Quote from: arklan on March 04, 2016, 05:08:58 AM

Quote from: Cconvert2G36 on March 04, 2016, 04:21:12 AM

Honestly, I'd rather just be posting memes and gifs while celebrating the halving, confident about growth after having added some modest capacity...

WRT Blockstream... it wasn't until after Scaling HK, when gmax's email to the mailing list literally became the Core roadmap, with no hard fork increase in sight, that the pieces began clicking into place. Others had already described the whole thing as a charade and a stalling tactic, but I was actually hoping for some kind of unifying plan, I should and do feel silly for that now.

Soft fork segwit as a "scaling solution" bifurcates the capabilities of full nodes without their knowledge or permission, completely changes fee economics with favoritism to settlement transactions, and apparently is having mysterious chain forks on its testnet a month before deployment... while being sold as the safe and "ready" option. [Shrugs] The pieces certainly haven't stopped fitting yet.

well, they can't exactly force us to upgrade our clients. i'm very much a "wait a see" kind of guy. if the release of segwit comes and things go kaboom, there will still be a record of how things were before that. if it's really bad, it's technically possible to revert to the pre-forked version and carry on, i think.

of course, "technically possible" and "easy" or even "kinda doable" are vastly different things, and frankly i might be talking total bullshit anyway. i'm not totally clear on the whole "soft fork" thing.

Unless we plan on keeping the 1MB cap always and forever... any HF to increase it will "force" you to upgrade your client, as it will reject any blocks from miners over 1MB. Soft fork segwit doesn't require anyone to upgrade, but it means your formerly full node client will not be verifying segwit tx signatures, as they would now be in the witness portion of the block, which your client doesn't understand. It becomes a sort of half-node. In the case of segwit tx, it sees them, but can't verify the sigs itself. If you hadn't been paying attention, you might not even know it had happened. In the case of a HF, you know, it's unambiguous that it's time to upgrade.

Segwit itself is actually a pretty cool idea, it has the big positive of ridding us of tx malleability problems altogether. I'm not against a version of the idea at all. I argue against aspects of it, like the soft fork deployment, the fee discount for sig heavy settlement tx, and that it's a fairly complex change to the system possibly being rushed and inadequately tested. I worry that the additional capacity it could bring is being dangled like a carrot to get us to turn a blind eye to the economic favoritism it carries, and it could stall out a HF blocksize increase way out to mid or late next year.