Pages:
Author

Topic: [ANN][MOTO] Motocoin - page 68. (Read 178259 times)

sr. member
Activity: 434
Merit: 250
June 07, 2014, 09:06:20 AM
Yes, but IMO as long as the chain is still hashing and no-one has demonstrated the proposed difficulty time-warp the coin is still safe and surviving, for now.

Though this could change entirely at any moment, I suppose.
It is not safe. Bot owners (including you) can fork the whole blockchain, it is not what people call safety.

Only if they can overcome the production rate of the other bot miners, though.  This is, after all, still a form of 51% attack.

Let's look at some facts:
 * To date, our longest fork was 13 blocks, this is *significantly* lower than historical forks of most other "fast" coins, including in particular HUC - the closest coin to moto in many ways.
 * There have been only 10 forks longer than 6 blocks, which is the "traditional" confirmation count for BTC at 10 minute blocks.  (Considering we average only 47 seconds per block since block 8000, we should "expect to need" almost 20 times as many confirmations (120) for the same level of integrity.)
 * The vast majority of forks are simple stales, 1 block forks.  These are entirely to be expected at such a low block interval.

Given these facts I'd argue that (historically, at least) the coin has shown itself to be pretty damned safe and secure!  Personally, I don't know of any other coin at all with such a high transaction confirmation speed and such a low fork rate.

newbie
Activity: 2
Merit: 0
June 07, 2014, 08:35:29 AM
motogame did not start.It appear a dialog that inform me that app crashed.
sr. member
Activity: 434
Merit: 250
June 07, 2014, 08:19:43 AM
Hello!
I've got a problem.I can't launch the game.
In the log there are some entries:
2014-06-07 10:03:06 trying connection 5.158.233.102:13107 lastseen=4,5hrs
2014-06-07 10:03:06 CreateNewBlock(): total size 1000
2014-06-07 10:03:07 CreateNewBlock(): total size 1000
2014-06-07 10:03:07 CreateNewBlock(): total size 1000
2014-06-07 10:03:08 CreateNewBlock(): total size 1000
2014-06-07 10:03:08 CreateNewBlock(): total size 1000
2014-06-07 10:03:09 CreateNewBlock(): total size 1000
2014-06-07 10:03:09 CreateNewBlock(): total size 1000
2014-06-07 10:03:10 CreateNewBlock(): total size 1000
2014-06-07 10:03:10 CreateNewBlock(): total size 1000
2014-06-07 10:03:11 CreateNewBlock(): total size 1000
2014-06-07 10:03:11 connection timeout
WTF?what am I doing wrong?

Are you able to run the "motogame" binary directly?  If you run it from a console does it generate any errors?

full member
Activity: 204
Merit: 100
June 07, 2014, 05:21:22 AM
Yes, but IMO as long as the chain is still hashing and no-one has demonstrated the proposed difficulty time-warp the coin is still safe and surviving, for now.

Though this could change entirely at any moment, I suppose.
It is not safe. Bot owners (including you) can fork the whole blockchain, it is not what people call safety.
full member
Activity: 204
Merit: 100
June 07, 2014, 05:16:17 AM
We are already very far from original satoshi idea... so I propose to use a floating threshold. If no one is able to find a block in 2 mins then threshold will be increased. Or better say that new block threshold should be somehow linked not only to the previous 2000 blocks and separation between them but also to separation between this block and previous. It is hardly to find any hole in there... looks fine to me at first sight. It may introduce some kind of oscillation but I do not see any big flaw that can be used in attack. If you are issuing new block in just one second then your solution shall be shorter than say 5 second. If you are issuing next block after a five minute of game than your solution can be very slow and not optimal.
How can you reliable measure time interval between adjacent blocks? I don't think there is a way.

Would it be possible to use the nonce patterns in identification of human played maps/blocks giving them priority?
Would it be possible to increase the complexity of the maps with higher nonce values?

