Quote from: cellard on January 02, 2018, 06:24:01 PM

Quote from: Samarkand on December 21, 2017, 03:49:44 PM

Quote from: nullius on December 21, 2017, 12:51:39 AM

...
Well said, except that 0-conf was never viable; and thats not a blocksize issue.

...

I agree that 0-confirmation transactions were never really secure. However,
many operators had a business model that was based on 0-conf transactions
(e.g. the now defunct betting site Directbet). Therefore something seems
to have changed, because otherwise these operators would still be able
to use 0-conf transactions.

I can´t imagine that this is only due to increased fraud by people attempting to double-spend
a gambling transaction (e.g. double-spend after their bet already lost on Directbet). There has
to be a technical explanation why 0-conf transactions are not widely used anymore
when that was still the case in 2014-2015 (at least at Bitcoin gambling sites).

Satoshi said that 0 confirmation transactions were never to rely upon, or at least he literally considered 0 confirmation users as "second class citizens" in the bitcoin network:

It's not clear to me if he had 0 confirmations to be part of the network model. Big blockers often argue that 0 confirmations are perfectly fine and the only reason they aren't now is due the blocks not being big enough. 0 confirmations never made any sense to me. Maybe for smaller transactions, but I would rather trust an LN channel than a 0 confirmation transaction.

0 confirmation is very secure & trustable for micro-transaction as long as the user put enough fee to make sure transaction can be confirmed quickly. In past many services, especially on-chain bitcoin casino/gambling accept 0 confirmation.

Quote from: cellard on January 03, 2018, 04:07:10 PM

Quote from: stantpro on January 03, 2018, 02:29:08 PM

Does the absence of 8 MB block imply inadequacy in decentralized crypto currency or does it make it even
more centralized?If the coders are reliable, does the 8 MB lag pose any threat that could raise any doubt of
non-performance?

Increasing the blocksize ALWAYS come with a centralizing factor, since the rule of "the bigger the blocksize, the harder to host your own node" always applies, therefore, you want to avoid raising the blocksize for as long as possible.

Reducing on-chain size with segwit and Schnorr sigs et al are ways to go around this, some consider anything but legacy transaction not safe, then again, if you don't consider that safe, you can't be a BCashier pretending BCash is a better idea.

If i'm right, SegWit and Schnorr sigs supposed to reduce transaction size, not on-size (aka block size). CMIIW.