Yeah but you all know how this goes, the noises uttered / keystrokes typed are simply neurolinguistic programming, dictionaries etc about what utterances mean are mere myths / superstitions of the listeners / readers, the real meaning is the effect it has, which of course is that the shitcoin / scamcoin goes ahead as planned. Noises (or even images or memes) about Satoshi, Hal Finney, Jesus, God, some stupid dog, or whatever are mere means to an end. Nothing you say is likely to make the scammer / bullshitter / neurolinguistic programmer not go ahead with the coin, regardless of what it / she / he has to type, utter, memefy or whatever in order to achieve the launch.

Heaven forbid something actually appropriate to Hal's stature be created in his memory, that might take actual work!

-MarkM-

Ok then lets move GeistGeld discussion to GeistGeld's thread then...

-MarkM-

You always need to pay the fees of the chain on which you are implementing your coloured coin or user-issued token or user-issued currency or whatever.

So in order to pay such fees directly with your community currency, your community currency would itself need to be the medium.

So you could make a community coin then implement coloured coins on top of it so that other people who also want to run coloured coins and do not want to do so on the Bitcoin blockchain could run on your community coin instead, paying fees to your community coin's miners or stakeholders or whatever.

Two problems though: one, how are you going to secure your community coin securely enough for folks to trust it enough to run their own coins as coloured coins on its blockchain; and two, coloured coin code is very low level against the blockchain, so every little detail about the blockchain you use, such as at which blocks it turns on which BIPs and block-versions and how malleable its transactions are and so on could impact your code.

If it was easy to port coloured coin code from blockchain to blockchain it would have been done thousands of times over by now; any new shitcoin/scamcoin released without its own version of coloured coins would be considered garbage, people would complain just like if there was no windows version client or no pools or whatever, it would just be one of the standard things considered normal for coins.

There are of course things like Ripple and NXT and Bitshares and so on designed to allow users to issue tokens over them.

Maybe one of their tokens-you-pay-the-fees-with is cheap enough (aka the fees are cheap enough) that you might find it usable?

-MarkM-

Thanks, that magic over-ride for pywallet did the trick it seems.

-MarkM-

I have been finding my ancient wallets left over from trying out various altcoins over the years, building the corresponding altcoins and getting up to date with their blockchains to access any coins those wallets might contain.

This went fine until I happened upon a wallet that crashed the coin daemon, asking for a database recovery to be done.

db_recover utility for BDB wants you to have a log not just the database (wallet) itself, so that was no good.

So I grabbed pywallet, a newer version of it said to be better than the older one someone else had on github.

Doing just a dump of the wallet using pywallet resulted in huge private keys with long strings of FFFFFFFF in them.

So I tried pywallet's recover routine, as it can work on a file not just on disk drive. (It is intended for scouring a disk or part of a disk trying to recover keys from deleted wallets).

The problem I have now is that the recover routine will not let me give no password to the wallet it creates to put the recovered keys into.

It made me a file with 207 recovered keys in it, but I had to tell it a password to use to encrypt that wallet, so it is an encrypted wallet.

I did mention I am talking about ancient coins, right? Right. Well guess what, some ancient coins do not have encrypted wallet capability even in their latest incarnations.

So the coin daemon cannot use this nice shiny new, but encrypted, wallet.

I asked pywallet to dump the wallet, I even told it the password, but instead of dumping un-encrypted keys the dump has all the private keys encrypted. It told me at the top of the dump that the password was correct, if I tell it a wrong password it tells me it is incorrect. So it does know how to decrypt the keys, heck it probably decrypts them (or at least one of them) to check whether or not the password I give is correct. But the damn dump does not dump decrypted keys.

So I am still trying to find out how the heck to import the damn keys or, probably simplest really, decrypt the damn wallet and leave it decrypted.

I had to tell pywallet the exact address-version the coin in question uses to get it to do things right, so it does not seem likely I could just load it with some other coin deamon entirely, one that does support wallet encryption, then tell that to decrypt the thing.

(A brief look toward that idea actually leads me to think maybe the coins that do support wallet encryption do not offer a decryption option to turn them back into decrypted wallets anyway, so even if I convert the whole thing over to some other coin's address version byte style I maybe still could not use some more modern coin-daemon to decrypt the damn thing.)

