$500 Budget PC Guide

If it's been a while since your last PC upgrade, chances are good that you stand to benefit from a refresh to your hardware. While in theory anything with a Core 2 Duo / Athlon X2 (the later variants at least) or newer is still "fast enough", gains in efficiency not to mention system responsiveness are definitely worth pursuing. And if you're having any sort of stability problems on an older PC, trying to fix things can often end up costing just as much time and money as simply starting anew.

While picking out parts for a new PC is relatively simple on one level, the trick is in choosing parts that provide a good combination of price, performance, reliability/stability, and features. Despite advances in technology, the reality remains that going out and buying the cheapest motherboard often means you get exactly what you paid for: a cheap motherboard. That might mean things like compatibility issues with certain devices (e.g. DIMMs or PCIe cards), instability, and/or other issues. Troubleshooting such problems is no fun, so saving $20 at the cost of hours of lost productivity isn't something I'd recommend.

For other parts, however, the choices are pretty straightforward. CPUs/APUs rarely cause instability or other problems on their own, so the only real factor in selecting a processor is how much performance you want and how much you're willing to spend. I'll be listing several potential processor choices below. Storage and memory are for the most part compatible with any good motherboard and are largely interchangeable -- again, the decision mostly boils down to performance and capacity versus price.

Graphics Card Guide, April 2015

I mentioned working on some other articles, so this is going to be my first real non-cryptocurrency article in some time. Given my background, it's only fitting to start with graphics cards. So you're thinking about buying a new graphics card (GPU), but not for cryptocurrency mining or anything serious -- you want to use your PC for something useful! You want to play games.

I enjoy games as well, and with a long history of benchmarking and testing games, I've been exploring this realm for a while. But what's different is that I'm not going to try and benchmark everything under the sun on a regular basis -- I have a separate site for that sort of thing. Instead, I want to talk about the bigger picture and focus on what's important and what ends up being marketing fluff.

Use your PC as intended: for gaming!

How Low Can You Go, Bitcoin?

Wow... I know a lot of people like to point at charts, talk about the cycle of investor emotions, etc. but I generally don't follow those too much. If a technology is good, it will (or at least should) eventually rise to the top, short of a catastrophic failure of marketing. Bitcoin is such a technology in my book, so I don't think it will ever fail and go to zero. The time for that was back in late 2011, and most people were saying, "See, BTC is over -- it was a con from the start. Let's move on." Those that sold at $20+ were patting themselves on the back and those that didn't were feeling sad. Except, there were others accumulating coins at a rapid pace, and just three years later if they're still holding they've reaped nearly 100X returns.

But that's all history now. What's going on today with the BTC price? I was surprised to see us go much below $300, and I figured $250 would be the likely bottom. Now we're sub-$200 again, and some are predicting an even bigger fall yet to come. I'm still skeptical, because for a huge panic to occur you would need the biggest supporters of BTC to start dumping. If they didn't dump at $500+, why would they dump now? We're still heading down, however, and until the trend changes those who like to draw lines on charts and talk about SMAs will continue to make guesses on the eventual bottom, and some of the guesses are bound to be correct.

I wrote in the post two days ago that CEX.io halted their cloud mining, and that was when the price was still at $265. At sub-$200 prices things are even worse. Consider this: ZeusMiner just sent out a newsletter about their latest "no limit bid" on the new Antminer S5. This is supposed to be one of the newest and most efficient SHA256 ASICs around, capable of 1155GH/s with a power draw of 550W -- so basically just over a 2:1 ratio. But how much would you pay for such a miner in today's market?

At the current price and difficulty, you would net about $30 per month (depending on the cost of your power). Think about that: it could take well over a year (500 days) to pay for an ASIC that only costs $500 for 1.15 TH/s. Ouch. The interesting thing is that the prices of ASICs have dropped a lot with the drop in BTC pricing, so really the manufacturers were making a killing and they've had to cut into their profit margins because obviously no one would want to pay $1000 for a miner that could take years to -- or potentially never -- reach ROI.

Of course I don't expect BTC prices to continue down indefinitely. The very reason we have the term "bottom" is because at some point the institutional day traders and investors expect it to rebound. The SNP500 is often cited as a great example of the cycle of investor emotions. Twice it reached ~1500 before dropping to half that value, but now it's at an all-time high of ~2000. Bitcoin could very well mimic that sort of behavior, just on a vastly accelerated time scale.

