Topic: [ANN] Catcoin - 0.9.1.1 - Old thread. Locked. Please use 0.9.2 thread. - page 13. (Read 131044 times)

I'm inclined to let the incumbent team give another shot at getting the Catcoin client right, just because they seem to be motivated, and apparently has resources and volunteers. These count for a lot.

However, I also want to point out that there seems to be a steady stream of people who are getting kicked out one at a time, sometimes (but not always) for ideological reasons. It seems that the incumbent ideological kicker-out peson has returned from retirement. As I have pointed out previously, the ideological aspect of people getting kicked out discourages innovation. As a coinholder, I would not want these people who got kicked out to necessarily disappear into the sunset - I would want them to stick around in case the current team fails to again deliver promised results - because they may be needed at some point if a new team will need to be constituted to handle the situation.

Since there seems to be a coherent governing structure of sorts already with coders involved, it would be redundant to set up any sort of committee consisting of coders and coinholders. Maybe a stakeholder group can be formed consisting of only substantial coinholders who can discuss amongsts themselves and report back preferences on what in their opinion will maximize coinholder value (i.e., make the coins worth more in the future), which the existing team can take under advisement, and possibly can accept as an authoritative requests or even can accept as commands especially if the group represents the vast majority of the coins in existence. I don't think it's a stretch to treat coinholders as analogous to shareholders in a corporation, so if the owners of the majority of the coins vote to fire employees of the coin, that should come to be recognized as a legitimate prerogative. There seems to be no other objectively measurable criteria by which any commands regarding the future of the coin should be regarded as authoritative. Certainly obnoxiousness and loudness should not count.

If anyone reading this believes they are in the top 10 largest coinholder category (criteria I am proposing), I am happy to act as a coordinating person to set up such a committee. Anyone can join this committee by buying Catcoins on the open market or mining and holding sufficiently to achieve top 10 largest coinholder status.

Etblvu1

Here's the thing about profit switching.

If done *transparently*, it's a good thing, and if it's not a good thing, the code is broken, not the ethics of the profit switcher. This seems to be at the core of the ideological fork that has been brewing for a long time. My advice is don't run any code unless you see some evidence the developers at least have a mental model of how and why profit switchers do what they do, and how their code is going to give a similar coins-per-megahash return to both large profit-switchers and small miners.

And, please be aware that if you mine catcoin 'for the community' without regard for your own return you are subsidizing someone else's profits, and at least try to be aware of where it's going, or accept you may be subsidizing and feeding the things you speak most loudly against.

What has gotten me really angry in the past is how the catcoin market seems to respond more to mind viruses ( http://www.nytimes.com/2014/10/19/business/economy/when-a-stock-market-theory-is-contagious.html?rref=upshot ) than to what I believe the fundamentals are. For example, does anyone have any better reasonable explanation as to why we have an increase in price and an average hashrate of 200mhash? There are several theories I have, and one of them is that the miners and investors are quantitatively responding positively to discussions about code and the fundamental mathematics of what is a reasonable way to adjust difficulty.

Also beware of anyone who asks you to trust them. Read my code, and trust the mathematics, but it would be a very bad idea to trust some cult-like ideal of what I should be if you do not want to take a great deal of time to understand why I do things, and what motivates me.

I can write code, and when I'm posting here I apparently write market-moving mind viruses, or I'm feeding some other virus if I write when I'm pissed off, which is why I'd really like to have another forum for discussion. If you can find those people that meet those criteria you mentioned above, then let's get a new *public* forum going.

I am on board, always have been, always will be.  Though i cant write C/++ I can write in a few others (scripting lang's).  If people have questions on how to run the clients, mine, etc, I'm willing to entertain any questions.

I'm almost always around during the day on IRC.

I'm back, skillface, testnet is back, and as far as we're concerned, CAT is back.  The childish BS that used to take place in -dev is a thing of the past.

And absolutely - there will be progress updates here.

For the members of the community that are interested, testnet is rolling again after a bit of a summer vacation.  Now that we are not in an emergency, there's plenty of time to get this coin debugged the right way.

One of many changes being made is to remove ALL of Hozer's code.  One of the kludges we accepted because of the emergency of the last fork was a 30 second minimum block time.  This was done to ease the severe 'galloping' the algo was doing when we got hit hard.  That is not an acceptable long-term solution because it potentially dumps valid blocks.  The also needs to properly adjust to hash rate changes without penalizing miners or stealing blocks they rightfully earned.

I want to make sure everyone reads and understands that because Hozer decided to run hacked code on personal CAT nodes and on one of the main network's three seed nodes, that his node has been removed from the DNS tables for the network.  We have already removed his seed node from the catcoinofficial/catcoinrelease GIT but it's not yet been compiled into the clients.  This will be pushed for the forthcoming fork.

