I'm starting a project in a statistical computing course and I'd like to do it on arbitrarily high precision matrix operations.  However, I'm having trouble finding applications.

I was thinking of doing something with Kirchkoff matrices and network connectivity, but if anyone has other ideas for computational projects that could use matrix operations but suffer from roundoff error, I'd appreciate hearing about them.

Thanks
- Jack

Alright, thanks. I was reading that from the bitcoin wiki and I didn't know if Litecoin used the same formula.

It seemed unlikely given the disparity (Bitcoin's difficulty is what... 1.1 million?), but considering that litcoin is still brand new, is built for CPUs, and has per-computer measurements in khashes instead of Mhashes, that makes sense.

What exactly does the difficulty rating in litecoin translate to?

Right now I'm seeing difficulty 0.64ish, but I couldn't find what that meant in terms of hashes expected to get a block.

Edit: From http://allchains.info/ could I hypothesize that since ~ 19 Mhash/s is being run on litecoin, and that since 28,800 LTC are generated daily, that 1 khash/s translates to ~1.51 LTC/day?

Would that mean that a difficulty of 1 implies approximately 1 LTC/day/khash/s (plus or minus a lot for layered approximations)?

Bump.

I'm also having a similar problem. Whenever I try to initiate pooled mining, minerd.exe crashes (Win Vista 64-bit). I'm just setting up litecoin now, so I'm not sure if solo mining works without minerd or if it just hasn't given me an error yet.

I look forward to hearing more about this in the future though.

That's reasonable. Didn't know it was such a difference.

Thanks for the quick answer.

I have a laptop with an onboard graphics card (Mobile Intel 4 Series Chipset on a Dell Inspiron 1440). Is it possible to mine using this instead of my CPU?

I've read something about the card requiring OpenCL support but I wasn't clear whether that was related to GUIminer, or to BitCoin in general that the support was needed.

Could someone give me a quick prognosis on my GPU mining prospects?

Thanks.
- Jack

Fair enough. Thanks for reading in any case.

Bump for new pitch and CV update.

It doesn't show on my CV because it's not published, but I've also done some work with text processing. Specifically whittling down box scores of professional hockey and cricket games into a clean database of each goal or each play.

Actually, it does sound feasible.  I can see there being a draw in being able to spend 'game money' on real services. Still though.. it would take a lot of time to work out the in-game mechanics so that you're not doling out huge amounts of BTC for little cash. Perhaps if the player had the option of becoming a miner for BTC in their in-game account? Then you could give item drops (to use an MMORPG term) in a structure that paid an expected value slightly less than the BTC value of their mining.

On the other hand, if the game got popular, mining value could go down the toilet.

Not so simple after all.

Virtex sounds pretty good. Looking at the list of trades, the spread is notable, but definitely worth the extra convenience, especially since I'm not looking to buy in volume. Many thanks.

A product that could only be purchased by BTC would be great for the currency, but would it be good for the product? Unless, as you mentioned, it would be something in which anonymous purchase was integral or an advantage already, it could cost the product maker a lot of sales by restricting its market.

One hazy idea I have is patent (or other IP) insurance. Patent applications can be expensive and time-consuming, and insurance that someone else doesn't beat you to it may not be ideal if you don't want to reveal that you have something.  I can see engineers wanting something like this if they were worried about lawsuits with their employers about IP rights.

I would prefer to sell these to someone in the greater Vancouver area to save on shipping, but otherwise, I don't see a reason why I can't ship anywhere in continental North America.

If interested, please make an offer here or in PM or e-mail jackd@sfu.ca . These are likely out of date, but may still be good for reference so I'm not expecting a lot for them.

Field Theory and its Classical Problems - Charles Robert Hadlock

Applied Partial Differential Equations with Fourier Series and Boundary Value Problems 4th ed - Richard Haberman

Introduction to Coding Theory - Ron M. Roth

Fundamentals and Differential and Boundary Value Problems 4th ed. - Nagle, Saff, and Snider

Modern Physics for Scientists and Engineers 2nd ed = Taylor, Zafiratos, and Dubson (Relativity and Quanta, mostly)

Thanks for all the replies.

BTCnearme sounds good. Still it depends there being someone nearby whom is interested, and the nearest seems to be 18km away and stale.

I'll keep trying. Selling goods is a good idea, as I have some textbooks that are still of marginal value that may be too obscure otherwise. I'm looking at BTC in general as a potential new market for statistical consulting, so I'd already considered something similar.

In short, I would be willing to pay a higher commission to the trading house if it were easier (for non US at least) to trade fiat currency for BTC.

I'm interested in buying a small amount of BTC to get exposure to the currency, but if I were to go through Mt. Gox, it seems like I would need another e-wallet account to put money in a Mt. Gox account to finally buy BTC.  As a Canadian, it feels like my options for these e-wallets are quite limited.

For the small scale at which I'm looking to enter into the system, the accumulated fees and hassle of going through a secondary and then a tertiary money holding system isn't currently worth it.

Is there another exchange that I can trade either CAD or USD for BTC more directly?  It seems like it's a major limiter to casual interest the currency otherwise.

Excellent, thank you!

That explains why the first 100,000 blocks were downloaded before I'd even noticed, and why it's chugging ever so slowly now that I'm nearly caught up.

I'm a recent MSc Statistics grad, and I'm trying to get some experience and exposure in statistical consulting. (eg. Data cleaning, data analysis, text processing).  To reach a new market, I'm considering working not only for $CDN and $USD, but also Bitcoins at a discount to attract clients other statisticians might not reach.

We can discuss price in IM or e-mail, but for some time I'm intending to charge near half the recommended consulting price and a further discount for payment in bitcoins.

My CV as of Oct. 2011 is up at http://dl.dropbox.com/u/44754267/Michael_Davis_CV_Oct2011.pdf if you're interested in seeing it.


Edit: Rewrote pitch, updated CV.

Not the same thing.

Pricepoint always seemed like more a marketing thing to me. Something could have a price of $16.31, which is what a volume purchaser would see, but could be belong to the $20 pricepoint to which a retail customer often simplifies the price, at least at a visceral level.

I may be off here, but as I see it pricepoints are psychologically rather than mathematically significant prices, or a commonly rounded version of that price. It was mentioned in a talk at DICE 2010, that Webkins toys take advantage the fact that "to parents $12 and $20 are essentially the same amount".

That's one block for every 10 minutes since Bitcoin started right? That would make sense, it implies about 2.85 years worth of blocks.

Is each block in the chain the same size and overheard, or does that increase with the number of transactions that happened in that block?  I can see that becoming a major hassle given that blocks in the chain never retire. If it becomes a problem, I would consider chipping in for a bounty for a solution.

I won't be doing much mining myself because my computer is already running BOINC. There's projects that I want to see done more than the mining. I'm assuming there are others with similar tradeoffs.

I've seen on the marketplace a lot of mining services for rental, but are commercial computing systems like Amazon.com's rent-a-grid buying or mining BTC?  As I see it, commercial grids could do that and then pay individual miners for CPU when their grid was over-capacity. Seeing as people are already set up to mine, would that be much different?

It seems like trading BTC for computing power would improve the market here and give commercial grids more flexibility to handle larger urgent jobs by buying outsourced CPU work later in exchange for spare CPU work now.

Perhaps this is already happening and I'm missing/misunderstanding something.  (Still a newbie).

That answers my question perfectly, thank you greatly.