And here's the thing to remember: Bitcoin is not a stock, bond, or even a commodity (though it bears the closest resemblance at times to the latter). When you try to predict BTC price movements based on tools used for stocks, you will encounter problems. And even when you apply the best technical analysis possible, you are still guessing. I always love seeing TA posts on BTC, as they end up presenting three possibilities: up, down, or stay the same. Rarely do they actually state unequivocally the direction they expect BTC to move, but a week later when there's a big change, rest assured they had that possibility covered the week before: "We were right! BTC went way down!"

Fundamentally, then, nothing has changed. BTC is the same technology today as it was last week, last month, and (mostly -- a few updates have undoubtedly occurred) last year. If it was a killer idea five years ago, it remains a killer idea, and nothing has yet been able to usurp BTC as the top cryptocurrency. The market cap is now down to $2.5 billion, sure, but that will change. The same people profiting from the drop in prices will eventually read their tea leaves and decide it's time to be bullish, and we'll see the inevitable reversal.

When will that happen? No idea, but I can't see any way BTC prices get much lower short of another major hack. $150 is a likely bottom, and double digits seems ludicrous to even consider. In a few years, I still plan to be holding onto the BTC I have, and by then I expect to be back into four digits. In the meantime, enjoy the roller coaster ride.

CEX.io "Suspends" Cloud Mining

It's been an interesting day to be sure, with PayBase announcing ways to improve their price and profitability long-term on the one hand while cloud mining services have been going belly up at a rapid pace. With the falling Bitcoin prices and continued high difficulty, it was inevitable that one of the biggest cloud mining services, CEX.io, would run into problems. Today in a blog post, CEX.io states that they are "temporarily" suspending their mining services. I put "temporarily" in quotes because there are really only a few scenarios that will lead to CEX.io resuming mining:

1. The price of Bitcoin climbs to the point where it's profitable again.
2. The difficulty of Bitcoin falls to the point where it's profitable to mine again.
3. CEX.io gets more efficient hardware that makes it viable to mine even at the current price/difficulty ratio.

I actually noticed the problems with CEX.io over the past month, when I saw that my BTC balance would sometimes drop, and the past week in particular has been bad. Now, I have long since pulled out of CEX.io, so I only had 0.5GH left on the service, but over the past week my balance dropped about 5% (maybe more?) due to maintenance fees. Even worse, selling all hashing power still seems to have only partially stopped the hemorrhaging, as referrals can still result in fees. Nice, isn't it?

The real question is whether the maintenance fees are even reasonable, or if they're just high in order to secure a profit for CEX.io. I've looked at quite a few cloud mining options over the past couple of months, and without fail those focused on Bitcoin mining have looked like they would never come close to ROI. Even the best ASICs are looking rather questionable right now, as the initial cost is too high to justify.

My prediction this round is that we're due for the first real drop in difficulty in a long time for BTC. We finally reached the point where all that hashing power was using too much electricity to sustain, and with falling BTC prices suddenly everyone is in the red. Note that we did see a few small drops in December (-0.73% and then -1.37%), but these were followed by a 3% and 8% jump in difficulty on the next two cycles -- and that last one coupled with a falling price really pushed things over the edge. People who own their own ASICs can keep running them, effectively paying extra in fiat power bills to avoid the trouble of buying BTC directly, but for cloud mining this is a complete loss: you pay maintenance in BTC and you receive rewards in BTC, so effectively it's like paying 1 BTC per day to get 0.9 BTC back (whatever the exact amount is). There's no reason to continue, period.

This is also part of the bigger problem with Bitcoin: ever increasing hashing power is only possible with ever increasing price. If the latter falls, the former must eventually follow. Assuming things get bad enough, we could actually see difficulty and network hashing power drop so far that a 51% attack would be possible, though I think that's unlikely as anyone with any sort of interest in Bitcoin doesn't want that to happen. Not surprisingly, this flaw is what has driven so many alt-coins -- including Paycoin -- to adopt a Proof of Stake distribution mechanism, as there the total power requirements are relatively minuscule in comparison to Proof of Work.

Long block confirmation times and large amounts of "wasted" power are two real issues in Bitcoin. Could we actually see the cryptocurrency collapse? Well, I doubt it, but nothing is certain. If BTC does end up failing, something else will end up taking over, and more likely than not that something will not use a Proof of Work hashing algorithm. Far more likely is that we're at the point where difficulty is going to become a lot more stable. I don't think Bitcoin prices are likely to drop too much further, but if they do you can rest assured that some big ASIC farms will pull the plug while they wait for the next difficulty adjustment to see if it's worth mining again.

