[ANN] GameCoin (GME) - Current version = v0.8.4.2

[Sondey10mg]

I'm currently testing a fix to the annoying rogue peers issue which causes the client to display that it's trying to sync even though it's not. This is an issue that has annoyed me and the dedicated miners for a long time and no doubt causes confusion for newcomers.

This fix will continue to allow the rogue peers to connect up until a deadline of August 15th 2013 at 00:00:00 (GMT) so there will still be a 2 week grace period where we will have to continue to put up with this annoyance but once the deadline is passed any peer who is not using the new client will be disconnected with extreme prejudice by the new clients.

Unfortunately this will also include all previous genuine versions of GameCoin so this is a compulsory update, hence the 2 week window to allow everyone a chance to update their clients. Failure to update will result in you being disconnected from all updated clients and therefore this will cause a hard fork between the updated client and the non-updated clients.

As I said I'm currently testing whether the fix will work as anticipated at the moment so I'll post more information later when I know the outcome of my tests.

Great work Petr1fied! Keep us updated.

[1714294688]

1714294688

[1714294688]

1714294688

[1714294688]

1714294688

[1714294688]

1714294688

[Petr1fied]

I'm currently testing a fix to the annoying rogue peers issue which causes the client to display that it's trying to sync even though it's not. This is an issue that has annoyed me and the dedicated miners for a long time and no doubt causes confusion for newcomers.

This fix will continue to allow the rogue peers to connect up until a deadline of August 15th 2013 at 00:00:00 (GMT) so there will still be a 2 week grace period where we will have to continue to put up with this annoyance but once the deadline is passed any peer who is not using the new client will be disconnected with extreme prejudice by the new clients.

Unfortunately this will also include all previous genuine versions of GameCoin so this is a compulsory update, hence the 2 week window to allow everyone a chance to update their clients. Failure to update will result in you being disconnected from all updated clients and therefore this will cause a hard fork between the updated client and the non-updated clients.

As I said I'm currently testing whether the fix will work as anticipated at the moment so I'll post more information later when I know the outcome of my tests.

Further to this, after extensive testing yesterday I've now pushed my changes to GitHub and the Windows Daemon and Windows QT are now up to date. I'm awaiting a response from maxpower regarding the Mac QT and I'll add the link for that in the first post as soon as it's available.

As mentioned above the deadline for updating is August 15th 2013 at 00:00:00 (GMT), until this deadline is reached, GameCoin v0.8.0 will still accept connections from older versions of GameCoin (and unfortunately rogue peers too). However, after this point GameCoin will only accept new peers running GameCoin v0.8.0 and above so it's absolutely essential that you update as soon as you can to avoid being blacklisted from the deadline onwards.

NOTE: If you're connected to peers which are not deemed as misbehaving after the deadline they may remain connected regardless of version. It should however not be taken for granted. All users really need to update to avoid being left without peers to connect to.

[Sondey10mg]

Updated!

[Petr1fied]

The Mac QT wallet has now been updated to v0.8.0 thanks to maxpower.  

[xorxor]

"Difficulty goes down 4 times faster than it goes up to protect against flash miners pushing the difficulty up too far."


This is serious hole in security. If this coin will ever start being worth anything, it will be destroyed immediately.

Basically: coin can be 13%'ed

[Petr1fied]

That statement is probably a little backwards and I should probably rewrite it. What I meant to say is that if the difficulty goes up too far and the flash miners leave then the difficulty will come down faster meaning the dedicated miners aren't left holding the bag.

With retargets every 48 blocks I hardly see it being destroyed when it gets to coinchoose, besides it's been field tested for about 6-7 weeks when very few people have been interested in it and I'm reliably informed it's been performing quite well to fluctuations in hash rate. I've also been monitoring it quite closely over the last few days and the difficulty has been staying within quite a tight range.

Anyway, just because it can potentially come down by as much as 8 times it's previous difficulty it doesn't necessarily mean it'll move that far, it just means that is the absolute limitation and is only twice the standard movement.

The standard is:
[up to 4x] up and [up to 4x] down.

GameCoin uses:
[up to 2x] up and [up to 8x] down.

[xorxor]

That statement is probably a little backwards and I should probably rewrite it. What I meant to say is that if the difficulty goes up too far and the flash miners leave then the difficulty will come down faster meaning the dedicated miners aren't left holding the bag.

With retargets every 48 blocks I hardly see it being destroyed when it gets to coinchoose, besides it's been field tested for about 6-7 weeks when very few people have been interested in it and I'm reliably informed it's been performing quite well to fluctuations in hash rate. I've also been monitoring it quite closely over the last few days and the difficulty has been staying within quite a tight range.

Anyway, just because it can potentially come down by as much as 8 times it's previous difficulty it doesn't necessarily mean it'll move that far, it just means that is the absolute limitation and is only twice the standard movement.

The standard is:
[up to 4x] up and [up to 4x] down.

GameCoin uses:
[up to 2x] up and [up to 8x] down.

Yes, exactly that makes it vulnerable to 13% attack.
And  that's short-term.
If used just right, you can make 0.01% attack and take over a year of mining.

[Petr1fied]

Yes, exactly that makes it vulnerable to 13% attack.

If that's the case then according to you, MinCoin is vulnerable to a 13% attack as these changes were based upon those made to Mincoin.

[xorxor]

Yes, exactly that makes it vulnerable to 13% attack.