Any insights/ideas/solutions?

(Hopefuly something faster than reverse-engineering pywallet to figure out how to hack it to allow not encrypting the recovered-keys wallet it creates. )

-MarkM-

I favour merged mining. So there is no moving hashpower from coin to coin, there is just adding more coins into one's merge.

When DOGE showed that Litecoin was so pathetically insecure that just some stupid meme could whip up almost overnight more hashing power than Litecoin had, I first urged them to implement merged mining then, when they refused, I dumped all my litecoins. And did not buy into DOGE, because who knows, the next stupid meme might be a "lets destroy all the scrypt based coins with the superior hashpower we memesters can whip up" meme blowing away DOGE as easily as DOGE could, had it been a destroy litecoin meme instead of a compete with litecoin with yet another crappy clone meme, have destroyed Litecoin.

Now that one can merged mine them I am re-thinking, heck maybe scrypt ASICs might be worth looking into if enough coins can be merged that between them all they can maybe attract over half the planet's scrypt hashing power and not lose it overnight to some stupid meme.

-MarkM-

GeistGeld only has the so called "dirty fix" of the RAM problem, not the full/clean fix.

That is still way better than no fix at all, of course.

The .git/config in the repo I have been using loks like that:


Which I think means I am using a specific branch or tag, one to do with the dirty fix, rather than using the bare main top of the repo you'd get if you just used the clone URL github shows you as the URL to clone the repo.

The main problem I have with p2pool that isc specific to GeistGeld is that until I get GeistGeld's difficulty high enough p2pool makes so many requests to it so fast that the operating system runs out of connection handles, then p2pool cannot open connections to anything.

So I would have to solo mine it at higher and higher hash rates until I got the difficulty high enough that p2pool would not choke on it.

This would be solved of course if umpteen people all merged it. It happened to me when no-one else or no-one else with much hash rate was mining it and my pool or main miners went down or off the net or whatever.

Geistgeld's difficulty would plummet fast as it is very adaptive difficulty-adjustment. So very son it wo0juld be at tiny difficulty, nice to CPU-mine but choking to a terahash of mining power on the p2pool.
(My problem with p2pool that is not specific to geistgeld is my Neptune would not get more than one terahash pointed at p2pool though it could get 2 or even 3.4 on mmpool).

-MarkM-

I just looked at coinmarketcap site and it lists the value of bitshares PTS as somewhere near 150 satoshis, which is way different from the 29 thousand satoshis that Cryptsy told me. If coinmarketcap, rather than Cryptsy, is right I maybe don't really have much of a stake in PTS afterall.

-MarkM-

Hiring!?!?! What a dirty word! We want devs to work for free!



-MarkM-

Once upon a time folks started thinking there was something wrong, probably a memory leak, with I0Coin, as it was using ridiculous amounts of RAM.

So much so that it got dropped from the merge of mmpool.org, the pool that merges the largest number of the merge-capable SHA256 coins.

Luckily someone followed up on it long enoguh and hard enough to figure out that the problem was not I0Coin's really at all, I0Coin was simply the fastest block speed merged coin merged by major pools, so it showed before slower coins a problem they were all eventually going to exhibit.

Someone (likely the discoverer of the real problem though I don't recall for sure) then came up with a "dirty fix", which I think was applied to all the coins (well except maybe namecoin, I don't know how namecoin relates to this whole excessive-RAM-usage problem). Or maybe some coins skipped the dirty fix and went directly to the clean(er?) fix, as a better, cleaner, more correct fix involves storing in a database the info that the dirty fix simply threw away.

However it looks like GeistGeld only has the so called dirty fix so far, so GeistGeld could benefit from the clean(er?) fix.

But now that a whole new generation of bitcoin classic satoshi code is coming out or already upon us, it would make sense to upgrade all the coins to be based off of that.

Basically, bitcoin only supports being the parent chain of a merge. So to base merged mined coins off of bitcoin code involves adding into it the "be a child chain in a merge" capability.

