bitHopper: Python Pool Hopper Proxy

[MaGNeT]

Try updating your user agent, looks like digbtc is messing around a little

Oddly a user agent like 'curl/7.21.0 (x86_64-pc-linux-gnu) libcurl/7.21.0 OpenSSL/0.9.8o zlib/1.2.3.4 libidn/1.18' works fine as well. wth...

Code:

diff --git a/work.py b/work.py
index bcac801..836ef1d 100644
--- a/work.py
+++ b/work.py
@@ -73,7 +73,7 @@ def jsonrpc_lpcall(agent,server, url, lp):
 
 @defer.inlineCallbacks
 def get(agent,url):
-    d = agent.request('GET', url, Headers({'User-Agent':['Mozilla/5.0 (Windows NT 5.1) AppleWebKit/535.1 (KHTML, like Gecko) Chrome/13.0.782.107 Safari/535.1']}),None)
+    d = agent.request('GET', url, Headers({'User-Agent':['Mozilla/5.0 (Windows NT 5.1) AppleWebKit/535.1 (KHTML, like Gecko) Chrome/13.0.782.109 Safari/535.1']}),None)
     response = yield d
     finish = Deferred()
     response.deliverBody(WorkProtocol(finish))

That works... What did I just do?

[ed64]

Ya, that's the messed up part, just went from Chrome/13.0.782.107 to Chrome/13.0.782.109

So they must be doing something whacky, because a user agent like "curl" will work, but the Chrome/13.0.782.107 won't. Seems pretty unreliable whatever is that is happening....

Maybe an IE user agent string is better, those don't change as often and would impact more users should they choose to do anything based on user agent

[error]

Ya, that's the messed up part, just went from Chrome/13.0.782.107 to Chrome/13.0.782.109

So they must be doing something whacky, because a user agent like "curl" will work, but the Chrome/13.0.782.107 won't. Seems pretty unreliable whatever is that is happening....

Maybe an IE user agent string is better, those don't change as often and would impact more users should they choose to do anything based on user agent

I'd rather select my own so that I can just make it match my usual web browser.

[user7516]

About russian pools:
itzod - score based
kiwipool - switch to pps in next round

[joulesbeef]

yeah that worked on the digbit.

sucks on the russian sites... maybe they will miss us like digbit did. digbit probably would still be looking for that first block without hoppers.


perfect time to put them both to info...(to help find deepbit blocks you dont want them disabled)

[simonk83]

There seems to be an error with the ars api as well if I'm reading that right?

I'm currently getting api errors for:

ars
btcpool24
digbtc
btunion
polmine

[owowo]

Well I get it now with polmine... it goes now to the "evilpools-list" in my book... disabled!

[djex]

Here's a tip that I found out the hard way. Do not set penalty to 0 for any pool or after about 10 min bitHopper will slowly die and lag out. It makes a real mess.

[joulesbeef]

btcpool24 and btunion are normal digbit we just fixed above.. ars is down too but they have no reason to screw with hoppers as they are PPS.. so they are PROBABLY network issues.. or changed the site.

btunion is doing upgrades today.. and we will have problems with them all day.


BTCpool24.. is a site to be wary of.. not sure what is going on there.. but it is odd. They could just be noobs and having noob issues.. but api errors are common.. reporting they found blocks when they havent.. seems common.. I mined some pps and some prop and leaving the site on info until i get paid.



quick question.. has anyone sent a single share to deepbit prop yet.. using the new c00w?

[simonk83]

btcpool24 and btunion are normal digbit we just fixed above..

How do I apply that fix without it being committed?   Sorry, bit new to all this python stuff

[paraipan]

Well I get it now with polmine... it goes now to the "evilpools-list" in my book... disabled!

don't make rush decisions, their site is down or lags bitgtime, just checked

edit: @simonk83 you open the *.py file in question with you favorite text editor (notepad will give you headaches) then search for the string you have to remove and replace with the new one, save file and restart bH

[bb]

Well I get it now with polmine... it goes now to the "evilpools-list" in my book... disabled!

don't make rush decisions, their site is down or lags bitgtime, just checked

Same here.

[joulesbeef]

I have polmine on info until i work it out.. i will watch both numbers and see which is a lie when a block is found

yeah simon dont use notepad.. get notepad plus or notepad2 or textpad or something else.

open work.py


search for @defer.inlineCallbacks

make the changes shown in his ed64's comment

the  minus is something to cut

the plus is something to add.

basically replace the user agent with the new one

[simonk83]

Well I get it now with polmine... it goes now to the "evilpools-list" in my book... disabled!

don't make rush decisions, their site is down or lags bitgtime, just checked

edit: @simonk83 you open the *.py file in question with you favorite text editor (notepad will give you headaches) then search for the string you have to remove and replace with the new one, save file and restart bH

Right, haha, that's what I did  Just wasn't sure if there was a fancy way of patching it

EDIT: Thanks joules as well.

[c00w]

Well patch is the actual program to do the patching. I don't know if its on windows however.

[bb]

Btw, c00w, index file names are broken again. But since I seem to be the only person in the world (!) actually needing this, I will fix it tomorrow.

c00w: Did you fix this yourself already?

[simonk83]

From what I can tell, any of the slicers reduce variance at the cost of payout. I'll try to write a sim mimicking it if someone can post a non code algo for me to translate.

I'm sticking with oldDefaultScheduler - I get around 250% on average with that which is about the max efficiency for 10 pools, normally scheduled. Perfect.

Hmm.  Now I don't know what to think (or use)

Well patch is the actual program to do the patching. I don't know if its on windows however.

I use TortoiseGit which has a patch function, but not too sure how to use it (need a separate py file I think).

[zybron]

I've seen this happen once or twice today. In the DefaultScheduler, I've had two pools running (Mt. Red and Triplemining, today) with the slicing alternating between the two. However, Mt. Red lagged out and, as it should, BH switched over to Triplemining. Once Mt. Red was no longer lagging, however, it's slice count was reset, while Triplemining was still around 2000 or so. So, BH seems to want to stick with Mt. Red now until it 'catches' back up with Triplemining. Perhaps all slice timing should reset when a new server becomes available, even if it was from lagging out?

[simonk83]

I've seen this happen once or twice today. In the DefaultScheduler, I've had two pools running (Mt. Red and Triplemining, today) with the slicing alternating between the two. However, Mt. Red lagged out and, as it should, BH switched over to Triplemining. Once Mt. Red was no longer lagging, however, it's slice count was reset, while Triplemining was still around 2000 or so. So, BH seems to want to stick with Mt. Red now until it 'catches' back up with Triplemining. Perhaps all slice timing should reset when a new server becomes available, even if it was from lagging out?

Great minds think alike:  https://bitcointalk.org/index.php?topic=26866.msg439969#msg439969

Another issue I've just noticed with the slicing is that if a pool turns red for whatever reason, the slices are lost, so if you're in the middle of alternating between 2 or more pools, and one goes red, it drops out of the slice.    Then when it gets picked up again it has to spend however long catching up to the other pools before they get hopped again. 

Maybe try and tell it to remember the slice value it was last up to before it turned red, then add some sort of monitor so that when it's available again it resumes from where it was.

On that, it'd be really good to have a timer on api_disable as well that checks every so often if the pool API is back up.     ozcoin must have had some issues earlier and by the time I got to it (a couple of hours) it was stuck on api_disable until I manually restarted it.

[c00w]

@bb
Um, I fixed a website.py bug. Not sure if that was the only one.

@zybron
thats a bug. I should fix it.