Steem pyramid scheme revealed

[iamnotback]

Smooth much better if you don't cash any out and buy more STEEM POWER. Why not put your $ where your confidence is?

I'll consider it

Anyway, I never claimed Steem is perfect nor that it will clearly succeed, but it does have a big first mover advantage, isn't terrible, has decent size user base for a crypto (even factoring in the large number of scammer accounts) and is still growing. Thus, I don't think it is certainly doomed. Uncertainty prevails.

The big uncertainty is whether they will remove the quadratic vote weighting and remove the whales' voting power. If they don't do that, they are dead. And if they remove the quadratic voting, they shift their target demographics significantly so there is a lot of risk in that change/transition.

Even if they do that, they are still vulnerable to competition, primarily because Bitcoin investors will prefer one that wasn't so egregiously "pre"-mined and also one where the blockchain is not prevented by license from being forked.

And also it seems there is no way it can be defended against a competitor which removes the stupid 50% yearly (currently 150%!) debasement of medium-term investors. Which I have a design that can do that.

In short, they probably can't make such big changes.

Those are the big picture reasons, that you apparently do not see clearly yet, but I do.

Also there are other issues that I have been mentioning such as better relevance and scaling design.

Moreover, why don't you buy any? As far as I know, you've never bought any significant amount of STEEM POWER. It was all "gifted" to you (okay some effort to research and mine it, also to build your reputation which made you a good asset for them to coddle).

All of us have our confirmation biases. Yours is that you didn't risk anything consumerate to obtain your stake and mine is obvious that I want to think I can compete.

[1714988886]

1714988886

[1714988886]

1714988886

[smooth]

Smooth much better if you don't cash any out and buy more STEEM POWER. Why not put your $ where your confidence is?

I'll consider it

Anyway, I never claimed Steem is perfect nor that it will clearly succeed, but it does have a big first mover advantage, isn't terrible, has decent size user base for a crypto (even factoring in the large number of scammer accounts) and is still growing. Thus, I don't think it is certainly doomed. Uncertainty prevails.

The big uncertainty is whether they will remove the quadratic vote weighting and remove the whales' voting power. If they don't do that, they are dead. And if they remove the quadratic voting, they shift their target demographics significantly so there is a lot of risk in that change/transition.

Even if they do that, they are still vulnerable to competition, primarily because Bitcoin investors will prefer one that wasn't so egregiously "pre"-mined and also one where the blockchain is not prevented by license from being forked.

And also it seems there is no way it can be defended against a competitor which removes the stupid 50% debasement of medium-term investors. Which I have a design that can do that.

In short, they probably can't make such big changes.

Those are the big picture reasons, that you apparently do not see clearly yet, but I do.

Also there are other issues that I have been mentioning such as better relevance and scaling design.

The main thing I disagree with there is that Bitcoin investors are almost entirely irrelevant. If it doesn't gain interest and investment from outside of cryptocurrency circles, then it is useless and will fail. That is the entirely of the value proposition here; I'm surprised you don't see that.

You are certainly incorrect about scaling, at least on the front end portion. I'm not sure whether the blockchain can scale to the necessary degree without redesign, but that certainly won''t be a problem for quite a while.

[iamnotback]

The main thing I disagree with there is that Bitcoin investors are almost entirely irrelevant. If it doesn't gain interest and investment from outside of cryptocurrency circles, then it is useless and will fail.

Bitcoin investors are the only source of capital while it grows from a small seed to the point where investors outside our sector would come it. We can't just jump across the ocean in one leap.

This a blockchain and crypto-currency. The investors must be comfortable with those concepts.

If you are thinking about big time funding rounds such as what Coinbase received, the problem is that the terms of their investment would be ruled by the Steem blockchain (which was not the case when they invested in Coinbase equity), which is not something traditional investors may be comfortable with. You can't just leap across the ocean.

You are committing a cardinal sin of marketing, by assuming that people change quickly. The key in marketing is to find an untapped synergy where people don't need to change what they already know and want to do.

You are certainly incorrect about scaling, at least on the front end portion.

