|
Finally donated. It accepted my card this time!
|
|
|
|
|
How does the main wallet know when a watching wallet generates new addresses?
Say 1000 customer's come to the site and each get a new address. Do I have to run something on the main wallet that makes it generate (the same) 1000 addresses, or is this done automatically somehow?
|
|
|
|
I would be careful here. After months of field testing with BAMT (a usb based linux that makes every attempt not to write to the USB more than required) we see that some USB keys will die after a surprisingly short time (sometimes only a few weeks, with considerably less writes than running p2pool would generate). It varies greatly from one model of key to another, but I would expect crappy keys to die in a matter of days running p2pool. Maybe you don't care, but it would be best to warn users of this potential clearly.
If you are doing everything you can to avoid writes, and you are still killing drives, the problem is the drives. I'll put a note at the top, but the answer in p2pcoin is to boot from CD or PXE if anyone is concerned about their flash drive. Or even just not bother taking the extra time to add persistent storage. So when you reboot you have to download the block chain again? That sounds like it could add a lot of downtime. EDIT: Oh. you mention rsync. I missed that. This sounds nice for someone with only one miner. It is even nicer for someone with a fleet of miners that wants to configure once and walk away. I have 4 rigs, each with 1 to 3 cards. After I made the 0.1 release this morning, I copied it up to my PXE server and rebooted all of the boxes. Bam, I'm done. If I didn't have PXE, I could just burn 4 CDs or write 4 new flash drives and swap them. Since all of the details are stored on the network, I don't have to customize the CD or USB stick for each rig. Also, if you have enough RAM to use the RAM disk for bitcoin, but you also have persistent storage, stopping the bitcoin service cleanly (which happens during a normal reboot) will update the stored copy. When it comes back up, it will only have to catch up to what happened when it was down. With a fleet of rigs, I feel like BAMT is a better choice. Running a p2pool node on every one of your miners is unnecessary. |
|
|
|
I would be careful here. After months of field testing with BAMT (a usb based linux that makes every attempt not to write to the USB more than required) we see that some USB keys will die after a surprisingly short time (sometimes only a few weeks, with considerably less writes than running p2pool would generate). It varies greatly from one model of key to another, but I would expect crappy keys to die in a matter of days running p2pool. Maybe you don't care, but it would be best to warn users of this potential clearly.
If you are doing everything you can to avoid writes, and you are still killing drives, the problem is the drives. I'll put a note at the top, but the answer in p2pcoin is to boot from CD or PXE if anyone is concerned about their flash drive. Or even just not bother taking the extra time to add persistent storage. So when you reboot you have to download the block chain again? That sounds like it could add a lot of downtime. EDIT: Oh. you mention rsync. I missed that. This sounds nice for someone with only one miner. |
|
|
|
|
The name is confusing. Makes me think it's another altcoin.
|
|
|
|
The source is modified to only serve difficulty 1 shares. P2pool calculates the hash rate based on the work submitted and I just log those hash rates at a regular interval and use that to calculate the reward distribution. It's not as sexy as modding the code to capture actual shares submitted but it works out mathematically the same and it is easier to maintain as the p2pool code gets updated. I'm just learning Python. Once I examine the p2pool source some more and get more comfortable with Python, I will probably modify the code so that I can capture actual shares submitted, or I could just ask forrestv if he could just add that capability.
Ah very nice. I think pools like this are a great idea for small miners. Goat was talking about wanting something similar. If you tracked the miner's shares, you should be able to figure out rewards on merged mining chains too. I'm sure this could be exposed as another page of json with not too much work. Ask in the main p2pool thread or maybe #p2pool in irc. |
|
|
|
|
This is a great idea.
How are you tracking the shares for your miners? Did you modify p2pool's source at all?
|
|
|
|
|
Wow. I wish I had switched to using BAMT earlier. It's awesome.
|
|
|
|
Someone should build an exchange in SL. Fill it with terminals for virwox and some sort of mt. gox bot. Could actually give me a reason to login lol.
