perhaps some other legitimate reasoning makes it very unlikely that there will be a BIP101 majority running it.  However, anyone can run any full node client that supports the same rules as the majority, so people technically have the option of modifying the core code and compiling themselves and/or switching clients multiple times in order to support some changes but not others.  Then again, most of the BIPs seem to be supported by miner votes and not client counts (which it may be possible to manipulate), so many of the points in this post may be moot.

[coinyard]

Whatever the argument about the future blockchain size and how to implement, we need at least 2MB by the end of next year.

Why 2MB, and not 4MB? (my prediction is that your reason is pretty arbitrary). What if we could increase the amount of transactions per megabyte? What if that meant only 1.75 MB blocks were needed to handle the load?


Bigger blocks is not the only way to increase the transaction rate, which is entirely the point. If bigger blocks couldn't fit any extra transactions in for some reason, presumably you'd still be really keen to double the resource requirements for no reason at all?

2MB or 4MB is just a limit. It is not a target. I think 4MB should be safe for the next two years.