When I wrote about Websockets versus REST, the scaling issue is of course on the serving end. How can a front end UI ever be a scaling problem since it runs on every client decentralized  

I'm not sure whether the blockchain can scale to the necessary degree without redesign, but that certainly won''t be a problem for quite a while.

It is a problem if a competitor starts to grow at the 100 times faster rate that Ello did. In a month, Steemit's first mover advantage would already been irrelevant.

[smooth]

You are certainly incorrect about scaling, at least on the front end portion.

When I wrote about Websockets versus REST, the scaling issue is of course on the serving end. How can a front end UI ever be a scaling problem since it runs on every client decentralized 

Front end including the web server. You are still incorrect.

I'm not sure whether the blockchain can scale to the necessary degree without redesign, but that certainly won''t be a problem for quite a while.

It is a problem if a competitor starts to grow at the 100 times faster rate that Ello did. In a month, Steemit's first mover advantage would already been irrelevant.

Yes but the blockchain scaling isn't the issue there.

[iamnotback]

You are certainly incorrect about scaling, at least on the front end portion.

When I wrote about Websockets versus REST, the scaling issue is of course on the serving end. How can a front end UI ever be a scaling problem since it runs on every client decentralized  

Front end including the web server. You are still incorrect.

With enough capital, the web server can be scaled. You could read about some of the things Facebook had to develop in order to scale out.

But the devil in the details in terms of costs. A certain incorrect design decision could lead to 10 or even 100 times higher costs for the (database) servers. And that is what I am referring to.

The extra reason my point is so salient, is because afaik there is no funding for the webserver at all. Steemit.com is paying for it out of their "pre"-mine. So if the serving is 100 times more costly we are all paying for it (less funding for signups, etc).

Also scaling up to 100 million users requires special expertise. If your design is 100 times worse than it could be, they will hit their headroom sooner, putting pressure on their ability to source the expertise and figure it all out while at the same time being slammed with the exponential growth.

It is not an issue of it can't scale assuming all other factors are irrelevant. My point was about design decisions putting them behind in a race with a competitor which can scale faster, easier, and less costly.

Note I haven't evaluated their design in detail, so it is possible that my generalized assumptions have some flaw due to some detail I am unaware of. But generally speaking, what I wrote about Websockets vs. REST is the correct intuition. And I have some specific ideas in mind of how the content can be optimized in the REST case that impacts the database load by perhaps orders-of-magnitude (will know more as I dig into details of coding it).

I'm not sure whether the blockchain can scale to the necessary degree without redesign, but that certainly won''t be a problem for quite a while.

It is a problem if a competitor starts to grow at the 100 times faster rate that Ello did. In a month, Steemit's first mover advantage would already been irrelevant.

Yes but the blockchain scaling isn't the issue there.

For blockchain scaling they have at least the issue of transactions per second and DPoS's issues there.

Also they might have issues with the database employed by the witnesses that may be impacted by the decision to allow so many real-time updates to pre-existing data. Data fragmentation issues, etc..

Remember that Monero slams SSDs so I am not so confident (until you inform me otherwise) that you all are highly expert on this point. (not to say that I am either but I am thinking about the issues)

[AlexGR]

And also it seems there is no way it can be defended against a competitor which removes the stupid 50% yearly (currently 150%!) debasement of medium-term investors. Which I have a design that can do that.

Eh? They can convert STEEM => SDs => STEEM => SDs, as they see fit. SDs will even give them 10% interest.

[iamnotback]

And also it seems there is no way it can be defended against a competitor which removes the stupid 50% yearly (currently 150%!) debasement of medium-term investors. Which I have a design that can do that.

Eh? They can convert STEEM => SDs => STEEM => SDs, as they see fit. SDs will even give them 10% interest.

That is not investment in Steem. There is no upside ROI for the investor coming from speculation about Steem by investing in Steem dollars which are pegged to the US dollar.

[smooth]

Remember that Monero slams SSDs so I am not so confident that you all are highly expert on this point. (not to say that I am either but I am thinking about the issues)

The person who literally is an expert on that (with a decade of experience developing databases for SSDs), said that particular FUD from an extremely shady competitor was "full of shit" (direct quote). (I also refuted it separately by measuring that routine laptop use generates SSD write volume that vastly dwarfs Monero's). In fact, he replied to you specifically in detail as to why your inexpert concern trolling on the matter was displaying ignorance. So why are you repeating it?