Worst-case, we could actually see a huge number of ASICs shut off in the near future, causing a large drop in difficulty that might take a month or more to happen. Then we'll see a huge jump in hashing power for the next cycle thanks to the difficulty drop, and suddenly all the problems alt-coins have experienced with too-long difficulty adjustment cycles will rear their ugly head in a big way with Bitcoin. There are probably enough "believers" that will keep mining come what may that Bitcoin won't have quite the roller coaster that we've seen with alt-coins, but this could definitely open the door for other cryptocurrencies to gain ground on what was once an unassailable position. It could prove to be a very interesting quarter or two....

PayCoin Finally Takes Its Correct Place on CoinMarketCap

I've been meaning to write about this for the past week or two, but CoinMarketCap has been discriminating quite heavily against PayCoin since the coin first launched. Well, they have apparently decided it's time to change their tune.

Up until now, CoinMarketCap has been using only the publicly mined portion of PayCoin when calculating the market capitalization, with a double asterisk denoting that there was a "significant premine". Note that the vast majority of PayCoin was put into the hands of HashPoint owners and miners ten days ago, so CoinMarketCap has been very slow to change their stance. What's ironic is that you could make that same case against many other coins, including coins that weren't mined at all (e.g. NXT, Ripple, MasterCoin, etc.) but the total number of coins in existence is still used for those. But I can't really blame them for being at least a bit cautious.

The net result is that instead of 12.325 million XPY multiplied by the current price, only 0.325 million was being used, resulting in XPY being ranked down around #20 (or lower) instead of in the top five. Now, however, the total circulation of XPY has been bumped up to 12,325,674 on CoinMarketCap, and with a current price of over $12 that puts PayCoin... right into the number three spot, surpassing even Litecoin. In fact, the market cap of PayCoin is now about 1.5X that of Litecoin. Wow!

This is, by far, the most successful launch of any coin seen to date. Within weeks of public availability, PayCoin has overtaken Litecoin, and let's be honest: Ripple isn't really in the same category as other cryptocurrencies. So effectively, PayCoin is now the number two cryptocurrency for market capitalization, and PayBase still hasn't come online yet. (But soon, my precious, soooon....)

Ironically, even now the current numbers on PayCoin are still incorrect, as there are 12,375,676 XPY in circulation (so CoinMarketCap is off by 50,000). I'm not sure that matters too much, and perhaps the charts are only updated a few times per day. With the number three spot secured (number two if you don't include Ripple), the next hurdle is a quite a bit further off. PayCoin is still only about 21% of Ripple's market cap, and more importantly Bitcoin is valued at roughly 27X PayCoin. But we're looking at 17 days old vs. nearly five years, so let's just wait and see.

Market cap isn't the only area where PayCoin is thriving either. Looking at trading volume -- and again, this is before PayBase is even fully launched -- PayCoin is ranked number four overall. In trade volume, LTC is still #2 with Ripple #3. What's interesting is that Bitcoin is only about 16X the trade volume of XPY, and with features like PaySave we could easily see the trading volume on PayCoin shoot up. It's going to be an interesting week, regardless!

HashProfit Update: DDoS and Hacking?

If you're curious about why I'm cautious with HashProfit, the answer can be found in the past 12 hours. Starting some time last night, there was supposedly a DDoS (Distributed Denial of Service) attack on their server(s). I don't know if that's true or not, but the site was giving a 502 error earlier this morning when I checked. The site is now back online, but payouts didn't happen last night that I can see, and what's more you get news posts like this one. The English translation is so poorly done that it's worth quoting here (and I wouldn't be surprised if the site disappears in the near future):

Attention! There is highly coordinated and massive DDoS attack on Hash Profit service taking place right now.
27.12.2014 15:18

Site and infrastructure of Hash Profit is under unprecended massive DDoS attack right now.

Attackers spreading the word of panic about collapse of our service and other nonsence with only one main target - to provoke mass panic.

Many users neglecting common safety or being unaccurate has compromented their accounts and that has led to availability of malefactors to sell 15% of their hashrate for PFC and sell these coins via the exchange service, which in turn led to downgrade of PFC exchange rate.

