It is close. My estimates were for a 'moderate' size post or comment of 1-2 KB. For larger or smaller the numbers will vary. The per byte charge is pretty low and dropping, so the effect of size is becoming smaller.

I'm not sure I've left a sub-2k post or reply more than 20% of the time I've been on this platform. Now, part of that is just because I'm comfortable with being verbose but part of it is because I like to actually have back-and-forth engagement.

The current per byte charge is fairly low and dropping, but that could be entirely different tomorrow. In fact, if other people's estimation of the impact of costs for node hosting are even in the same ballpark of reasonable, and they seem to be, then the current 1/10 RC cost adjustment is long-term untenable to do what the system is supposedly intended to accomplish – actually charge people in line with the amount of resources that they consume for using the platform. Given that the per byte charge has to include an assumed future cost not just for uploading the content and replicating it but serving it, it probably has to be adjusted upwards if we are to believe the system is to function as it's described.

As far as I'm concerned, that whole cost could go away – but it's not going to.

These aren't currently adjustable by witnesses. The rules are fixed (absent a software upgrade), although the prices are determined by market/usage conditions so in that sense you can't know in advance, only estimate. C'est la vie.

That's funny. The most recent post by Steem Inc. was fairly explicit that the 1/10 RC adjustment was achieved without having to do a blockchain replay because the transaction costs were witness adjustable, and in fact were one of the things that witnesses are responsible for coming to a consensus on. Perhaps I misinterpreted or they misspoke, but it seemed pretty clear.

It would certainly be in line with the current design of how the witness node configuration is set up for such values to be witness-adjustable, so much so that it would be remarkable and noteworthy were they not. This might be an issue that you want to double check.

The relative unpredictability is one of the things that makes it less appealing for new users to engage with the system. As it stands, and as far as I know, no client which deals with engagement with the blockchain provides an estimated cost versus the resources that you have in your pocket before you commit a post or reply. This is sort of like going to the grocery store, leaving your wallet by the door, picking up stuff without prices marked, and walking out, hoping that you had enough in there so that they'll let you leave.

Or, in other words, it's a situation that no sane person would engage in if they care about whatever it is they're doing. And it's a scenario that Steem Inc. knew about as long as they've known what they were going to implement in hardfork 20 and should've known that it was a bad design decision multiplied by not getting engaged with the popular clients in getting them on board with setting things up so that user expectations and user experience was maintained and cultivated.

The pool formulas have a half life of a week, and they were pre-filled to 90%. As such it won't take that long. There is still some adjustment going on as things balance out but it isn't enormous. When the 1/10 as applied there was an immediate dramatic effect (15 SP accounts were limited to 1-2 comments previously). There will be further adjustment to equilibrium but only around the edges.

They were prefilled to 90%, and that was in no way anywhere near what was necessary to allow continued function of the blockchain, which should have been obvious from even cursory testing. Others have mentioned that some of those involved with the testing provided feedback that the pool formulas were flawed and functioning wouldn't be able to continue.

And yet…

It's impossible to tell me that "there is still some adjustment going on is things balance out but it is an enormous" on the heels of costs being modified by an order of magnitude and people still reporting that the vast teeming underclass still has insufficient access to participate in what is ostensibly intended to be a social network.

You are absolutely correct that when the 1/10 was applied there was an immediate dramatic effect – but everything that has come down from on high suggests that the more restricted resource environment is what is intended because some of the side effects are thought to be beneficial. Telling me that "there will be further adjustment to equilibrium but only around the edges" in the wake of an order of magnitude adjustment to costs are just words.

They mean very little to me as a user, but they fly directly in the face of experience for someone who was developing a digital application for this blockchain. Having an entirely new set of mechanics dropped on and dropped in, breaking content generation for a week, adjusting costs by an order of magnitude along the way, and then having someone say "further adjustments are coming but I'm sure they're going to be small" is high comedy.

I'm firmly of the belief that more changes are inbound but they won't be just around the edges. The entire mechanic has to be examined and a much broader description of intent has to be published if anyone wants investors or application developers to be secure in deploying their resources to this blockchain.

That's not my problem, thankfully. I would've done my best to avoid this in the first place.