I suppose the card that performs badly is the one that doesn't have the monitor attached to it.

A video card without a monitor is considered "not used" by the OS, so it gets clocked down to save power. A dummy plug will take care of this.

That it fits in the description, whether you include payment reference or not.

We must have different internets...

Did you include a payment reference? If you do, many bank will indeed cut off the description after the first 35 characters.

In you case, only Deposit to the btc- e.com username would get forwarded to the receiving bank.

Note that FFC ACC.9311.23733 ACC I D S59357E has 34 characters.


In my personal experience (in general, not with BTC-e), transactions that require manual intervention take about four weeks.

Probably true. A Core i7 3930K can get to 66.6 MHash/s, but almost nobody has one of those. At 6.7 MHash/s (mining speed of a Core i7 2600K), it would take over 82,000 CPUs to match the Jupiter's 550 GHash/s. Although the largest botnet ever had 30,000,000 bots...


Not a good analogy. 100,000 bicycles are still slower than a jet airplane, but 100,000 CPUs could match a Jupiter.

More or less. Moderators moderate.

Besides, there's no big different between having a default trust list a trusting just moderators. With a trust depth of 1, both come down to letting theymos choose who can label scammers as such.


Feedback from users who aren't on your trust list won't affect the trust rating you see. The untrusted feedback is even hidden by default, so it has little impact.


Isn't that what Bitcoin is all about?

The new system might generate a few false positives (I haven't formed a personal opinion about what happened between you and Goat), but the old system had too many false negatives. THeymos was the only person handing out scammer tags, so it was him against thousands of scammers...

Scammer tags have been deprecated in favor of the trust system. Your trust rating is negative, so it seems to work.

Of course they might still have your BTC, but what does that have to do with anything? They received BTC and sent USD. Whether those BTC were yours or not is rather irrelevant. It's also unlikely, since exchanging them would require waiting for two transactions to confirm, and payments via BitPay are instantaneous.

You bought an item listed in USD. Instead of exchanging your BTC for USD and then paying with USD using either an online service or the banking system, you used BitPay to eliminate a few additional, time consuming and potentially expensive steps. Without BitPay, you'd get your refund in USD as well.

Think about it:

You buy an item worth 1,000 USD from a seller while the BTC/USD exchange rate is at 500, so you send 2 BTC to Bitpay and the seller receives 1000 USD. Shortly after paying, the exchange rate climbs to 1,000. If it was up to you, you'd immediately demand a refund, the seller would returns 1,000 USD to Bitpay and Bitpay would return 2 BTC to you.

What prevents somebody from sending 1,000 or 10,000 or even 100,000 USD worth of BTC to either a friend or himself when the course is highly volatile, demand a refund if the exchange rate rises and keep the USD if it falls? It's essentially a sell with a buy-back option. Neither BitPay nor anybody else is going to offer that.

I don't know about you, but if somebody who scammed me told me he lives in Africa, I'd expect him to be everywhere but there.

Where did you get the Bitcoin client from? The only executables I have encountered so far that would access an FDD periodically were either malware scanners or malware...

Sort of. Passwords should be hashed, not encrypted. Encryption is reversible and would require an encryption key that has to get stored as well on the server. If somebody obtains access to the encrypted passwords and the key, obtaining the actual passwords is straightforward.

Luckily, this is not the case here. The first two lines of the dump say:


"$1" means that the MD5 hash of the user's password salted with "E1xAsgR1" is "vPt0d/L3f81Ys3SxJ7rIh/". As long as somebody is using a strong enough password, MD5 works reasonably well for this purpose, i.e., it isn't possible to obtain the password from the salted hash.

However, MD5's speed makes brute-force attacks on weak passwords considerably less expensive than deliberately slow functions like bcrypt, scrypt of simply thousands of iterations of SHA-512. Even my OS uses the latter by default. I'd expect the same fro a service handling my money...

Bottom line: Don't use weak passwords! Never, ever, reuse a password!

Inputs.io
redtwitz:0.08209722:0

So far, no response to my email.

Why so complicated? Why wouldn't you just install the wallets in Ubuntu?


Software for 12.04 is more up-to-date, so I'd choose that one.

If the CPU is slow and/or old, the paravirtualization is going to be a much bigger problem than the GUI. In any case, you could install GNOME Fallback (sudo apt-get install gnome-session-fallback) if you don't want to waste resources on Unity or install a lighter distro (Xubuntu, Lubuntu, Mint Cinnamon/MATE), like maaku suggested.

Not even close. I just made a payment with BitPay. Blockchain.info said it would take 20 hours, but it confirmed after a little less than four minutes.

With a low priority transactions, confirmation time is basically luck. If you looks at TF's wallet, you'll see that after every 10 to 20 unconfirmed transactions, there suddenly is a confirmed one.

--temp-overheat won't prevent this kind of failure. CGMiner monitors the GPU temperature, not the VRM temperature.

No. The 85% efficient PSU draws 1840 W from the wall. 15% get converted directly into heat, the remaining 85% get converted into heat by the rig's remaining components.


By far too wasteful. Since the rigs will be mining coins only while heating, you should run them 24/7. An 18000 W rig running 8 hours per day will produce the same heat as a 6000 W rig running continuously, but it will cost thrice as much and produce only one third of the coins.

That's what I said (or at least tried to). I was merely pointing out that considering the GPU's power consumption alone isn't enough, since the remaining components require electricity as well and the PSU has to draw more power than it delivers to those components.

This is complete nonsense. Whether heat is conceptually wanted or not doesn't matter at all. A simple example is the traditional light bulb, which converts over 95% of the consumed electricity directly into heat. The visible light it produces gets converted into heat as well when it is absorbed by the bulb's surroundings (there are some exceptions), so while a light bulb is only 5% efficient as a light source, it is 100% efficient as a heater.

Energy cannot be created nor destroyed (Law of conservation of energy), so all the electricity your computer consumes has to get saved or converted into some other kind of energy. In the case of a GPU, all energy is converted into heat in the transistors and the electric circuits. Even a video card's fans actually heat up the ...


To calculate how much heat every 6990 in the setup will produce, you have to consider how much energy the computers containing them will draw from the wall. PSUs aren't 100% efficient as a power converter; the "lost" energy gets converted into heat as well.

Calculating 65 W power consumption outside the video cards, we get a total of 1565 W for a computer holding four 6990s. If the PSU operates at 85% efficiency, the whole setup will draw approximately 1840 W from the wall, giving 460 W per video card.