Anyway that is off topic to Steem. The blockchain can clearly scale to orders of magnitude more than current usage. Whether it runs into problems at some point is an issue, but not one for the short term, even with fairly rapid growth.

[AlexGR]

And also it seems there is no way it can be defended against a competitor which removes the stupid 50% yearly (currently 150%!) debasement of medium-term investors. Which I have a design that can do that.

Eh? They can convert STEEM => SDs => STEEM => SDs, as they see fit. SDs will even give them 10% interest.

That is not investment in Steem. There is no upside ROI for the investor coming from speculation about Steem by investing in Steem dollars which are pegged to the US dollar.

Ah well, you can't have everything. There are 3 tokens and you know one of them is debased a lot, the other is protected significantly and the other is giving you 10% interest. If you want upside potential, then you'll either trade it when low, or ride the fluctuations in the SD/STEEM markets (selling SDs when 1 or above, buying below 0.9 / 0.8, etc).

[iamnotback]

Remember that Monero slams SSDs so I am not so confident that you all are highly expert on this point. (not to say that I am either but I am thinking about the issues)

The person who literally is an expert on that (with a decade of experience developing databases for SSDs), said that particular FUD from an extremely shady competitor was "full of shit" (direct quote). (I also refuted it separately by measuring that routine laptop use generates SSD write volume that vastly dwarfs Monero's). In fact, he replied to you specifically in detail as to why your inexpert concern trolling on the matter was displaying ignorance. So why are you repeating it?

I replied saying that I personally know people who've told me their laptop becomes unusable while syncing Monero, which I consider to be 'slamming'.

I remember his reply indicating that my points were valid and mentioning some of the ways to tackle those points. I had basically said need to always write to the end of the file and never insert data at random locations into a file. His stance was coming more from how to deal with databases that must do random writing. My stance was coming from eliminating random writing as a design goal. So perhaps you misunderstood my context (and I wasn't eager to point that out to him, since he reply went off on issues about random access databases which wasn't even my focus at all and didn't see any point in wasting my time). Paradigm-shifts can pull the rug out from under an expert.

Feel free to point us back to the post you have in my mind (it is in the Monero Technical thread afair), in case my recollection is different from what you think was discussed.

Btw, 'slam' is not referring to John Conner's accusation that Monero kills every SSD (or what ever he wrote). But it is true that with excessive random writing, the MTBF of the SSD will drop due to several factors of the way an SSD operates at the low-level.

Maybe I shouldn't give away so much information for free (versus being cryptic and having people underestimate me, which has its trade-offs), but as long as I am getting back information then it is worth it.

[smooth]

Btw, 'slam' is not referring to John Conner's accusation that Monero kills every SSD (or what ever he wrote). But it is true that with excessive random writing, the MTBF of the SSD will drop due to several factors of the way an SSD operates at the low-level.

Okay I misunderstood then. FWIW. LMDB does do most writing to the end of the file, not randomly. Of course any database must some point do some random writes, but it is minimized. I think the main reason for Monero having a high user-perceived impact is just that it is optimized to do a lot of precessing in a short term. When you take the sync time from several days down to an hour and do the same work, the intensity is obviously vastly increased. Probably some work is needed on backgrounding and throttling it better (i.e. take two hours on the same equipment at least while the computer is otherwise in use, but in a less annoying way)

I don't know the details of the database in Steem/Graphine but several different people have said they've tested it at 100x the current transaction volume of Steem and had no problems with ordinary hardware. That takes daily active users to about 4 million, or higher with better hardware.

[r0ach]

Remember that Monero slams SSDs so I am not so confident (until you inform me otherwise) that you all are highly expert on this point. (not to say that I am either but I am thinking about the issues)

I'm glad you brought this up.  A couple days ago I was syncing up a Bitcoin Core full wallet via USB3 to external USB3 mechanical hard drive.  The chain was already something like 95% done beforehand and I had synced it the same way many times before, but this time after like 10 minutes of going at it with the crunching sound you hear when doing intensive hard drive stuff, a Bitcoin QT has stopped responding message pops up and the drive just stops as every USB port on the computer also stops working (mouse and keyboard dead).  

