Topic: [ANN] Catcoin - 0.9.1.1 - Old thread. Locked. Please use 0.9.2 thread. - page 5. (Read 131044 times)

Thanks guys.  Blak found the problem - stay tuned for a new wallet.

In the mean time, please note that you can find block info on the network chart posted here:  http://catcoinwallets.com/ and via the only blockchain explorer that is maintained by the community here: http://catcoinwallets.com:2750/chain/Catcoin  Cryptsy maintains a separate block explorer and they keep it current as well.  Nobody maintains the old page at catstat.info *** and it is at least one code revision behind.  Do not rely on that explorer.

Andy

*** edit...   Apparently that domain has been orphaned and has been gathered by our resident terrorist, hozer.  He's running his own hacked clients and is consistently and intentionally creating his own mini-forks with the three nodes he controls.  The real CAT network is strong and there's no way his code will harm our network.  But that's not stopping him from using that to 'claim' there's a problem.  Don't fall for his lies.

Catcoin dev:

On current github source, main.cpp line 1118:
unsigned int i;

This needs to be changed to:
int i;

Prevents both Linux and Windows QT wallets from dying while importing block 2015.

james

Glad to see this isnt just me - i get this exact message with the new wallet.
Assert suggests to me that (in src/main.cpp: 1222) that the chain has separated somehow; as Assert is the same error you get when incorrectly cloning a wallet (ie. differs away from chain derived from Genesis block).

wallet problem 

Seems to be separate chains advertised around block 43869, 43977 and 44176.
Catcoin block explorer keeps flicking between 43977 and 44176.

Hi all,

Seems to be more than one version of the Catcoin chain being distributed at the moment.
The block explorer at http://catstat.info:2750/chain/Catcoin indicated that we were up near 44,xxx - has just dropped back as of 10 minutes ago to block 43977.

I'm also unable to get any Windows QT wallet to sync correctly; it seems to crash at a number of checkpoints (i've tried v0.1, 9.1.1 and 9.2) which may indicate the libdb used was incorrect or there are indeed multiple forks of the chain (particularly the Assert error shown).
Linux catcoind appears to get to 43,xxx and rejects new blocks (timestamp too early for xx).

Can someone try syncing either wallet from scratch and letting me know the results?

james

Working pool.

https://pool.bejjan.net/pool/CAT

Proportional, 1% fee, free transactions.

Welcome!

/Bejjan

Ok everyone - Catcoin has new code on the streets.

Windows wallet:
http://catcoinwallets.com/win/Catcoin_0.9.2.0.zip
SHA256: A3FAC5C54F2E22B2918823140A4D121BA25160DFB3E2AEDB073A14E30A5204B9

Windows build instructions are here:  http://catcoinwallets.com/win/catcoin_wallet_build_windows.pdf

Source:
https://github.com/CatcoinOfficial/CatcoinRelease

Mac wallet is on the way.

Cryptsy has been notified.  The seednodes and catchain explorer are updated.  The Geekhash pool is updated.  We'll get the word out to other pool operators next.

The fork block is 46331 - we should hit it in mid-July.


And yes...work on the next release has already begun.

Andy

/edit...updated Windows wallet hash - bug fix posted.

The final release candidate code is on testnet.  This is a full test of the code that'll be posted on catcoinrelease.

http://testnet.geekhash.org/

Comments?  Code?  Exploits?  Suggestions? Alcohol recommendations?  Bring them soon - after the last couple of years I need a drink. 

skillface - sorry to see you go.  Best of luck.

Vondi1122 - thanks for posting the PM.  I didn't want to without your permission.  As Blak already outlined (and as I outlined in my response to you), none of the current dev team was involved in the project at launch.  I truly do not know the intent of the original crew.  I think we can all agree that it wasn't a well thought-out release and that the first few months were rough.  During that time - the time I started to ask questions here - my 'gut' was screaming to me that while it appeared that a number of people were working for CAT and her community, that 'something' wasn't quite right and that it appeared someone was also working against CAT.  I think it's pretty clear from recent activity that there's at least one person that has proven to be working to damage the coin and her community.  I don't know if they're acting alone or not.