All this is happening right now because lead companies of this market has understood the threat from HP service for their place in the market and they're doing everything possible to drown the company.

Right now HP team is working hard to reflect all the types of threats and the most important is defending of keeping and payout system - that's the main target of attackers.

Also our email is overcrowded with spam and rubbish from hacked users' mails, and that makes fast responses to real users questions very hard.

Main help from our service's users in solution of this difficult problem:

1. Don't panic.
2. Do everything possible to defend your own accounts - mails, BTC-wallets and computers.

We've succesfully deflected serious threats and attacks on our service earlier and we're sire that together we can do this even in a situation this hard.

The gist of the message is pretty clear -- "Hash Profit is fine, don't worry!" The problem is that I've seen and heard that same story so many times in the cryptocurrency world that it's practically a chorus. Basically they're claiming some users have had their accounts hacked, but it's only because those users "neglected common safety", not because of any error on the side of Hash Profit. The only way this would be true is if the users in question didn't set up any two factor authentication (2FA) and if they used the same password on multiple sites.

Assuming that actually happened (again), shame on the users for being stupid. More likely however is that the site itself was hacked and they're just trying to pass the buck. Or perhaps the site was a scam since the start and this is just part of the scam. The hack has supposedly allowed the hackers to sell hashing power for Profitcoin (PFC), which has then been dumped on the open market resulting in a massive crash in the price of PFC -- it's down to around 0.00077 BTC per PFC now, less than a fourth of the price just a few days ago.

Here's the real question: can Hash Profit recover and continue to mine and pay out BTC users? Or is this really just a big ponzi scheme that is now imploding and they're making a story as part of their exit strategy? "Sorry, we got hacked. Sucks to be us, and we lost all your coins. Dasvidaniya!" That's the worst case scenario right now, but it's a real possibility and in fact quite likely; hence I can't recommend anyone buy any more hashing power or PFC right now.

If you like to gamble a bit, there's a potential bright side. Let's say Hash Profit really hasn't lost a ton of money, they're not a ponzi scheme, and they still have a bunch of hardware that can happily hash away and make money for them. Their profits to date have been rather suspect -- no other site really is paying out as well as they have, and their "proprietary SmartMining algorithm" sounds like a bunch of hot air. Anyway, if you were to exchange 1BTC into PFC right now and then buy SmartMining KH at 11 KH per PFC, a single Bitcoin could buy about 14000 KH/s. 14000 KH at the current rate of payouts (assuming those continue) would pay for itself in roughly ten days. Ha!

You know the old saying: if it sounds too good to be true, it probably is. I will be very surprised at this stage if payouts continue at the previous rate, or even if they pay out at all. In fact, until I see more BTC paid out (e.g. tonight), Hash Profit has a huge red flag waving in the breeze. I'll update this post as things develop. Hopefully I'm just being overly pessimistic, and I stand to lose a moderate chunk of BTC if the site ends up being a scam, but worse things have happened to me with cryptocurrencies already.

Update, 12/27/2014 @ 11:32 PM PST: My BTC balance at Hash Profit has updated, and so have the balances of at least two friends that I checked with. The problem now is that the automated payouts haven't happened in well over 24 hours. Oh, and the PFC price temporarily rebounded and then took a further tumble. ROI time is now a ludicrous six days, but there's no way they can sell this amount of hashing power (income) at such low prices. Until I see otherwise, my suspicion is that all BTC at Hash Profit will remain at Hash Profit.

Dasvidanya, Hash Profit, 12/28/2014 @ 10:30 AM PST: The site is offline once more, payouts didn't happen, and some users have reported receiving email responses but nothing that's worth anything. LTCGear imploded in a similar fashion not too long ago, and it looks like today it was Hash Profit's turn. It's looking more and more like the days of profitable cloud mining are over... except for GAW, which continues to do well. Let's hope they at least can buck the trend.

BTC: Up, Up and Away!

It's interesting to see patterns start to emerge with Bitcoin, and right now after more or less stagnating for the past three or four months, Bitcoin looks to be rebounding. How high will it go this time? I have no idea, but it sure would be awesome if it could break the previous top from last December! I have been hoping something like this would happen, and right on schedule (two weeks before Thanksgiving) we're starting to pick up steam. So here's the question: do you buy BTC now and speculate, or just sit back and watch?