Additionally - we strongly recommend that the community NOT run any of Hozer's code.  None of it is official and none of it should be run on the main CAT network.  Hozer has decided that the way to 'fix' the coin is to kludge the code so the 'charts look good' - and he aims to do that by setting 'approved' block times closer to 10 minutes.  In other words, he expects miners to accept that most of their work will be dumped into the trash can - he expects miners to remain on the network even when they only get credit for SOME of the blocks they find.  As Blaksmith pointed out, this is the crypto equivalent of withdrawing $200 from a teller machine and only getting $100 or $50.  Put in very simple terms:  Hozer has decided that to fix the algo AND to attract 'quality' miners, he thinks we should agree to steal from miners.  After me, Blak, Zerodrama and others pulled out of shock and made it clear to him that this was completely unacceptable behavior, he took it on himself to hack mainnet. This is not a political disagreement, or one being done in secret by a closed group - this is a basic moral and ethical 'line in the sand' that the CAT development team are not now and will NEVER be thieves.  Because he violated our oath to NEVER run non-release code on mainnet, and because he thinks theft is a viable option that should be 'debated' by the team, he has been kicked and banned from the dev team and the dev channels.

So far on testnet, we ran 'pre-filter' function Hozer designed for the current PID.  Two rounds of his filter were run with his parameters and they failed.  I suggested a shorter set of moving averages for his filter and that failed as well.  What's being done now is that Blaksmith, the author of the current PID algo and a professional programmer paid to code PID for his day job, has re-written the PID to use different timings and some more sophisticated analysis of block time changes.  The first version of this is on testnet now.  It will take a number of days for it to get the rest of the data it needs, then we'll start abusing it.  More news on progress as it happens.  http://testnet.geekhash.org/

Bottom line folks - we believe in CAT, we believe in the trust of the algo, the block chain, the community, and the dev team - and we will not do anything to violate this trust, and we will not allow anyone in our group to, either.

Stay tuned for the best and most stable network yet - just in time for CAT's first birthday!

Andy

I am taking a look at this - looks interesting - but also am not sure if I will have enough time to run it through its paces enough to put it into production. But will see what I can do.

Etblvu1

I think there are a lot of legitimate stakeholders in Catcoin, and it is hard to quantify their stakes - some have put in a lot of sweat equity, but do not hold very many coins; some have put in a lot of effort but have very little mining hashpower; some have more massive mining power, but may have engaged in profit switching at is hard to prove.

My thought is that it is pretty hard to fake if someone has done any substantial coding on Catcoin or related projects; and it is hard to fake how many Catcoins someone controls. That is why I think these two categories should get representation of 3 each. People who have a stake in Catcoins who want into this group can work on buying up coins to get into it that way (thereby helping to capitalize the coin market), or get busy coding (or sponsor a coder who agrees to let the sponsor sit in his position; thereby helping to jump-start the catcoin development process that has been stagnant for awhile). And also, by making this group consist solely of coders and coinholders, does not mean everyone else is shut out - it just means others contribute suggestions which will be taken under advisement by the core group.

I think code contribution is easy to verify and measure; and current coin holding is easy to verify and measure. Does anyone have ideas for any other criteria that is easy to very and measure? Otherwise, I think 3 each of these two would do nicely for our core group.

Etblvu1

I think in all probability this aspect will not be changing any time soon. I think the focus right now will be on getting the block timings to regularize around 10 minutes.

Etblvu1

will cat stay at scrypt?

I'd be happy to assist *another developer* to implement whatever is needed in an MIT licensed reference client. I would probably even submit my patches under said license. But if I'm the only one writing code, it's going to have a viral copyleft license so I can charge people who profit from my work by including it in something else.

What I do find rather encouraging is the market response today, and the additional mining that has shown up. I'd like to believe the market price bump is because we have been reasonable in our discussions here, AND the mining is because of the price increase. But I have no way of knowing either unless the buyers or miners say so.

But, while I'm speculating, I'd say the posts on this thread today almost doubled Catcoin's market cap.

Think what we can do if we can collaboratively develop a real, honest, transparent, and profitable business plan to build the Catcoin trade economy.

Hozer - I agree in principle that the developer needs some way to monetize the development effort - but I respectfully suggest requiring people to buy licenses can be counterproductive (as in, exchanges may drop the coin, which would not make anybody happy). One idea is to allocate 20% of mining proceeds at the discretion of the miner, to any one or more of qualified wallets held by nonprofit entities (such as cat charities), plus we can add on that list, development team. Then there can be a regular stream of revenue, and depending on how good a job the developer does, more miners will choose to donate to the development effort (versus other nonprofits, probably we should focus on helping out literal cats, so we'd get a nice list of cat-related nonprofits participating). Your thoughts.

etblvu, if I remember correctly you know .net, can you install https://github.com/Falkvinge/Swarmops on a server and we can see if it works?