We're a very small team, and frankly it's taken way too much time away from the project dealing with folks trying to hoze the coin.  One thing I do know about the folks working on the coin is that we completely adopt and live by the original intent of crypto - that the process be completely open, that we all run the same code, and that modified code must ONLY be run on testnet.  We understand this to be critical:  "It's crucial that clients follow certain rules in order to maintain consistency across the network, and to protect the Bitcoin security guarantees."  

I think that it's pretty clear from the network hash rate and the rise in CAT's value that the community and the coin is alive and well.  I also think that the new wallet will result in a dramatic improvement in block times while significantly reducing or eliminating the problems with PID.

Please give the testnet code a look if you're so inclined, and if you can see ways to improve it please DO let us know.  If you don't want to post the code here, PM me or Blak, or hit us on #catcoin-dev.

For everyone in the community - please only pull code from https://github.com/CatcoinOfficial/CatcoinRelease or use binaries from catcoins.org - this is the only code that should be run on the network.

Thanks everyone.
Andy

My take is early-on a bunch of people insta-mined Catcoin, and didn't contribute much of anything, except possibly some just-as-shady promotion.

So if we want to be taken seriously, we need to address two things:

1) hashrate-proportional distribution of block rewards
2) address the early-instamining concerns.

Figuring out 1) is both technically challenging, and controversial. (if you happen to remember discussions about block time or rewarding loyal miners)

I think we might be able to address 2) by implementing Demurrage (which, I think, for Catcoin, should be called Depurrage), and since Freicoin has code that does this, I'd say it's a relatively minor technical change to borrow their code. This leaves the controversy. Would anyone run a hard-fork upgrade in which coins lose value over time? (see http://p2pfoundation.net/Freicoin#Why_Demurrage )

And quite sadly, NONE of the original developers are involved in this coin any more.  I didn't get involved in this coin until around 3 months after launch, and by then, the damage was done.  If there is still interest from the community as a whole, then we can still try and pull this CAT back up.  I would like to hear from more of the community, not just one person.

Selfquote, this is what i wrote Andy aka Slimepuppy 4 months ago:


Cheers

I say do whatever - I'm done.

This coin's been dead for too long.

What I thought I heard you say is there is a small group of people that want to hardfork Catcoin by adding a difficulty adjustment that is selected by the nTime provided by the miner. Is that accurate?

If that is accurate, I think it's best for any of you advocating such a thing to be very clear how and why you think this is better, and that you can answer my concerns about why I think it's exploitable.

The reason for asking cryptsy to halt trading is to remove any potential accusations that someone is going to profit by releasing code they know is broken. It is to YOUR benefit to have them halt trading if you are going to ask them to run a new hardfork upgrade.

I am not making any bullshit claims about being 'official'. I have published code that modifies the minimum block time, and let the community make a free choice if they want to run it. Apparently we have differences of opinion on if this is a good idea, and whether doing so constitutes 'manipulation' (which is an awefully vague term). What is it going to take for you to consider that maybe the community actually wants a longer minimum block time, and they would rather express their opinions by what code they run and what chains they mine on, instead of having to put up with inflammatory FUD all the time?

If you want to have an 'official' release where it is a crime to run unauthorized code, then change the software license, and then you can send a DMCA notice to anyone that is not following exactly as you wish.

I have the same problem you do, these idiotic arguments take time away from more important things, like posting cat pictures.

I keep talking, because I keep hoping there are people listening who are sick of accusations and attacks, and would rather have an upgraded catcoin that includes features from the latest bitcoin-core, and is secure, reliable, and has consistent block times, even with a fraction of the total scrypt hashrate. I can see this happening with a 3-9 minute minimum block time, and maybe Heavycoin's temporal retargeting. If you want this, mention #catcoin somewhere other than this toxic forum.