If there is a way to force bots to emulate the way humans play, would that mean safety and survival of the coin?
I'm very tired of answering the same questions. Nonce is useless, you can just change your coinbase address and always use nonce=0.
newbie
Activity: 2
Merit: 0
June 07, 2014, 05:06:24 AM
Hello!
I've got a problem.I can't launch the game.
In the log there are some entries:
2014-06-07 10:03:06 trying connection 5.158.233.102:13107 lastseen=4,5hrs
2014-06-07 10:03:06 CreateNewBlock(): total size 1000
2014-06-07 10:03:07 CreateNewBlock(): total size 1000
2014-06-07 10:03:07 CreateNewBlock(): total size 1000
2014-06-07 10:03:08 CreateNewBlock(): total size 1000
2014-06-07 10:03:08 CreateNewBlock(): total size 1000
2014-06-07 10:03:09 CreateNewBlock(): total size 1000
2014-06-07 10:03:09 CreateNewBlock(): total size 1000
2014-06-07 10:03:10 CreateNewBlock(): total size 1000
2014-06-07 10:03:10 CreateNewBlock(): total size 1000
2014-06-07 10:03:11 CreateNewBlock(): total size 1000
2014-06-07 10:03:11 connection timeout
WTF?what am I doing wrong?
hero member
Activity: 2296
Merit: 506
Cryptocasino.com
June 07, 2014, 04:20:05 AM
like now. you get 22 seconds to complete the map. This is humanly impossible having in mind the complexity of the maps.
nobody will play this except bots.

the 60 second target should not be variant.
sr. member
Activity: 434
Merit: 250
June 06, 2014, 08:29:25 PM
Would it be possible to use the nonce patterns in identification of human played maps/blocks giving them priority?

Not really, it will become increasingly hard to discern bots from humans over time anyway.

Quote
Would it be possible to increase the complexity of the maps with higher nonce values?

This wouldn't be much different from just limiting the number of maps available, I suspect.

Quote
If there is a way to force bots to emulate the way humans play, would that mean safety and survival of the coin?

Yes, but IMO as long as the chain is still hashing and no-one has demonstrated the proposed difficulty time-warp the coin is still safe and surviving, for now.

Though this could change entirely at any moment, I suppose.
sr. member
Activity: 280
Merit: 250
June 06, 2014, 07:34:49 PM
Would it be possible to use the nonce patterns in identification of human played maps/blocks giving them priority?
Would it be possible to increase the complexity of the maps with higher nonce values?

If there is a way to force bots to emulate the way humans play, would that mean safety and survival of the coin?
sr. member
Activity: 434
Merit: 250
June 06, 2014, 07:11:32 PM
nevermind, ignore me.  I must be getting tired.
member
Activity: 81
Merit: 10
June 06, 2014, 05:16:26 PM
I think you just answered your own question, sort of: simply don't decrease below the lowest threshold ever seen in a valid block. (Whether that block is in the main chain or otherwise!)

This is absolutely the same flawed idea as we have now. You can build a long chain started from the genesis block and each solution in a chain will be slow and far from optimal.
member
Activity: 81
Merit: 10
June 06, 2014, 05:10:12 PM
Nice. There is a vulnerability in motocoin that can be exploited by a bot owner to easily make a blockchain fork. Anyway this is too risky to invest money in this coin unless they close the bug.
How do you propose to fix that bug? You can't just decrease target time if there were no maps that were completed in time less than new target time because you may just accidently decrease it to a point when it becomes unachievable.

We are already very far from original satoshi idea... so I propose to use a floating threshold. If no one is able to find a block in 2 mins then threshold will be increased. Or better say that new block threshold should be somehow linked not only to the previous 2000 blocks and separation between them but also to separation between this block and previous. It is hardly to find any hole in there... looks fine to me at first sight. It may introduce some kind of oscillation but I do not see any big flaw that can be used in attack. If you are issuing new block in just one second then your solution shall be shorter than say 5 second. If you are issuing next block after a five minute of game than your solution can be very slow and not optimal.
sr. member
Activity: 434
Merit: 250
June 06, 2014, 04:53:10 PM
Nice. There is a vulnerability in motocoin that can be exploited by a bot owner to easily make a blockchain fork. Anyway this is too risky to invest money in this coin unless they close the bug.
How do you propose to fix that bug? You can't just decrease target time if there were no maps that were completed in time less than new target time because you may just accidently decrease it to a point when it becomes unachievable.

(Edit: removed poorly worded and distracting junk.  I'll try to better convey what I meant later, after more pressing matters are handled.)

