News for hackers. Stuff that matters.
Re:Zynga Links Up With Bitpay For a Bitcoin Payment Test in ... by jnbiche on Saturday January 04, 2014 @09:51AM (#7011174)
The network split had nothing to do with transactions per hour. It was due to a very subtle difference between two versions of BerkeleyDB used in different versions of the Bitcoin client. Basically, a perfect storm:

http://www.reddit.com/r/Bitcoin/comments/1a51xx/now_that_its...

Re:Zynga Links Up With Bitpay For a Bitcoin Payment Test in ... by makomk on Saturday January 04, 2014 @10:07AM (#7011238)
Not really. The issue with BerkeleyDB was actually triggered by a new version of the Bitcoin client that didn't use BDB to store the blockchain and that simultaneously raised the number of transactions allowed in a block so that more could be processed. Older BDB-based versions of the Bitcoin client hit a previously unknown limit when attempting to process that many transactions, causing them to fork onto their own version of the network. The issue hadn't been noticed before because the number of transactions was previously restricted to a low enough level that it couldn't trigger the bug.
Re:Zynga Links Up With Bitpay For a Bitcoin Payment Test in ... by ewillbefull on Saturday January 04, 2014 @11:14AM (#7011522)
> and that simultaneously raised the number of transactions allowed in a block so that more could be processed

Source? I don't believe the hard limit in blocks has ever been changed; bitcoin's devs are currently struggling to convince people to let them do it. The soft limit will NOT cause hard forks.

The only reason that fork happened was an unknown limit that existed with one database in one version and which didn't exist with an improved database in another version. It was a bug.

Here's an evaluation of the chainfork: https://github.com/bitcoin/bips/blob/master/bip-0050.mediawi...