I think what you are describing is a reasonable human-oriented approach to the problem. With 3 coders and 3 coin holders that have a good working relationship, you might disagree but at least be able to understand each other and have reasonable discussions, and then agree to work together to ensure that the group always has the capability to deploy checkpoints or lease mining to maintain the blockchain consensus. Another alternative could also be 2 coin holders, 2 coders, and 2 miners who can demonstrate they have individually mined say 10% of the blocks. But how do we include merchants or places that might want to sell goods for catcoin in this governance group? This seems to be a very important thing we are currently missing.

What about the following: always 2 coders (if we can't get at least 2 coders, the coin holder should hire someone), and 3 to 5 more people who can prove one of the following:
* large coin holdings (investor)
* mining capacity (by showing what they have mined)
* transactions (by selling/buying/trading something else with catcoin)

The ideal way I see to do this is open up your catcoin client and vote on changesets that developers have proposed. I'd be willing to implement this in the Catoshi client, but it's going to take awhile, and the code will be AGPLv3, which may not be usable for some exchanges and merchant businesses unless they buy a GPL exception license from me.

In the meantime, we can start with 2 large coin holders, and maybe mining capacity. But until those people come forward and we get a forum acceptable for discussion (is reddit acceptable because we have up/downvotes?) , we are in the wild-west where control of catcoin is determined by who's willing to 51% it.

I suggest a core team consisting of the three most active developers who have contributed code to the source code of the coin or related project; plus three largest holders of Catcoins who can prove their holdings. For any changes to take effect, it would require majority vote, which means 4 out of 6 votes. So it would be hard for any major changes to come to the coin without substantial buy-in from both developers and major coin holders. Once we have a stable coin, we can figure out a long-term governing structure, but this would give us at least the hope of moving forward with minimum of politics and chaos. Community members - please voice your opinions.

Here is something to consider - I think the idea of the blockchain being consensus based is more aspirational rather than factual. With the blockchain technology, we  can aspire to a stable situation where you have a massive number of miners and users, such that it can be said it is consensus based. But the reality is that if there are competing blockchains, it can create a lot of chaos, dissatisfied users, loss of coins, etc., so it is better to work on finding a consensus at the human level rather than at the code level. Otherwise what we get is not consensus but whoever can afford to 51% attack everyone else (which in practice means any whale that wants to get involved can single-handedly take over the coin).

Once we have the coin to the point where it has stable 10 minute blocks and functionally finally where it should have been all along, we'll have everyone running the code "by consensus" and it won't be a problem with chaos or contention. The challenge right now is to get to that point.

Etblvu1

The brilliance of blockchains is that consensus is measured by the blockchain proof-of-work. (in other words, the 'log2_work=' output in the debug.log file)

The way we move forward is by posting proof-of-work, and in a decentralized distributed system, we all make our own choices about what code to run to mine those proof-of-works.

May the most profitable code win.

If we're going to have a catcoin central planning committee, then let's at least make the process to be in the cabal transparent and open to anyone who wants to:

a) prove their existing catcoin holdings
b) buy cheap catcoins (and prove it)
c) buy dedicated miners, and prove you keep them mining catcoins, rather than profit switching
d) post code that does any of the above, or improves the return on investment of the dedicated catcoin miners
e) post any code at all

Also, you know something is very wrong if every working client on the network reports the same version string. That, by definition, is NOT a decentralized system.

Keep this coin alive boys!

I think we need to have a way to measure consensus on how to move forward with Catcoin. I don't think it works to give everyone equal weight (as in straight democracy), because some people are more affected than others by others what happens to this coin, and some people are more needed than others for the continued development and support of this coin. I respectfully suggest that loudness of the voice should not be the measure of anything. I propose that a small working group be formed, consisting only of those with substantial provable holdings in Catcoins, and those who have done actual coding on the Catcoin source code or related project. Input from others would be welcome as well and taken under advisement, but this core group has the most at stake (and has the power to affect the future of the coin), and should have final say. Proceeding in this way would cut down on a lot of drama and politics, and get us back on track. Everyone in the community - please comment.

Etblvu1

If the owner of catcoin address 9Uueyr4T8dmzeZHNhYPjYdZbPpxmE9tUMh is willing to modify their miner code to generate blocks with a minimum timestamp gap of between 1 and 9 minutes, we can unstick the blockchain and have more reasonable block times with no hardfork or orphans.

( see http://catchain.info/block/be27362249fade88cd6e51197b563e1f1dd7e13a82a5b8515c36f5d44830c926 )

Please take a look at my code on bitbucket if you need any assistance. I'dalso be willing to do a commercial GPL-exception license for the Catoshi client if you can help us out. I think the market reaction for having stable block times again will be well worth the effort.