Another approach would be to do something along the lines of a gravity well, or to introduce a small difficulty-reduction-over-time-by-default.  (When multi-pools started destroying chains inadvertently these solutions became popular to keep chains from stalling out.  I think this particular conundrum is similar.)
sr. member
Activity: 434
Merit: 250
June 06, 2014, 04:48:45 PM
EDIT: Wait, maybe I do see what you mean after all... basically "median" is both not actual median, and also ultimately has no impact on the retarget if target==actual. (Or is very very close.)  Am I on the right track now?
actual has no impact on the retarget if current==median.

Ah, yes, so it works both ways. (Or rather fails both ways, heh.)
sr. member
Activity: 280
Merit: 250
June 06, 2014, 04:44:16 PM
HunterMinerCrafter, WilliamLie2, you guys are awesome.
I have enjoyed just about everything with this coin so far.
Thanks for the nonce graphics as well WilliamLie, very interesting.
full member
Activity: 204
Merit: 100
June 06, 2014, 03:36:23 PM
Nice. There is a vulnerability in motocoin that can be exploited by a bot owner to easily make a blockchain fork. Anyway this is too risky to invest money in this coin unless they close the bug.
How do you propose to fix that bug? You can't just decrease target time if there were no maps that were completed in time less than new target time because you may just accidently decrease it to a point when it becomes unachievable.
full member
Activity: 204
Merit: 100
June 06, 2014, 03:32:18 PM
EDIT: Wait, maybe I do see what you mean after all... basically "median" is both not actual median, and also ultimately has no impact on the retarget if target==actual. (Or is very very close.)  Am I on the right track now?
actual has no impact on the retarget if current==median.
sr. member
Activity: 477
Merit: 250
Blockchain Just Entered The Real World
June 06, 2014, 03:31:39 PM
Dont really understand tech discussion but, I freakin love the idea. Love that game, such classic!
sr. member
Activity: 434
Merit: 250
June 06, 2014, 03:23:31 PM
I'm still not convinced that this is meaningfully exploitable.  Anyone care to demonstrate?  Cool
How fast your bot can mine new blocks? Can it generate one block at least every few seconds? If so, just start to generate blocks right from the genesis block untill you will have more blocks then there are currently, then send all your newly mined blocks to the network, voila, and you will own all of motocoins except for my premine. This is just a 51% attack.

No it is not a 51% attack because it is not involve much resources to accomplish. Because in motocoin threshold value is not linked to block issue time attacker can produce blocks whose solution is stupid and slow.

It is like producing a blockchain for bitcoin where eachblock have only one leading zero. This is not a 51% attack by definition because attacker can have far less resources than the rest of the net. He only shall be able to make form blocks fast enough.
I'm just saying that current network difficulty is too low and you don't actually need to use this vulnerability to perform attack.

I still don't really see how this is any different from a 51% against any coin.  On any coin if one dominates hashing power (by some multiples) they can rewrite block history - including difficulty.  If you had sufficient hashing power you could produce a btc chain with a different final difficulty as well.  (Probably not as low as 1 bit, but certainly lower than the current difficulty.)  It would just have to be a longer chain and end with a block holding a timestamp close enough to the wall clock.  What am I missing here?

EDIT: Wait, maybe I do see what you mean after all... basically "median" is both not actual median, and also ultimately has no impact on the retarget if target==actual. (Or is very very close.)  Am I on the right track now?


full member
Activity: 204
Merit: 100
June 06, 2014, 03:02:14 PM
I'm still not convinced that this is meaningfully exploitable.  Anyone care to demonstrate?  Cool
How fast your bot can mine new blocks? Can it generate one block at least every few seconds? If so, just start to generate blocks right from the genesis block untill you will have more blocks then there are currently, then send all your newly mined blocks to the network, voila, and you will own all of motocoins except for my premine. This is just a 51% attack.

No it is not a 51% attack because it is not involve much resources to accomplish. Because in motocoin threshold value is not linked to block issue time attacker can produce blocks whose solution is stupid and slow.

It is like producing a blockchain for bitcoin where eachblock have only one leading zero. This is not a 51% attack by definition because attacker can have far less resources than the rest of the net. He only shall be able to make form blocks fast enough.
I'm just saying that current network difficulty is too low and you don't actually need to use this vulnerability to perform attack.
Pages:
Jump to: