Topic: NA - page 272. (Read 893613 times)

This is true, I should of bolded the a lot better.

We are already doing better than Litecoin my friend, now we just need to get everyone to see it too ;-)

The moral of the story is we just have to be a lot better then litecoin. End of!

We will help make that a reality.

Would be good to do tests with the simulator soon.

I think we will end the day with less than 300 blocks and an average block time of more than 5 min.
Right now that worries me more than what CM is doing (less profitable for them, so less blocks they mine).

The average diff is steady getting higher while the hashrate does not (rather getting a bit lower).

I did a count of the times between the last 50 blocks
17 blocks with more than 3 min block time
6 blocks with 2 -3 min block time
26 blocks with up to 2 min block time

And look at the network status at https://explorer.guldencoin.com/#/ and compare the 7d avg. against the 24h avg.

Looking at the diff change between blocks it roughly fits with what to expect (max +/- 33% change), so no surprise here.

So what is causing this trend (decreasing blocks per day, increasing block times per day) and how to revert it no normal?

I was thinking about an additional DiffAdjustmentFactor based on the time of the last 576 blocks, but it would be like "let's try this screw and see what happens" without finding the cause for the actual trend first.

They started yesterday halfway, also at 30% of the blocks.  30% of less blocks today is less dump. It was at 33% today and is dropping a bit. Now at 30,7%, so no much change in that. If I was Terk, I go mine other coins.

CM has mined more blocks today than yesterday and with less total blocks. Edit: 30% like I calculated.

http://nlgstats.iblogger.org/

30% is better then 50-70%. But I can't imagine, CM still mining NLG now with this mean blocktime atm.

Better leave your upset with Terk for now Fuse, all is calmed down now and Terk is at least a bit cooperative after months.

CM won't leave.  Terk seriously thinks he's doing us a favor by mining our coin.  He's delusional.  He will mine NLG until the day that we find a way to kick him off for good.

-Fuse

Today even lower as it seems. Less dump from Clever. I think CM has to leave soon, because it is not profitable for them anymore this way  

If someone can provide this I am also looking for an API which returns total received coins on a given Guldencoin address.
This is for a little project I am working on, so Guldencoin will benefit from this for sure.

Great you join the collective mining! More people should follow your example 

I talked with a friend of my with a nice rig that will jump to NLG, He has been mining Digi for months now and whas looking for something else. He will hold just like he did on digi.
I hope it will help a little  
I also will jump in with the collective mining for 0.1 btc.

Or let /GeertJohan work on his simulator and have GuldenDiffAlgo, that will also work in future situations.

I think Fuse has hit the nail on the head here.  IMHO, digi was designed for a coin with shorter block times.  I'm no coding guru, far from it, but perhaps we should consider bringing the coin specs more in line with the designed specs of the algo. I can't help but think that 150 second block times and digi don't quite mesh properly.

If that was done, with of course a reduced block reward for a shorter block time, and a slight tweak to the diff adjustments, I'm pretty sure that things would be nearly perfect.

These are just 2 of the many posts I just read catching up on the forum after a long day's work.  Inspirational.  I'm proud to be part of this community and biting my tongue pretty hard to put the back and forth between Terk behind us.   Just remember guys.. slow and steady.  Putting up a big 800 satoshi buy wall on bittrex tomorrow only raises the maximum difficulty clever mines NLG at and could create larger block gaps.

So I have been following this coin in silence since I heard about it on the sterlingcoin ANN.
Guldencoin is in a very fortunate position that it has big enough support to withstand jump pools. These same jump pools destroyed a lot of coins into the ground as first dedicated miners left and then their weak communities gave up. Guldencoin community has a strong will which I truly believe can overcome any obstacle. I wish this was the case for Britcoin but it wasn't.

Yesterday I spoke to one of my mates about Guldencoin because I believe in this project just like most of you do and he bought some up yesterday on bittrex. Which brings me to my next point, I see plenty people wasting time crying about the Hitlers of altcoin currencies also know as jump pools. Now the only way to get nullify the effect of these devils is to grow your coin even more. More users means more purchases and more spending, higher price means more dedicated miners come and smooth out the chain. Litecoin does not have to worry about jump pools, so that coin is free of that evil that plagues the altcoin market.

Now instead of wasting time on what these parasites are doing , spend the time on promoting the coin. Don't be hesitant to promote it because it is constantly getting better and I truly believe this project will succeed even with these devils taking 45% of the coins.

I'll admit that I wasn't expecting some of the results we've been having.  I believed that we should have been having a mix of short blocks and long blocks that averaged out, like you mention.

We're overshooting the difficulty needed to effectively keep the chain moving.  Instead we're going from 500 to 2000 in the span of 3 blocks.  Then the long block time causes the maximum drop, and we end up going back down to the level that is profitable for MPs.  It's part of the reason why I recommended trimming the max adjustment.  The adjustments will happen just as fast, but they won't move as far.

Additionally, we have a lot less hashrate keeping those higher blocks moving.  So it's back and forth with high block, MP block.  Digi is doing it's job, we just need to tweak it a little.  Or we can start talking about the shorter block times.

Of course, the ideal solution is more miners, or more dedicated hashrate.

-Fuse

From https://explorer.guldencoin.com/#/  (guldencointrader.nl shows basically the same):

                       latest        24h avg       7d avg.
Hashrate               12.7 GH/s     12.1 GH/s     14.9 GH/s
Difficulty             1,068.35      722.82        710.57
Time between blocks    1 m 49s       4m 10s        3m 25s


And from http://nlgstats.iblogger.org/ :

    date  | blocks
+---------+-------+
|  Feb 01 |  491  |
|  Feb 02 |  398  |
|  Feb 03 |  377  | 
|  Feb 04 |  410  | 
|  Feb 05 |  370  |
|  Feb 06 |  337  |

Feb 6 with only 337 blocks is very low (~60% of expected 576 blocks).

Looks like while diff goes up, blocks per day goes down (talking about average here).
And this happens while the hashrate goes down too?

DIGI should and is reacting faster than DGW3, so it should level between longer and shorter block times faster too.
So why isn't the average diff lower together with the lower average hashrate?
Looking at our (simple) DIGI code again and again without seeing why.
(Even the old DGW3 with it's one-off bug and 24 blocks interval did a better job there and kept us closer to the default number of blocks per day).
Anyone able to explain what is going on here?

Let's get an actual point of reference.

Started with (edit: balance):
114927 <-- index number, not block number  (block number: 196194)    2015-02-06 08:15:48 -0500    Sent    1,000.00000000    7,000.00000000

(Added:) First block found:
114928 <-- index number, not block number  (block number: 196195)    2015-02-06 08:17:55 -0500 Received 1,000.00000000 8,000.00000000

Last found at time of the writing of this post:
114972 <-- index number, not block number  (block number: 196324)   2015-02-06 18:11:44 -0500 Received 1,000.00000000 44,000.20115000

CM had a positive, non-zero balance for the 4 days. CM has also already sent 4000 NLG since CM started back up today.
So 41 blocks in 10 hours, out of 130 blocks. 31% of blocks for ten hours.
My original post came up with a higher percent of blocks, and at the time they were banging out more blocks per hour.
I really wanted to wait to see after 24 hours what the equation would show us, but I was off although I was taking the info straight from the explorer. I would rather get closer to on point since my first post was 20-25% higher than it actually is.


-edited-