Once that basic template, bitcoin plus being a child chain in a merge, is done the rest is basically cosmetic, the usual "make a clone-coin" stuff of changing the names, the block times, the magic bytes including what prefix addresses get, the ports, the images, the visible references to the name of the coin, the block-rewards schedule, and all the nitpicky things like exactly which BIPs are supported starting at which block, what if any protections against transaction malleability are to be in there, starting at which blocks, how difficulty adapts to changes in hashing power, what the checkpoints are and so on.

The merged mined coins supported currently by mmpool are bitcoin (as parent chain), namecoin, devcoin, groupcoin, ixcoin, i0coin and unobtanium. Other old classic merged mined coins are CoiLedCoin and GeistGeld, of which the most likely to get into the merge any time soon is CoiLedCoin.

There is also an interesting coin that has a game in its blockchain, huntercoin, which claims to be able to be merged mined using SHA256 but so far no one seems to have that working, so once familiar with the whole scheme of things maybe you could be the pioneer who actually gets that sucker working so it too can get onto mmpool.

-MarkM-

A password? Oops. Gosh knows what it might have taken as my password in all of the stuff I typed that it kept rejecting for one reason or another.

It might turn out that once I exit pts_client I might not be able to get back into its wallet.

No big deal if so, I can just go through the import stuff again.

I am off to try to build bitshares itself now, which apparently should work out of the git repo as it looked like the bug I hit in bitshares PTS repo was already fixed in bitshares repo.

(And if not well I know how to work around it now.)

If you only end up with maybe four or so times as many bitshares as bitshares-PTS it looks like that will be maybe only about half as much value as your bitshares-PTS, since it looks like currently bitshares itself is priced at about a tenth the price per unit of bitshares-PTS.

I do remember giving a password for the wallet I created, just not for the account I created within that wallet.

-MarkM-

Oh yeah, nice thought.

Maybe not a wash though as buying or dumping either way they have to hear about the industry so overall maybe publicity is generally a good thing.

-MarkM-

Yeah there are lots of developing coins that could use some development work!

Looks like it is time again to upgrade all the SHA256 merged mined coins, for example. That could be nice routine work you could do to prove you can work well with the common codebase so many coins share...

-MarkM-

I managed to use the wallet_import_bitcoin command to import my wallet.dat from the old protoshares into the shiny new pts_client.

It was complicated by the need to first create a wallet in the pts_client and create an account to import it into, which apparently requires inventing a lower-case account-name.

At first blush it does look like a nice number of units of the stuff so I guess it is time to go try to build a shiny new bts_client to import into...

-MarkM-

EDIT: And oh yeah Stan, it looks a nice enough number that maybe I will go sign up at your forum, being maybe now stakeholder enough it should be worth my time to do so.

In other words you don't care about end users, screw them all if by doing so you can profit by making the coins they use unstable by sucking away the miners.

"Lets de-stabilise this hash-type so we can profit by pretending the people will be better off coming over to us."

Sound like, like DOGE did to Litecoin, your main contribution really is just to show that the hash type involved is acctually not secure because some new garbage can suck away the miners.

DOGE showed that litecoin was - and likely still is - so pathetically insecure that some stupid meme could whip up more hashing power than it had almost overnight.

II haven't seen a meme from you yet so I guess you are demonstrating that whatever hash type you guys use is so even more pathetically insecure than Litecoin that more hashing power than it had can be whipped up almost overnight without even using a meme!

For a while Litecoin and DOGE were like you guys, fuck the end uers, lets split up the miner base by refusing to merged mine.

Eventually though DOGE did implement merged mining so at last the litecoin miner-base can secure all the scrypt coins other than those that, like you apparently, do not actually give a damn about end users and still want to try to weaken the scrypt family in general by trying to fragment the scrypt miner-base...

With luck maybe some meme will hit soon, saying lets destroy all the non merged scrypt coins, instead of, like DOGE did, saying lets make yet another crapcoin to try to fragment the miner-base making all coins that use this algorithm that much less secure...

Litecoin was lucky the meme that hit them was just a make another crapcoin meme not a lets attack litecoin's chain and doublespend it meme...