After reboot, everything works perfectly fine again.  I have never once seen anything like this occur before in Bitcoin, but I often hear people bring up wild claims about Monero doing stuff like this.  The mechanical hard drive wasn't even hot or anything, but it seems like it could not handle the load from the Bitcoin wallet somehow.

[iamnotback]

I don't know the details of the database in Steem/Graphine but several different people have said they've tested it at 100x the current transaction volume of Steem and had no problems with ordinary hardware. That takes daily active users to about 4 million, or higher with better hardware.

Note by transaction rate tests you mention, I presume they are referring to computational load and not database load (since for one reason I doubt they even could simulate the real world data load patterns a priori). I am suspecting (not confirmed, as I didn't study the code) they may not be prepared for database load (this is usually the myopia of every blockchain upstart, even Monero went through this learning experience).

As for computational and network load issues, I think the issue is likely only if we want decentralization and performance.

Since I apparently realized upthread how the 19 witnesses can thwart DDoS using disposable IP addresses, it doesn't appear that the a priori round-robin attribute of DPoS is a blockchain DDoS vulnerability. As for reliability, if we assume the 19 witnesses are 100% loyal and have invested significantly in infrastructure, then reliability should be robust. All of this seems to require some coordination, so this is more or less a very centralized blockchain in terms of control. And it requires there not be disagreement. When disagreement comes, and the proof-of-stake is not 50% controlled by a few whales or the control of the whales diverges from the desires of the majority of users/investors (not a majority by stake due to power-law distribution), then chaos could result. But again that is not a near-term issue but I am not confident it is an anti-fragile design.

Afaics, the speed of transactions probably can't improve much lower than the current 3 seconds with DPoS, but that won't matter near-term.

[smooth]

I don't know the details of the database in Steem/Graphine but several different people have said they've tested it at 100x the current transaction volume of Steem and had no problems with ordinary hardware. That takes daily active users to about 4 million, or higher with better hardware.

Note by transaction rate tests you mention, I presume they are referring to computational load and not database load (since for one reason I doubt they even could simulate the real world data load patterns a priori).

I'm not. They ran the ordinary node which writes to a database. Real world load patterns are of course unknowable and I don't know how realistic any of these tests were but there is also a pretty (or very) wide margin between running on ordinary cheap unoptimized hardware and something actually optimized to host a database.

I'm pretty sure ability to handle only the computational load is much, much higher, but that is completely unrealistic, so pointless to even quote those "Wow!" numbers.

There is a deployment scaling issue in that the database not fully ACID and recovery from corruption is slow (more precisely will become slow if the tx volume increases; currently it isn't too bad). The only way to deal with that currently is with redundant systems and trailing backups. It is a pain in the ass, but workable.

All of this seems to require some coordination, so this is more or less a very centralized blockchain in terms of control.

Yes I agree there.

But again that is not a near-term issue but I am not confident it is an anti-fragile design.

Also there.

A couple days ago I was syncing up a Bitcoin Core full wallet via USB3 to external USB3 mechanical hard drive.  The chain was already something like 95% done beforehand and I had synced it the same way many times before, but this time after like 10 minutes of going at it with the crunching sound you hear when doing intensive hard drive stuff, a Bitcoin QT has stopped responding message pops up and the drive just stops as every USB port on the computer also stops working (mouse and keyboard dead).  

After reboot, everything works perfectly fine again.  I have never once seen anything like this occur before in Bitcoin, but I often hear people bring up wild claims about Monero doing stuff like this.  The mechanical hard drive wasn't even hot or anything, but it seems like it could not handle the load from the Bitcoin wallet somehow.

Bitcoin transaction volume has steadily increased. It may have finally gotten to the point where recent blocks just can't be indexed on cheap hard drives any more (using Bitcoin Core). i don't know what happened with the mouse and keyboard being dead though.

