[ANN] Catcoin - 0.9.1.1 - Old thread. Locked. Please use 0.9.2 thread.

[mac-coin]

Looks like either team catcoin or coinium have folk cat coin,

They are on different block and different difficulty.

My wallet is on the same block as team catcoin

My wallet is at block 22705 at 17.52 GMT

[horus1]

I'm kinda new here so maybe im reading it wrong somehow but,

Processed 22813 blocks of transaction history in my wallet,

coinium Current Block    22716

teamcatcoin Current Block    22715

[vondi1122]

Can someone confirm if any of the major pools on a wrong blockchain? Greetings vondi1122

[SlimePuppy]

DEV FLASH

We have figured out the root cause of all of our network and fork problems and are testing a solution now.

TL:DR Beware of Innovation!

- Our code has DDoS protection.
- Our code can blacklist nodes that send bad blocks

Both of these are good and useful capabilities that many coins do not have!  However:

- We released a 1 block retarget method that required a cycle to stabilize - and that meant we had to let the diff settle to near 1 before climbing to stabilize the network.  We know it works because we tortured it on testnet before release.

- During the first low diff period, we got high hash inputs and pools were finding blocks very quickly.  The sudden influx of quick blocks with slightly different difficulty numbers triggered the DDoS protection - and network nodes (primarily the pools) were temporarily blacklisted.  This resulted in micro-forks until the diff adjustment slowed things down, everyone resynchronized, and the network stabilized.

Unfortunately, this disruption also reset the diff adjustment algo - and it dropped back to near 1 for a reset - and we started all over.  We think this is why the real-world performance of the new difficulty adjust code is so different from the testnet performance - we didn't trigger DDoS or forks on testnet.

TL:DR part 2:

We've got root cause, we've got code written, we're deploying test nodes to positively confirm that this works.  We'll release a 'maintenance release' wallet that will fix our network.  This is NOT a FORK.

Once we stabilize the main network, the exchanges will synchronize and everyone's transactions should complete and confirm.

Jonhhy_non called it:  We have just de-wormed the CAT.  We will absolutely be back in sync very soon and we will be ready for the influx of new CAT lovers when the Oprah Network highlights our community on episode 6 of the Diamond Collar.

Life's good and getting better folks!

Andy

DEV FLASH

[horus1]

......and we will be ready for the influx of new CAT lovers when the Oprah Network highlights our community on episode 6 of the Diamond Collar.

Will this be in the near future?


Good to hear that u'r working on a solution! Almost had given up.. Do you think it will take hours or days till the new wallet?

[ozzymax]

Awesome news SlimePuppy. thanks for the update and all yours and the rest of the Dev teams hard work.

[SlimePuppy]

......and we will be ready for the influx of new CAT lovers when the Oprah Network highlights our community on episode 6 of the Diamond Collar.

Will this be in the near future?


Good to hear that u'r working on a solution! Almost had given up.. Do you think it will take hours or days till the new wallet?

We have testnodes running new code now.  Once we confirm this fixes the problem we'll get production code and wallets on the streets.  Should be a matter of a few days, but that's a guess right now.

We have about a month to be upgraded and stable for episode six of the Diamond Collar show.  That's our 'make or break' point.  The network will be stable well before that, however.

Thanks for not giving up!

PS:  "Official" song of the Dev team: The Year of the Cat.   

[hozer]

DEV FLASH

We have figured out the root cause of all of our network and fork problems and are testing a solution now.

TL:DR Beware of Innovation!

- Our code has DDoS protection.
- Our code can blacklist nodes that send bad blocks

And pools that get more than 2 blocks in a row look like an attacker to the rest of the network.

The only pool that does not have this property is p2pool, because the block generation is distributed among the p2pool participating nodes.

The lesson: if you are a pool operator, you will eventually get forked and banned by the rest of the network if you and your users get greedy.

If you want to properly rate-limit your pool, or use p2pool under the hood so this is not an issue, I am happy to offer professional development services to help you do this for free for any catcoin pools. Other coins have to pay. In catcoin

Why? Because izzy

[TheTribesman]

......and we will be ready for the influx of new CAT lovers when the Oprah Network highlights our community on episode 6 of the Diamond Collar.

Will this be in the near future?


