Quote from: NeuroticFish on January 23, 2017, 08:44:26 AM

Quote from: deisik on January 23, 2017, 08:24:11 AM

Quote from: NeuroticFish on January 23, 2017, 08:19:53 AM

If dynamic block has some problems, I think that the best option would be to overcome/fix them. An option would be to set an upper limit of the dynamic block, at least it would reduce the flood problem

Then blocks would be no longer dynamic

In other words, with the upper limit set on the block size the current implementation of Bitcoin can be thought of as dynamic too. Some rogue miners choose not to include any transactions in the block they find (apart from the block generating transaction), some include only the transactions with the highest fees, and the size of the blocks is only a few kilobytes (or even less than that). In this manner, blocks are as dynamic as they could possibly get

You do have a valid point. But I see it only partly valid.

An older discussion was that some miners cheat or something is defective and until that is fixed any other solution is only a lie. It happened at one of the first big spam attacks I know of.

From what I know, now the block size is fixed. So whether the miner includes 0 transactions or max possible, the block size will be the same.
I see the dynamic block ... dynamic. If it has an upper size, we will have the same problem as now sometimes, if there will be huge spam attacks (if the upper limit is 16 MB or 64 MB, you can imagine..)
In the normal days, the blocks will be like now, even smaller sometimes, leading to a not-so-big increase of the blockchain over time.
Does this sound ok or is there something extremely wrong in this logic?

I was actually me who had raised this issue why there were blocks half-empty or just empty when literally many dozens of thousands of transactions had been queued and didn't get confirmed in time. Some guy (don't remember his name) who is behind some mining software came up with an explanation that it was due to misconfiguration on the miners' part as well as deliberate actions of rogue miners (allegedly for the sake of efficiency). Regarding fixed block sizes irrespective of the block content (i.e. how many transaction it includes), this simply makes no sense and creates useless or even detrimental overhead...

Really, what's the point of adding empty space to the block just to keep its size to 1Mb in all cases?