As much as I enjoy open source, I'm more concerned with performance at this point. So closed source drivers would work for me. The performance of the open source drivers is fine for desktop use. You'd only care for gaming. However, the closed source drivers have a convenient control panel: https://lh3.ggpht.com/_Lb-i_VCD1tE/SvCmBK29usI/AAAAAAAAANg/MCkrQ1gv66Q/s400/nvidia-settings.pngIt makes it a piece of cake to set up the "span" type desktop like you want. At the moment i'm using a 5770 ATI Card with the ATI Properietary Driver. Could check the version if anyone wanted to know. My 5770 is glitchy on Linux just like you described with the ATI closed source drivers. The card is fine; the drivers just suck. So nvidia card(s) is the recommended solution, over ATI cards ? Yes. ATI's drivers are less-good on Windows and just terrible on Linux.
|
|
|
I suggest using TwinView on a multihead nVidia card. I've been using that on my Ubuntu desktop for a few years and it works great. I've had problems with ATI cards having graphics glitches like you describe.
Just like on Windows, it comes down to getting a card with good drivers. nVidia's binary drivers are excellent and fast, but closed source. nouveau (the open source driver) tends to work well, but has less features and is slower. ATI's drivers are kind of marginal. Intel's cards work very well, but they're only on integrated chipsets. It's a shame they don't make a multihead card.
What are you using now?
|
|
|
Adding a small proof of work would probably solve it. Include a timestamp and nonce in the block, take a SHA256 of the whole thing, and set the difficulty so that it's a few million times harder to generate than to verify, and don't accept blocks that are more than a few minutes old. The difficulty could be static because there's no need to keep up with Moore's law - it doesn't really matter if they can produce bad blocks more quickly in the future because the cost to verify will also decline.
|
|
|
I suggest the transaction fee should be kept as low as possible while preventing the blockchain from growing at an unreasonable rate.
Long term it may be necessary to use it to encourage mining for security, but that's not a concern until generation is much lower.
|
|
|
Install Python and Twisted (see directions at the bottom of that page), then:
pywallet.py --datadir=WHEREVER --dumpwallet
|
|
|
If you're just accepting payments, you don't need the private keys to be online at all. Just generate a bunch of keypairs then copy ONLY the public keys / Bitcoin addresses to the server. When people send you payments you can watch for them by address. You can then transfer the funds later using a secure PC, IE one that isn't connected to the web site at all. Edit: Here's a useful tool to create bulk addresses: https://www.bitaddress.org/bitaddress.org-v1.6-SHA1-162d1ff4fd1e09222cbaca6c282672ee6c195e1b.html
|
|
|
bitcoind does what you need. From "bitcoind --help": -blocknotify=<cmd> Execute command when the best block changes (%s in cmd is replaced by block hash) Then whenever you're notified of a new block use getreceivedbyaddress or listreceivedbyaddress to see if your address has received anything. is there any way to get a json representation of a specific block which includes all transaction in it? Yes, getblock does exactly that.
|
|
|
Related: https://bitcointalk.org/index.php?topic=62137.0;alltl;dr: instead of trying to connect to every miner everywhere, you simply send your transaction over the network with a generous fee. Miners include a pubkey with each block, and if they consider your fee sufficient, they use that key to pre-confirm your transaction, and flood that signature. Pro: simpler and faster; con: more p2p traffic.
|
|
|
I doubt 1gb will be that expensive in the future to store, storage capacity is still growing faster than blockchain size. Hard drives double about every two years. If the block chain is doubling every year then it will become a problem. Transaction fees are currently very low to help adoption, but I think it's inevitable that they will have to be much higher in the future.
|
|
|
Hm, I thought that it had the ability to sync to a random node. I guess not. Sorry. ![Smiley](https://bitcointalk.org/Smileys/default/smiley.gif)
|
|
|
LFS is great if you want to learn how things work, but I think he's looking more for a no-hassles, easy to use system.
|
|
|
Technically it's not that hard to do, but I don't think it's a problem that needs to be solved, and it creates a whole bunch of new problems like having to compare split history to make sense of past discussions.
It's much easier for people to use "bitcents", "millibtc", and "microbtc" if and when the need arises.
|
|
|
|