Good to hear that u'r working on a solution! Almost had given up.. Do you think it will take hours or days till the new wallet?

We have testnodes running new code now.  Once we confirm this fixes the problem we'll get production code and wallets on the streets.  Should be a matter of a few days, but that's a guess right now.

We have about a month to be upgraded and stable for episode six of the Diamond Collar show.  That's our 'make or break' point.  The network will be stable well before that, however.

Thanks for not giving up!

PS:  "Official" song of the Dev team: The Year of the Cat.   

Excellent news! CAT seems to have been very sick lately.
 

[ParadoximA]

Thanks for the hard work and feedback SlimePuppy!

Should we be mining before the fix comes out? Which pools are on the right fork? TeamCatCoin hasn't found a block for 34 hours now - that doesn't seem right. p2pool found a block half an hour ago - if they're on the right fork, I'll be moving to them until we know what's happening at TeamCatCoin.

[raveneye]

DEV FLASH
- During the first low diff period, we got high hash inputs and pools were finding blocks very quickly.  The sudden influx of quick blocks with slightly different difficulty numbers triggered the DDoS protection - and network nodes (primarily the pools) were temporarily blacklisted.  This resulted in micro-forks until the diff adjustment slowed things down, everyone resynchronized, and the network stabilized.

It is not "slightly different difficulty numbers" that are causing the DoS logic to kick in, it is a bug in ComputeMinWork and out-of-date checkpoints.

Here is what I am seeing, clients are getting disconnected in ProcessBlock by the DoS logic with this error message:
ERROR: ProcessBlock() : block with too little proof-of-work

The Blocks are being rejected because:
1) The latest checkpoint is on block 20000 from Dec 30
2) deltaTime passed into ComputeMinWork is about 3921600 (about 6 weeks)
3) ComputeMinWork is using nTargetTimespanOld*4 (2 weeks * 4) instead of nTargetTimespan*4 (6 hours * 4) because nHeight of the last checkpoint is less than 20290.
4) The diff at the last checkpoint was 64, causing ComputeMinWork to return a Diff of 16 instead of a proper value below 0.015625.
5) Pools are attempting to submit blocks with a diff that is valid (above the 2-3 diff currently floating around), but is below 16 that ComputeMinWork is returning.
6) The pool is DoS'ed for publishing a "invalid" block and cut off of the network.
7) The pool continues to build what it thinks are valid blocks in the 2-16 diff range the the rest of the network rejects, and it lands on a fork.

(The reason you did not see it on testnet is because ComputeMinWork returns bnProofOfWorkLimit.GetCompact() for testnet, and it never gets to the spot where the bug is.)


The solution is simple:
Create a new checkpoint with nHeight > 20290.

[mac-coin]

So once again

Teamcat at block 22785, but they havn't found a block for 36hours,  (last block found 22647) block still moving forward thou and still in sync with my wallet

Coinium at block 22747, and they are still finding block


who's right, who's wrong??

[homey108]

I won't be doing any cat mining until I receive the coins that I've already mined. Fool me once, shame on you, fool me twice and shame on me.


 I will say that it's a bit unethical to not shutdown mining at a pool while there is such a big issue. Just letting people come and waste resources and electricity/$ when they won't be paid is bad karma and just plain wrong. I have much more respect for a pool that would just shut everything down and warn people about the problem instead of ignoring emails/trouble tickets. This is the new frontier, there will be bumps on the road for sure. It is understandable and to be expected. It would be nice if pool admins had the common decency to step up to the plate and deal with these bumps responsibly and ethically.

On a positive note, us cat groupies are quite lucky to have good folks actively researching and brainstorming solutions to these challenges. The day there are no dedicated developers involved is the day the dead cat doesn't bounce back. Kudos to Slimepuppy and the team!  

[vbal]

Doge is running fast, I hope cat is not dead for now.

[bennet]

Who knows what will happend with all round shares mined till period of pool problems? Reset...?

[vondi1122]

I'd also take the chance and say thank you developers for doing your awesome work. kudos. greetings vondi1122

[anotherlateminer]

[madoka]

conium is taking all blocks

[vondi1122]

conium is taking all blocks

is it on the right chain?

[madoka]

I'm not sure, but are the only blocks that are