It seems to me we are grappling with a multi-dimensional problem and there might not be any one "perfect" set of parameters and certainly what might have been good enough at one time could be less than good at a later time.  I shall attempt to list the dimensions;

• strategic intent
• functional goal
• durability
• stability
• robustness
• participants & their roles
• consensus
• PoW vs. PoS vs. TaPoS vs., etc. {what is the general name of this?}
• block size
• transaction fee
• block reward
• target time between blocks

I just bet there are others.  It seems unreasonable to expect everyone to agree on a common set of parameters across all of the dimensions and across time.  Many/most/all of these dimensions are interrelated.  Someone/anyone espousing a parameter choice for a single dimension risks being misunderstood/abused.

Given that no one set of parameters could possibly satisfy everyone, it is completely natural to have multiple solutions.  To resist this natural process and attempt to "force" everyone to agree is a disagreeable thing to watch.  Moreover, I think it will lead to disillusionment both within and outside the community.

No, I did not leave my wallet up.  I did see my transaction on Blockr.io, etc., so that gave me some confidence.

So, I recommend a reasonable fee but if you're in no hurry at all and can tolerate 11 days or more then it is still possible to get a no-fee transaction to go through.  I have no idea if they will eventually stop going through altogether.  So, be prepared to recover if they don't go through in the time frame you desire.

Oh, good point.  Hmm, eliminate those pesky block rewards?  Man, that would drive the fees sky high.  Hmm, let all the block rewards stand during a fork merge?  Violates the 21 million cap leading to more inflation/debasing.  Hmm.  Thank you for indulging my dreaming instead of calling me an idiot or something.

Is there any sense to letting forks live side-by-side for awhile?  Naturally the transactions in each fork would need to be unrelated.  Eventually we would need a block bringing forks back together when the history of transactions begin to intersect.

More dreaming.

In theory then a smaller block size (smaller even than 1MB) will promote decentralization?

Suppose we go with this idea and take it to the extreme.  Then perhaps a block should be exactly two transactions; the coinbase transaction and one other non-coinbase transaction (such that both fit together inside of 1MB?) -- um, that is, until the block reward drops to zero and then one and only one transaction gets packed into each block.  Naturally the highest fee transaction is the one going into the next block.  Hmm, maybe we should have these single transaction blocks coming out more often than once every 10 minutes (on average).

Just dreaming.

To me it makes sense to adjust things so that miners can remain profitable on block rewards forever but it's a little too early to really consider that closely yet.

Another approach would be to recognize the social value of Bitcoin and have the public subsidize mining (this will not be a popular approach with many of the Bitcoin community).

When the block reward dwindles enough to make mining unprofitable (ignoring fees for the moment) then the miners will face a crisis.  They could drop out leaving the business to others.  As they drop out the difficulty will decline; perhaps enough so that some remaining miners could still make a profit.  If the difficulty drops enough (highly unlikely) then I personally (along with a bunch of other enthusiasts) will fire up mining rigs and run them even if it is unprofitable (it entertains me/us to do so).

If we include fees then the story looks only a little different.  If the miners and users can reach a mutually agreeable fee then the miners can remain profitable on fees alone.  If not then some miners will drop out and we get the story above.

@DannyHamilton:  Naturally it is the low end of the range that is the relevant/interesting part.  The Inf is not especially helpful.  The low end of the range, e.g. 16, 23, 25, etc., represents a best case scenario.  My transaction took 11 days or approximately 1,584 blocks.  Certainly this is in the range but I'm trying to get a feeling for how many blocks to really expect.  If Bitfury (and other miners?) are still processing any no-fee transactions then that is interesting.  Do *any* no-fee transactions ever get through in at the low end of the reported range?  If the https://bitcoinfees.21.co/ guys reported 1584-Inf then that would help folks understand what to expect better.  If there comes a day when so few or no miners are packing no-fee transactions into blocks then I would hope the 21.co guys would report Inf-Inf or something to make it clear no-fee transaction are never getting through.

I didn't contact let alone pay Bitfury although I do appreciate them doing it.

Hey, what do you know, it went through.  So, some no-fee transactions are still getting through eventually.  Mine only took 11 days.

So, to recap; per Nash et al, Bitcoin can help guide/force fiat into being ideal money.  I just knew Bitcoin was good for something.