Thankfully, I'm in the position of having some BTC already, so I'm just hanging onto it and seeing where things go -- I'm in this for the long haul at this point. However, I'm also in no hurry to dump thousands of dollars into BTC in hopes that it goes even higher after the investment. Sadly, I just don't have that sort of disposable income hanging around. For those who do, however, I think if you can just dump a few thousand (or a few tens of thousands) into Bitcoin and then turn a blind eye for a couple years, this is about as good a time as any to take that chance -- actually, it was a better time last month when we were at $300 or so, but whatever.

The fact is, more and more businesses are starting to adopt Bitcoin, and I think some of them are going to start saying, "Hey, instead of cashing out 100% to guarantee our profits, we ought to hang on to at least a few percent and see where things go." Can you imagine what would happen if places like Newegg decided to hold just 3% of all Bitcoins they receive? I don't know what sort of BTC volume they're doing, but it's definitely more than zero, and over months they would likely end up sitting on thousands of Bitcoins. If more companies take that same approach -- a calculated risk, so to speak -- supply and demand dictates that there will be fewer Bitcoins in circulation, and thus price will trend up.

It's important at times like this to remember: there will only ever be 21 million Bitcoins. The same coins can be used again and again with no deterioration in quality, yes, but more importantly if big companies start using Bitcoins -- and this is already starting to happen -- then a place that does millions of dollars in business per day will certainly cause some ripples by supporting BTC. In another year, we could easily be looking at five digits for Bitcoin.

I've mined and sold a lot of BTC over the past few years, but I'm glad that I've finally reached the point where I don't need to sell BTC to cover expenses. That means I can join the ranks of the true believers and simply sock away everything I earn going forward. I won't be surprised if 20-30 years from now when I'm retired, most of my retirement savings consist of BTC that I can live off for the remainder of my days. Now if I can just get universities to accept BTC for tuition, I might have my kids' education paid for as well in the coming months.

Predicting Difficulty, ROI, and Price of BTC

If you read that title and your first thought was, "Anyone that can do those three things accurately is guaranteed to make a ton of money!" Well, you're absolutely right... and your next thought was probably something along the lines of, "Too bad most of that stuff boils down to pseudo-science and lame charts that are wrong 90% of the time," again, you're right.

One of my recent posts basically had a couple of people call me out and say something to the effect of, "But you're not even factoring in the increasing difficulty of BTC mining for your calculations, so they're all wrong!" To which I responded, effectively, "Yes, and I didn't factor in price either."

Of course course the difficulty of mining BTC -- or any other cryptocurrency -- changes the ROI prospects. The price of the target cryptocurrency likewise changes ROI prospects. Factoring in both of these is basically... well, at best it's just pure guesswork, and so I've stopped bothering. At this point, we should all know that difficulty and pricing for cryptocurrencies change regularly, but there's no telling what they'll actually do.

I've said before that difficulty follows price, but that's only part of the equation. Difficulty follows price as well as hardware efficiency, with other less significant factors also playing a role. BTC has had steadily increasing difficulties every cycle for so long that everyone seems to have forgotten that at one point, the difficulty actually plummeted! It did so right after the first big crash from $30 to $2. At some point it will almost certainly happen again, so if you're really smart you'll sell all of your Bitcoin right before the next big crash and then buy back in at the bottom so you can make tons of money. (Good luck doing that, by the way.)

So again: difficulty basically follows price as well as the efficiency of the hardware, and unless you can predict both accurately, any guess at difficulty is just that: a guess. Let me give you some great examples.

If we look at April to October 2013; the average increase in difficulty for BTC was 22% per cycle (a bit less than every 12 days). Then if we look at the next six months, October 2013 to April 2014, the increase per cycle averaged 25% (with a few instances where it jumped more than 40%). Wow, difficulty never stops increasing, right? But then we go the the past six months (April 2014 to October 2014) and the average increase is now only 12.5%. More importantly perhaps, look at just the last month where difficulty only increased 7% per cycle -- or the last two increases of 1% and 3%!

The reason the difficulty increases have fallen is that the price of Bitcoin has trended down lately (about the lowest it's been since a year ago), coupled with the fact that there's not really any more efficient SHA256 ASIC hardware coming online right now. As I've pointed out previously, the best SHA256 ASICs are doing something like 0.5W per GH, and in the not too distant future (~3 months?) we might see 0.25W per GH... but the next bump after could be much further out and eventually we might only get 0.1W per GH before we basically are stuck following Moore's Curves. That means the efficiency will be very difficult to improve without a process technology shrink, those only come every two years or so now, and in many cases a process shrink only improves efficiency by 30% (or maybe 50%).