Isn't that sort of already the case ? Right now theres ATMs on random street corners and the official virwox exchange building is kind of boring. I want a place that acutally looks like an exchange that has people in it. |
|
|
|
"I am not capable to parse myself the text greater than 1Kb in size"
What does this mean? |
|
|
|
|
Someone should build an exchange in SL. Fill it with terminals for virwox and some sort of mt. gox bot. Could actually give me a reason to login lol.
|
|
|
|
My issue is the 'could' in your first sentence. API's and protocols are about 'should' not 'could'. All of your examples can be solved without it too. Just because you can (could) doesn't mean it's a good idea.
While a service can replicate the functionality that these BIPs provide, a service requires trust . Requiring a minimal amount of trust is a pretty central component of Bitcoin IMHO 'should' = mandatory parameter 'could' = optional parameter pretty sure most APIs/protocols have both kinds. I am pretty sure you either totally misunderstood what this change is about, or have not thought enough about what requiring two completely separate security keys to sign a transaction could be used for. I don't think "could" is being used to signify an optional parameter in this sentence. What are you trying to say? |
|
|
|
My issue is the 'could' in your first sentence. API's and protocols are about 'should' not 'could'. All of your examples can be solved without it too. Just because you can (could) doesn't mean it's a good idea.
While a service can replicate the functionality that these BIPs provide, a service requires trust . Requiring a minimal amount of trust is a pretty central component of Bitcoin IMHO |
|
|
|
I haven't been able to get cgminer to load at startup on BAMT 0.5 personally. Not sure why. So I'm sticking with 0.4 for a while until I figure it out. But if you look in /opt/bamt/examples/ there will be another bamt.conf file that will show you what to change to use cgminer.
Thanks. I just got another card and am playing with 0.5. I'll look in /opt/bamt/examples for some help. |
|
|
|
Any word on getting this open sourced? I guess BAMT is a more capable substitute, but linuxcoin source would be nice to have.
What part of LinuxCoin is not open source? Last version I used was all open source stuff, as far as I noticed. Do you have access to the live-build files? That's what I want. LinuxCoin uses open source, but the source to build a custom LinuxCoin iso/img is not available. That is what I want. ah ok. No sorry I don't have access to the live build config that generates lc. OTOH, I would think if someone wanted to create an image very much like LinuxCoin from scratch, it would be a fairly simple process. There is nothing stopping someone from just taking the package list from linuxcoin, i.e. dpkg --list, and using that as the packagelist to live build. There would be a few other things not in apt, but not a big deal. I know it isn't hard, but it's already been done. Why duplicate his work? I've played around with BitSafe's live-build and its actually really easy to add things, so that's what I'm hoping for from LinuxCoin. |
|
|
|
Any word on getting this open sourced? I guess BAMT is a more capable substitute, but linuxcoin source would be nice to have.
What part of LinuxCoin is not open source? Last version I used was all open source stuff, as far as I noticed. LinuxCoin uses open source, but the source to build a custom LinuxCoin iso/img is not available. That is what I want. The author said he would if there was interest, but then hasn't been back to the forums. Hey guys !! Been off the radar for a while dealing with real life. Was thinking about reopening this project for something to ease myself back into things. Anyone interested ?  Was going to leave the mining stuff to someone else and start with a fresh platform and all source made public. If anyone's interested in getting involved PM me and get involved. Even if it's maintaining a few packages it would make the dev of this project more pleasurable lol !! |
|
|
|
|
Any word on getting this open sourced? I guess BAMT is a more capable substitute, but linuxcoin source would be nice to have.
|
|
|
|
I've made it to level 3! Not too bad so far  On which site? The OP's. I know what I have to do, I'm just not sure how yet. |
|
|
|
I've made it to level 3! Not too bad so far |
|
|
|
|
Mine with Tor. It's way easier than setting up a VPN. Slush and Eligius both have hidden services if you really want to hide your mining.
I didn't mine very long over Tor since i was just doing it to test, but the latency didn't seem to cause any problems.
|
|
|
|
|
Show Posts
