NuPond

[masterOfDisaster]

Perfect!
Only the first reported line was below 100% (efficiency: 91.67% rejects: 0 missings: 1), all succeeding lines 100% so far!
But right after placing the order it’s often below 100%.

What you are going to propose seems reasonable compared to other grants of this type.
Do I understand your proposed grant right if I read the 1000 NBT as the nupond maximum that will be rewarded?
At least that would fit the 7.5% interest rate and the total compensation (for funds) of 75 NBT.
Is this 1000 NBT limit for a test run and do you plan to increase it later?
I think a bit more than 50 NBT should be appropriate, because of the efforts - not only for the renting the server…
You might consider changing it in future grant proposals…

[Nagalim]

Yes, all numbers will be changed in future proposals, and I am very open to trying things other than dutch auction given proper testing. the 1000NBT is for 500 buy + 500 sell and represents the total liquidity I will be compensating for. This is sure to increase in future proposals if we continue supporting nbt/btc. I have sent an email to Bter telling them about my operation plans and encouraging them to add NBT/CNY and that I would be happy to support that pair.

I mean, I’m totally down to take more money. How about 75 NBT for an even grant total of 150 NBT?

[masterOfDisaster]

Let’s see what others think.
Ultimately it will be decided based on the vote.
But I find it fair to raise your fee from 50 NBT to 75 NBT.

[Nagalim]

yah it’s hard to know what to ask for, but @woolly_sammoth made a good point about setting precedent for competition, so I think I’ll go with the higher number.

I guess I’ll go ahead and make a draft post.

[Nagalim]

But right after placing the order it’s often below 100%.

These are probably missings rather than rejects. It’s the client not putting up the order before the server does a sweep, whereas rejects are when the server tries to make an API call in the name of the client and does not succeed.

I’m still not sure yet why my past servers had so many rejects.

[cryptog]

Here the bot is rather unstable. The efficiency alternates between 100% and 0%…

I ask for 50 NBT operators fee th

It might look a bit too much compared to this for example.

[Nagalim]

Here the bot is rather unstable. The efficiency alternates between 100% and 0%…

I’ve been getting 100% for many hours now. When you get 0%, do you have missings or rejects, or is it just 0’s all across the board?

It might look a bit too much

The percentage is high, I agree, and that is what turned me off to asking for a big fee in the first place. However, as I have not earned shareholder trust, a big fee and a small target make a bigger promise that I will follow through with my plans. Perhaps I should ask for the money after the 30 days rather than upfront?

edit: I’m asking for 75 NBT upfront for the retainer and 75 NBT after the 30 days. This gives me good motivation to follow through on my proposal, as I am untested in shareholder eyes.

[cryptog]

I’ve been getting 100% for many hours now.

Getting 100% now most of the time.

[masterOfDisaster]

Same here: 100% most of the time!
And the outliers might be related to my RaspberryPi - I wouldn’t dare to blame the client application or the server
At second glance it might be related to the server, because it’s “rejects” which cause the degradation.
The question is: what leads to rejects?

[Nagalim]

Rejects are when a server API call in the name of the client fails. I’m still trying to sort out why my old servers had such high rejection rates. Maybe when I figure that out I can tell you more.
If you’ve got 100% most of the time and only a rejection or two every hour or so, that is pretty good I think. How often do you get rejections and how many at once?

[Nagalim]

I got a response from Bter staff. They are excited to have Nu support again, and are talking about also adding a NBT/USD pair when they enable the USD markets.

[masterOfDisaster]

Between
2015/04/29-22:57:40 and
2015/04/30-23:56:06 there have been

1488 samples of which
1451 have been at 100 % and 37 below.

Thats a ratio of 97.5% perfect hits!

3 of the 37 have been due to missings,
34 have been (partly multiple) rejects.

9 times single reject
5 times double reject
1 times triple reject
3 times quadruple reject
3 times quintuple reject
1 time sextuple reject
etc. (too lazy to grep any further )

I’m happy to provide further information it if helps tracking down issues!

[Nagalim]

Super helpful, thanks. I’m still getting 100% all the time, so information about your case is particularly helpful.

[masterOfDisaster]

Please keep in mind that I have the client running on a RaspberryPi.
This one is a RPI 1 and there are some other processes running like peerunity daemon etc. but only to keep the block chain in sync and not to mint.
Anyway there might be issues on my side!

[Nagalim]

ok, cool. I just restarted the server. I’m hoping to test out payment soon.

[masterOfDisaster]

Right, I got submit errors…
Then I restart the client. Makes digging in the log files easier.
It seems the balance was reset when the server was restarted. Not that I would be interested in the dust balance that was acquired by the client.
I’m more interested in testing NuPond

1 missing (91.67%) right after placing the bid and the next 3 log entries are at 100% efficiency!
I’ll check that again in some hours.
Great work so far!

[Nagalim]

Yah, I noticed the balances reset. I still have the credits files and can pay out manually when I figure it all out. I don’t have any nbt on the server anyway, I think I’m going to make a new server this weekend sometime.

[cryptog]

Working here with no issue now.

[Nagalim]

By the way, I found out the reason I was getting such a high rejection rate was because I chose the IPv6 option when creating the droplet.

[willy]

Do you have a more in-depth explanation for that?