Quote from: hdbuck on November 26, 2015, 10:05:38 PM

Quote from: jtoomim on November 26, 2015, 09:52:42 PM

Quote from: gmaxwell on November 26, 2015, 12:56:38 AM

Actually, no, they've been having huge problems with it; with nodes crashing all over the place and such. Of course: Bitcoin Core nodes on testnet are unaffected: They're just ignoring the XT chain entirely, banning those peers, and continuing on as if they didn't exist.

We did have some trouble with a couple of features that I was testing out in parallel -- namely, Mike Hearn's alpha thin blocks patch (which works 99% of the time, unfortunately) and the BIP65 + low-S patch for BitcoinXT (which has some problems when heavily reorging and removing a block that contains either a high-S transaction or an invalid OP_CLTV instruction). However, these are unrelated to BIP101 and large blocks. So far, as far as I know, we have not had any crashes or problems due to large blocks themselves.

it is not about crashes..

Crashes are rarely a show-stopper. Most of the time such things are figured out fairly quickly and some stability is achieved (but not always.)

I believe that I did predict a quarter or two ago that the XT-ish ilk will need to find an excuse to switch over to Conformal's 'btcd' protocol implementation. A no-brainer for this would be 'inherent instability' in the Satoshi-based C++ implementation.

It was around three months ago IIRC that Gavin made a quip about learning go-lang. I remember that when he did I already had my prediction made about the bloatchain fork, btcd, (and etc) so his comment simply added some more confidence to my prediction about how the takeover attempt would be going down.