[generalizethis]

As far as interest versus numbers, does anyone have any numbers for Linkedin and twitter? These could give you a gauge into media and corporate interest. I'll put the current numbers here in case there is no alexa-like stat-machine:

(assuming these pages are official)

https://www.linkedin.com/company/steemit -- 115 (followers)

https://twitter.com/Steemit -- 1570 (seem to remember this a few hundred lower a week or so ago, but memory is well-known as a flawed content aggregate....)

[smooth]

SteemServices is not "official" it is a consortium of a few whales (not me) who do marketing and provide other supporting services for the platform. That's a pretty good number of followers for them at half the official twitter. I doubt they literally bought them but they may have run some sort of promotion to encourage it, I don't remember.

[generalizethis]

SteemServices is not "official" it is a consortium of a few whales (not me) who do marketing and provide other supporting services for the platform. That's a pretty good number of followers for them at half the official twitter. I doubt they literally bought them but they may have run some sort of promotion to encourage it, I don't remember.

*fixed

As far as the witnesses go, is that equivalent to Parker's servers? Or can they adapt in crisis?

[iamnotback]

Here is my response to smooth's Alexa point:

https://steemit.com/steem/@anonymint/steemit-s-skyrocketing-alexa-rank-on-the-road-to-world-domination

The performance of the above post will be yet another indicator of where Steem is good at displaying relevant blog posts or whether there is controlling hand that prevents objective criticism/questions. The fact that already one user commented that he didn't even know about Ello, indicates this is a very relevant blog for that audience. As for new audiences, well Steemit has to gain a lot wider usership before we can write to them. As the cosplayer girl discovered, her cosplaying audience was too small on Steemit at this time.

Also note this:

https://www.thesaleslion.com/numbers-rankings-alexa-waste-time-stupid/

[iamnotback]

I don't know the details of the database in Steem/Graphine but several different people have said they've tested it at 100x the current transaction volume of Steem and had no problems with ordinary hardware. That takes daily active users to about 4 million, or higher with better hardware.

Note by transaction rate tests you mention, I presume they are referring to computational load and not database load (since for one reason I doubt they even could simulate the real world data load patterns a priori).

I'm not. They ran the ordinary node which writes to a database. Real world load patterns are of course unknowable and I don't know how realistic any of these tests were but there is also a pretty (or very) wide margin between running on ordinary cheap unoptimized hardware and something actually optimized to host a database.

I'm pretty sure ability to handle only the computational load is much, much higher, but that is completely unrealistic, so pointless to even quote those "Wow!" numbers.

Since afaik, they haven't published the data in an easily accessible format, I have no way to confirm nor refute unpublished benchmarks.

But I note you refer to "transactions", and I am referring to edits of blogs and comments. They may have simulated adding these at 100X current rate, but perhaps not random editing of content updates to the blockchain data, i.e. random writes (and then random requests for data to the web server). Since random edits can be so much more costly (and also require a less optimal database design), they can have a big impact even if they are not the most numerous case.

Note I presume they can scale any design with enough capital to spend on infrastructure and engineering. But if this is a significant cost (and maybe it is not, I dunno), it will come out of the pockets of the users one way or the other. For example, it is not unfathomable that it might possibly cause a problem in that user's stake becomes insufficient over time and users have to buy stake to continue, but instead users might leave for the more efficient system that doesn't need to charge them.

When I refer to scaling, I don't mean that it is physically implausible to scale it. I am referring to how the cost and complexity grow through scaling (i.e. big O notation), and potential impacts thereof.

There is a deployment scaling issue in that the database not fully ACID and recovery from corruption is slow (more precisely will become slow if the tx volume increases; currently it isn't too bad). The only way to deal with that currently is with redundant systems and trailing backups. It is a pain in the ass, but workable.

Thanks for sharing that.

Note this is also a weakness of DPoS. If witnesses were instead fungible amongst all nodes, then it doesn't matter when a database is corrupted on any subset of witnesses.

[Divinespark]

Steem is getting rogered here, and on bittrex at least I am only seeing sell orders at the moment. Think the market will keep sliding from here. What surprises me is that synereo (as the alternative) hasn't really risen