-MarkM-

Sounds like maybe both coins refused to implement merged mining, thus are both deliberately avoiding participating in making the entire hash type they use more secure.

So seems likely that both ought to be headed for the trash heap as investors seek out coins that are actually sincere about wanting to provide security thus avoid fragmenting the hash type's miner-base between coins, instead having all the coins merged so as to all benefit from each other's miner-base...

-MarkM-

I did create it, as hired hand working for Unthinkingbit to help him create DeVCoin.

We made GRouPcoin first as a testbed to test various coding approaches toward what he wanted DeVCoin to do, then once we figured out how to code DeVCoin GRouPcoin was left as a simple 50 coins per block forever coin.

He would have been fine with just abandoning it but the gamers who had mined it to help us run the testing didn't want their characters in games to wake up without their GRouPcoins in their backpacks kind of thing so we just kept it running so the folks using it could keep on using it.

Gamers can be great for playtesting stuff but taking away their pile of sand after they have built their sandcastles with it just isn't kosher.

So technically it is just slightly older than DeVCoin, but is part of DeVCoin development so maybe should be thought of more as part of the DeVCoin project than as a rival coin that preceded DeVCoin.

(As I mentioned, gamers used to copper silver gold platinum etc at fixed exchange rates kind of hoped it would simply function as a 1000-DeVCoins denomination coin.)

Cinnamon, I am sending you two more sends of 1000 GRP each, do you think you could put your github repo at a less confusing URL, one that actually says GRouPcoin instead of I0Coin and without another URL embedded in the URL ?

-MarkM-

Coinmarketcap dot com claims it is at $3.99, what is this over-five you seem to be referring to?

-MarkM-

Yup. First I have to convince PTS to dump me those privkeys. Its CLI mode might not be best for that unless it has built in less/more pager function, I think I might kill it and try oneshot commands with output piped to less or more to capture the help output.

Or maybe I will try to see if it can talk to its daemon-self like typicalcoind can (or could before the talker was broken out into a -cli program). I dunno yet how similar to bitcoin-based daemons it is.

Too tired to do it now, I guess its sleep first then start exploring my shiny new pts_client when I wake up...

-MarkM-

I sent you two sends of 1000 GRP each. Partly following above precedent maybe but partly simply because when initially shipping the coins out of the wallets they are mined into 1000 is a small enough collection of 50 coin mining rewards that the client can build the transaction.

Players used to roleplaying games where there are so many copper coins to a silver coin, so many silvers to a gold, so many golds to a platinum and so on used to figure that logically GRouPcoin should be a thousand-devcoin-piece since it mints 1/1000 as many coins as DeVCoin does.

Miners might tend to argue though that such a convention ought to be adjusted for difficulty, with GRouPcoin less difficult than DeVCoin thus presumably less secure, it should be worth less than a thousand DeVCoins per coin.

Looking at http://galaxies.mygamesonline.org/ingroupcoins.html (Asset values expressed in GRouPcoins) it looks like folks have been losing faith in miners' ability to restrain themselves from throwing away the GRouPcoins for free or close to free; current latest values file shows them worth less than a DeVCoin each!

I wish that the gnuplot plots were working, it'd be nice if the DeVCoin folks would offer a DeVCoin bounty for figuring out what broke with those, basically the plot pages worked nicely when I was running gnuplot on Fedora but when I changed over my home systems to Ubuntu suddenly all the HTML pages of plots produced and uploaded to the webserver just look like blank pages now with the Firefox that comes with Ubuntu.

( Links to tables and plots are shown on the menu page http://galaxies.mygamesonline.org/digitalisassets.html )

Still, one can scroll down the http://galaxies.mygamesonline.org/ingroupcoins.html page to see that GRouPcoin was not valued so low in the past...

Last time bitpop offered to sell me some at the rates shown in the latest rates include-file (from historical archives of which the tables and plots are generated) I said I kind of had enough for now and he said well maybe he will just give them to me someday then. I anecdotally mentioned that conversation to some players (of Galactic Milieu of course) and pretty soon they all seem to have decided to just sit and wait for miners to give them away.

So thanks bitpop for at least giving them as bounties to someone deserving instead of just giving them to a rich old fart such as me.

-MarkM-