I do worry that the SegWit BU schism might undermine (ha!) the role of Bitcoin in this mission.

I recommend compromise.  Second best would be to just walk away from each other.

Is the ETH run up over or is there more to come?  I may pull out of Bitcoin altogether.

Let's suppose SegWit is pretty good stuff but is falling behind in the public relations department.  Is there a way to incorporate SegWit without forcing everyone to move their coins to new SegWit-compatible addresses?  Or am I confused?

I'm lazy and don't want to have to do anything.  Besides which there will be the fees to be paid to get my investment/holdings moved to the new addresses.  If I don't move them then eventually they will become valueless, right?

If all of this is right then Satoshi's coins will either have to move (interesting) or become worthless (interesting).

If forced to chose at this very moment then I would be inclined to chose BU.  I would much rather see the two camps compromise.

How much does the https://bitnodes.21.co/dashboard/ User Agents chart matter to folks?

Oh, thank you both for the sites.  If BitcoinChain.com and BlockTrail.com (and my full node) all still know about my no-fee transaction then are the miners still being told about it?  It would be great to be able to see the mempool of the miners.  If my no-fee transaction still has a chance of going through (even if it takes many days) then I am content.  I see at https://bitcoinfees.21.co/ that at least some no-fee transaction appear to go through (is that reliable/accurate/up-to-date?).  Can anyone point me at a recent no-fee transaction that did indeed get committed?

I run a non-mining full node (it entertains me to do so).  I'm delighted to feel like I am contributing in some small way (8 outgoing and 52 incoming connections --> ~200GB per month).  But, I realize if I turn if off then the miners will plow forward mining without me.  The users would have to connect to a miner (or some other non-mining full node).  *If* all of the non-mining full nodes were turned off then the miners would plow forward mining without us.  The users would have to connect to a miner, period.

I think maybe the point is miners don't just mine, they also handle connections from users.  If a miner just mined and didn't handle connections from users then they wouldn't have any transactions to pack into blocks.  I suppose they could do that but they can do that without me.  In fact, to the miners, my non-mining full node just looks like a user (well, except that I do forward transactions that aren't mine to start with).  In fact, miners look like users to other miners.  If a miner restricted their connections to just other miners (those that produced a kept block in the last little while) then they would be ignoring users but hardly suffer at all.  In fact, a miner could not forward transactions it receives to keep other miners from seeing them and so keep the transaction fee to themselves.

So, I guess unrestricted non-mining full nodes do serve a purpose beyond self-checking the blocks the miners announce.  They help tell all miners about new transactions to help avoid hoarding/selfish miners.

Maybe the users should deliberately avoid connecting to miners.

What's the big deal?  GBTC has been/is available right now.  Yes, the GBTC 2% per year fee is obnoxious and I would have preferred the 0.4% per fee on COIN but so it goes.  If any investors want in bad enough then they can already get in via GBTC.  The lower annual fee of COIN isn't that big a deal.

I am willing to wait many days for a no-fee transaction 74ccfd796cf2cb404a2245f699fd0e5cdf2a3a6b8a7b9cf1169ff2f87c124557 to go through but how can I tell if it is still in the mempool of the miners?  Neither Blockchain.info nor Blockr.io recognize it anymore.  If it is completely forgotten then so be it; I just don't know how to determine if it is or not.  If it is indeed forgotten then I suppose I will need to take steps to adjust my wallet accordingly.  Oh, btw, the non-mining full node I run still knows about it (as seen via the console using gettransaction).

I feel sad.  I want Core and BU to come together but I fear they won't.  I don't know what to do.  I have to wait until they do come together or split.  If they come together then I will be ok.  If they split then I will have to figure out how to move/split my investment.  This violates my desire to be lazy.

I run a 0.13.2 Core full node with 8 outgoing and 52 incoming connections and am thinking about going to 0.14.0 soon.  I have run a BU version in the not too distant past but I really want both SegWit and BU.  I suppose I could alternate every so often but that seems cumbersome.  I could run two full nodes but I just want to run one.  Would someone build me a version which runs both SegWit and BU?  Please.

I have GBTC and am holding; I don't want to miss a run up.  If COIN becomes a reality then I will evaluate jumping to it.