We might see some other "tricks" where companies focus on making truly efficient ASIC designs and get a boost without a process shrink (e.g. look at NVIDIA's recent GM204 launch, where they increased performance and reduced power use while staying on the same 28nm process technology), but given SHA256 hashing isn't really all that complex I wouldn't expect major breakthroughs going forward.

Short summary: difficulty increases in Bitcoin are going to slow down, and at some point we'll even see difficulty decrease. That will happen when the price and difficulty make it unprofitable to continue mining for a bunch of ASICs, at which point they get shut down (just like my AntMiner S2 ASICs are now resting peacefully). The only thing that will get those ASICs turned back on is higher BTC prices to compensate for their power use. And if we see a massive jump in Bitcoin prices to $10,000 or whatever, you can bet that difficulty will skyrocket as well.

I of course make no claims of being able to accurately predict the future difficulty or price of Bitcoin, so when I say "best-case, the current ROI for buying a Hashlet Prime is 429 days", that means that if you buy a Hashlet Prime at $34 from the Hash Market and Bitcoin stays at $338 with difficulty also staying static (which means the payouts stay at 0.0003 BTC per MH, plus 50% for Double Dipping), you'd need 429 days to break even. If the payouts drop the time to ROI will increase, and if the price of BTC goes up the time will decrease.

Hopefully that clears things up.

Hash Profit: Free 200KH/s Test

So here's an interesting one for you: Hash Profit is claiming that through a variety of coin mining algorithms, they're able to pay a whopping 0.00698 BTC per 1000 KH per day. What's more, there's an option to get a free 200 KH/s for seven days as a trial. And since it's free, I figure why not give it a try?

There are a couple options. One is that they've misplaced a decimal point (which would be crazy, since this should be based on running some calculations and automatically generating a figure, not manually updating a field). The other is more likely: they're not talking about Scrypt MH/s or any number of other algorithms, so in fact their 0.00698 "MH/s" rate is some apparently arbitrary number that they've come up with.

I'll report back tomorrow with the actual daily returns. It looks like the free 200 KH for seven days will net you about 0.01 BTC by the end of the week at the current rates. So basically, for signing up you get a free $3.50 or so of Bitcoin; I suppose there are worse ways to spend your time. :-)

In the meantime, if you're wondering, the price for 1 MH/s (1000 KH/s) is currently 1.28457 BTC, which explains the apparently crazy returns per MH. See, if you're earning 0.00698 BTC per day off of a 1.28457 BTC investment, you're time to ROI is 184 days. Six months to hit ROI would be great on most investments, but in the world of cryptocurrencies there are far too many companies that don't last that long.

Still, if you're interested in something other than GAW or LTC Gear, you could give Hash Profit a shot.

Farewell, MintPal? The Alt-Coin Apocalypse Continues

I haven't been doing much on most of the exchanges lately. With the alt-coin apocalypse, so many of the coins that have been traded quickly became more or less "worthless", so I consolidated all of my holdings primarily into BTC and called it a day. I'm also basically done with GPU mining, as the best returns only just pay for power and may a bit extra.

Anyway, I just saw today that MintPal, one of the more well-known cryptocurrency exchanges (and by no means one of the best, as they've had hacks/problems in the past) is apparently going away. Well, maybe not quite, but the managing company is filing for bankruptcy and they've "sourced a new management team" for the site. There have apparently been some additional problems with missing/lost balances over at MintPal, and if that's the true I can see how that could quickly lead to closing shop at this stage.

On that note, with so many alt-coins basically dying and becoming worthless, I suspect we'll see something similar happen with many of the exchanges that primarily deal in alt-coins. Again, I've pulled my coins off of all of the exchanges (well, except for coins that are basically worthless -- the exchanges can keep those if they'd like), and I'd strongly suggest others do likewise.

For that matter, if you're in Bitcoin for the long haul, you might want to just move your coins into cold storage rather than trusting Coinbase or any of the other big names with your holdings. I haven't gone that far yet, mostly because I still do things with my Bitcoins (like buying Hashlets, ASICs and other goods), but better safe than sorry. And if a company you're doing business with seems to be acting a bit crazy, perhaps it's because they are?