Phil

What were your temps before swapping the fans? Only thing I see is from your screenshot of the webpage, which shows 85/85/86 for the max temperatures.

Beside the power save, did the sanyo denki make any difference temp wise? I notice the miner shuts down if max temp touches 99 or 100.

Across several b52s, I'm already close at 96 on one and 94 on other two. Looking to sort out these fans before summer.

I'm on a b52 here by halong, but the standard api port works:
echo '{"command":"summary+stats+pools"}' | nc $ip 4028

I personally plan to. And I don't see any issues with cygnusxi doing it as well since he publically has his name on his g+ page.

Just been a busy week at work.

I'll do it

What's up with 14 blocks with the same masternode paid... forks?

you need to check your numbers: http://folding.extremeoverclocking.com/user_summary.php?s=&u=657895
Day    Points    WUs
05.26.14    0    0
05.25.14    36,500    3

You have 36k points for 5/25, not 200k ppd. You'll probably see payment for that 36k tomorrow.

The folding link you posted http://fah-web.stanford.edu/cgi-bin/main.py?qtype=userpage&teamnum=224497&username=coma1013 also only shows 36734 points.

Have you seen any scrypt asics that have been worthwhile investments? Why not?

Asics for bitcoins came onto the scene and established the rules.. The rules are.. the market is dumb. And you don't need to price your hardware competitively as long as you can own the market.

You price your asic at a projected break-even point of a 3-6 months of mining, depending on how many suckers you can get to buy it.

What's the benefit of building asics for yourself? You own a % of the market and take on all the risk of making, maintaining hw.

What's the benefit of selling asics? You own 100% of the market, because you can price it so high that people will pay you 3-6 months worth of mining up front.. And then you sell them 1st gen stuff which will be obsolete when 2nd gen comes out in 3-6 months. Miners are doing calculations to see if it's profitable to buy the next gen.. Meanwhile, the hw you sell is ALL profit. Small base manufacturing price, 90%+ "R&D" margin.


So this plays to my point. Why would an asic manufacturer not announce as soon as they have one developed? What can they hope to gain by keeping it secret? Pennies compared what they can make if they announce.

It would be retarded for asic manufacturer to make an asic and keep it in the dark...

ahhh right.. skimmed over your post and misread.. yeah, get new iptables rules.. it'd be easy to post them, but better that people have to look it up and try to understand than just copy/paste linux commands, which never works well in my experience

No. The MAJOR difference is that chaeplin's howto specifically ASSUMES that amazon will block all ports for you. He even has a line in his iptables to ACCEPT ALL INCOMING TRAFFIC.

That would be a huge mistake to make on another setup. I specifically told him so.

6 masternodes..

3 different forks

debug log gets recycled.. no need to worry about the size.. you can also delete the debug.log file yourself without any problems

I have a question about masternode voting I brought to IRC but it was not answered.

<dimesz> can no one answer this question? what is the voting that the block miner gets to do? it decides which masternode to use?
<eu-Robert> dimesz: google gave me this brief explanation https://darkcointalk.org/threads/darkcoin-update-masternode-requirements-masternode-payments.225/ @ "How will the payments work?"
<dimesz> yeah.. that's the post with such little information that it led me to ask my question in the first place
<dimesz> each block miner only votes for one masternode? so if 20 (pools) control 99% of the hash rate, then about 20 masternodes will get paid?
<dimesz> let me clarify, then by the end of a day, only 20 masternodes will have seen payments
<InternetApe> dimesz, every block a new masternode is selected,
<InternetApe> So even if there are 500 Masternodes they all should get at least one payment a day
<dimesz> but the person who mines the block gets to vote with 1/6 of the decision which masternode gets paid...
<dimesz> so what's to stop the top 20 pools from always voting for their own masternode to get the payment?
<dimesz> if it was in a random fashion, i get that all masternodes would get elected eventually and you can use the formula blockreward*.1*576 / masternodes, etc
<dimesz> but then it says here that the masternode that gets paid is voted on.. by miners.. which will be a pool 99.99% of the time..
<dimesz> why would a pool vote for a random masternode when they can all vote for their own masternode to get the payment
<InternetApe> There is no way for them to SET/HARDCODE who they vote for
<dimesz> ahhh ok.. that's the small bit of information i was looking for
<dimesz> because when you say "vote for" it seems like they have the power to decide
<curious_> dimesz even if they could, they would lose miners for messing with the coin just as pools that approach 51% lose miners. Not worth it for a higher share of the 10% masternodes get
<dimesz> InternetApe: from the wiki, it says "The aim of this mechanism is to make reliable payments to well-behaving masternodes."
<dimesz> if that statement's true, then some masternodes will get preference
<dimesz> in what way do they get this?

Too lazy to stylize so irc logs it is.. Here is all the information I have about it so far:
From Evan:

From wiki:

The main concern here is that a pool (statistically speaking, there are no miners and only pools..) gets some kind of say on which masternodes will get the payment. How much say do they get? It's important to clarify this point.

A couple of things are the cause of this...

1. Folding at home is not an algorithm. See a previous explanation I gave in this post: (https://bitcointalk.org/index.php?topic=603757.msg6708528#msg6708528). Therefore, it is highly dependent on the project and the core it is using what the GPU is doing. So no, people who program for crypto currencies will know fuckall about maximizing gpus for F@H. This is not technically feasible because
 1a. the code must be hardware agnostic (read: NOT optimized) in order to run on the most combination of platforms, gpu generations, hw vendor configurations, driver versions.
 1b. Any optimizations would only apply to specific projects running on cores, etc.

2. Some offtopic commentary about "fully optimizing algorithms". Of course, in a project such as F@H, it would be good to maximize GPUs. But for proof of "work" cryptocurrencies, it's wasteful and INEFFICIENCY should be valued. Consider scrypt, which runs the hottest on GPUs and keeps it at almost peak wattage. How does this give end users ANY benefit? Every gpu gets a share of its hash as a % of the total hash of everyone else using GPUs (not going to bother talking about asics). Therefore, the less efficient the algorithm, the better, as everyone gets the same % while using less electricity.

I like how you can ask a question and then answer it yourself, but you can't read the first page of this post.

Do you have 6 million points total or 6 million ppd (points per day). Those 2 are different things.. What is your username?

You understand that 2.5 million points is the amount of points you've accumulated total but not the amount of points you get paid for EVERY DAY right?

You're getting 300k ppd (points per day), which you're getting credited for. This is correct. Your understanding of how maths work, that's wrong.

The folding community is VERY old. While curecoin is doing a great job at adding power to the project, we should NOT rename our team in order to snipe users from other teams. This is very frowned upon in the folding community.

Also, good post by someone else earlier that says it smacks of desperation. And by the chart above I just posted, I'm not sure why there should be any desperation at all. Let things come as they do.

Look at that trail we're leaving behind in our wake.

Then I stand corrected.

My main point on my last post however (which I forgot to make) is that I bought in on drk because despite these 3 things not being the most amazing things, drk has a good balance and remains good steady ground with all 3.

However, the achilles heel of this is that they all came from 1 person, and not a development team.