If that's the case then according to you, MinCoin is vulnerable to a 13% attack as these changes were based upon those made to Mincoin.

That's true. If it is taken from MinCoin, MinCoin is in fact completely open to attack.

[Petr1fied]

Much as I don't want to keep messing people around and keep releasing new clients. I certainly don't want to leave GameCoin vulnerable to attack after the hard work we have all put into reviving it so perhaps another hard fork is advisable to put the difficulty swing back in balance.

[Petr1fied]

OK doing some calculations I figure we should reach block 60,000 within a day or two of August 15th. Coincidently the block is exactly divisible by 48 blocks so I think I'll aim for that block for the switch back to 4x either way. The switch back can be made with a single line change.

Fun, fun, fun.

[Petr1fied]

No, originally it was in balance. both on 4x so that's 50:50. I'm merely reverting that part of the hard fork. It'll still adjust the difficulty at 48 block intervals.

[xorxor]

No, originally it was in balance. both on 4x so that's 50:50. I'm merely reverting that part of the hard fork. It'll still adjust the difficulty at 48 block intervals.

that is safe from <51% attacks, but still worse algo that it was. If you want to shorten time you need to limit max change , 4x is to much for fast retarget.  It would be unstable, instad of slowdowns and some stealing from regular miners, you would get massive oscliations and huge stealing.

If you shorten retarget time from 2016 blocks to 48 blocks, you will not make the algo 42 times faster... it wil be :

4'835'703'278'458'516'698'824'704 times faster. YES, THAT MUCH! almost 5 octillion times....

If you limit the change @ 1.1X it will still be ~14 times faster that original. You can make it 1.2x for ~1760 times faster algo.

[Petr1fied]

It's certainly an idea for the future but I can't just keep pushing new client versions out so fast, especially as I'm relying on third parties to compile the QT wallets for me.

Also I'm not sure of the ramifications of switching to an extremely low multiplication factor. In doing so the difficulty will creep up very gradually and also fall very gradually. In these times of flash miners you don't necessarily want the movement to be too gradual, there needs to be some room for movement.

My main priority is to get past the August 15th deadline and hopefully put an end to the rogue peers that have been plaguing GameCoin forever. And of course to ensure I've not left the door open to potential attacks which is why I've acted on your advice so fast.

[Sondey10mg]

Updated to 0.8.1

So the chain is secure now?

[Petr1fied]

Updated to 0.8.1

So the chain is secure now?

Nothing has physically changed as yet.

If everything goes to plan then on August 15th 2013 at 00:00:00 (GMT) GameCoin v0.8.1 will start rejecting peers not using at least the same version. Further to that if we remain on schedule and average a block every 2.5 minutes, we should hit block 60,000 sometime on August 16th.

After this point the potential danger should be over.

Due to this becoming a guaranteed hard fork all users need to update to GameCoin v0.8.1 as soon as is possible, I'm sorry to keep spitting out new clients and calling them a compulsory update but this latest update is essential to ensure we cannot be attacked without someone having over 50% of the network.

[Petr1fied]

I guess that's a decision for a bitcointalk staff member.

Everyone really needs to update to v0.8.1 ASAP though.

[Sondey10mg]

Bump! Update required!

[ortchi]

Please test my GAMEcoin Pool

http://coin-base.net/gamecoin/

1% POOL FEE

Payoutsystem prop

only 120 confirms

[Stratum]
./minerd --url stratum+tcp://coin-base.net:3428/ --userpass Weblogin.Worker:Worker_password
./cgminer --scrypt -o stratum+tcp://coin-base.net:3428 -u Weblogin.Worker -p Worker_password
+
[LongPoll]
./minerd --url http://coin-base.net:8428/ --userpass Weblogin.Worker:Worker_password
./cgminer --scrypt -o http://coin-base.net:8428 -u Weblogin.Worker -p Worker_password

[Petr1fied]

I've been working on a payment gateway for merchants based upon the blockchain API. It basically uses all the same rules as their Payment Gateway.

If anyone wants to test it here is the url:

Code:

http://gamecoin.kicks-ass.net/api/receive?method=create&address=$receiving_address&callback=$callback_url

$receiving_address should be the GameCoin Address you want the funds to be forwarded to
$callback_url should be an URL Encoded url.

Please refer to the information on the Blockchain API, it should be pretty much identical and you should send *ok* from your callback url once you've acknowledged that you have enough confirmations on the forwarded transaction. Please give feedback if you get chance to test it. Forwarding should happen within 2.5 minutes (block target) of you sending the payment and there is a 0% fee.

If anyone wants to test but doesn't physically have any GameCoin. Please download the wallet (linked in the first post) and send me your wallet address via PM and I'll hook you up with a few coins to test with.

NOTE: The system I've written could be easily adapted to other coins.  

Please test my GAMEcoin Pool

http://coin-base.net/gamecoin/

1% POOL FEE

Payoutsystem prop

only 120 confirms

[Stratum]
./minerd --url stratum+tcp://coin-base.net:3428/ --userpass Weblogin.Worker:Worker_password
./cgminer --scrypt -o stratum+tcp://coin-base.net:3428 -u Weblogin.Worker -p Worker_password
+
[LongPoll]
./minerd --url http://coin-base.net:8428/ --userpass Weblogin.Worker:Worker_password
./cgminer --scrypt -o http://coin-base.net:8428 -u Weblogin.Worker -p Worker_password

Excellent, thanks for posting. I'll update the main post.