If you are going to try and be THE power to tell Cryptsy to halt all trading, because YOU want to fork the coin, then you are not only proving that you are trying to manipulate the market for your own pocket, but also proving that you are one hell of a crook that should be put behind bars.  Market Manipulation of ANY kind, be it Crypto, Fiat, Precious Metals, etc.. is a federal offense.

I'm done arguing with the village idiots.  So far, only ONE person has admitted to manipulating both their miner AND running unofficial code.  I'm not going to waste any more of my time or bits of data replying to someone that all they want to do is take precious time away.

Blak

The code everyone is running is the minimum block time. What's up for debate (for me) is should that be 3 minutes, or even longer. Leaving it at 30 seconds has proven to be rather non-functional for a 10 minute block time. 3 Minutes seems to work okay with the existing PID algorithm.

It's not even semi-random, the timestamp (ntime) is pretty much entirely under control of the miner. Go read https://mining.bitcoin.cz/user-manual/stratum-protocol/ and give me an argument about why you think it's a good idea to use something chosen by the miner to determine what difficulty algorithm to use for the next block. This basically lets a big pool select whatever algorithm they want to insta-mine for 10 or 20 blocks and then leave it hung on whichever algorithm is going to make everyone else expend the most effort on finding the next block before they come back and do it all over again.

Agreed. How do you want to do that?

My suggestion is that we have Cryptsy suspend trading for 2 weeks.

At the end of those two weeks, whichever chain has the highest log2_work wins, and continues to be called Catcoin.

The other chain can either pick a new name and trading symbol, or stop mining and become an orphan fork.

Sound fair? Or do you have another idea?

That's the whole point of the semi-random algo used.  According to your statement, that so-called attacker (most likely you that CLAIMS that everyone is running your code, which looking at the blockchain, only YOU are running that code), could set the timestamp of each block....

ok,  say someone (you) does come up with a clever way to TRY and submit blocks within the certain period of the certain algo picked, that is still a VERY small period of time to submit a shit-ton of illegal blocks, as your code would try to submit, as you are still trying to manipulate the coin with low-ish fake hash.  What would happen if a LEGIT block domes through at the same time as your hacking attempts, it would throw you off 100000%, as the whole algo was changed, and then, it would probably take a main system another few hours to try to decipher what the next attack would be, and then in the mean time LEGIT blocks would be happening, and you, the hacker would only get 1 or two blocks out of the whole deal.

Hell, I'm even willing to tighten it up and add more algo's in there, with smaller timestamp resolutions.

Let's let the WHOLE community decide.. not ONE asshole who wants to try and hack the whole system EVERY FUCKING TIME!

Blak

My biggest concern with that code is that by going back to the original GetMedianTimePast function, in combination with using the timestamp to select the difficulty algorithm, you are basically giving the miners the ability to manipulate the difficulty by selecting timestamps that cause the difficulty to do whatever they want. This is why early disclosure and discussion of the code is so critical, so that we have a chance to talk about how or why it might be exploitable.

I see a repeat of some exploitable things in Dogecoin going on here.. https://github.com/dogecoin/dogecoin/issues/323#issuecomment-37419607

GetMedianTimePast works in Bitcoin because they only change difficulty every 2016 blocks, and an attacker would have to maintain a large hashrate for at least that many blocks to subvert the block timestamps. With Catcoin and adjusting every block, you only need 5-10 blocks to cause dramatic difficulty changes, which means there's now a profitable motive to run the difficulty way up, and then mine on another fork with different timestamps and lower difficulty, but more consistent block times.

As far as I can tell, there are only two ways out of that mess:
1) do what Dogecoin did and merge-mine with another coin. (What if all the cat-themed scypt coins that are still alive agreed to do AuxPOW merge-mining?)
2) longer minimum block times

There also might be a way to use Heavycoin's temporal retargeting (https://heavycoin.github.io/about.html#temporal-retargeting) but I haven't quite figured out if that'd be exploitable in the same way or not.