Yes indeed, I'd forgotten about that photo.  Hopefully it means they have enough chips in hand for batch 1.  And zefir's post is very positive news.

Thanks

roy

It was my assumption that they have received production wafers now but are only able to spare small numbers of chips from their first production run to offer as sample chips for DIY projects (because they need most of them to fulfull batch 1).

I'm hoping it's not the case that they've only received engineering samples from the fab so far and not production wafers.  I suspect that most of the doomsayers are assuming only engineering samples (which would mean they probably only have a dozen or so chips in hand at most, and we could be waiting weeks/months for bulk chips).

Personally I believe the former (if they only had engineering samples in hand I really doubt they could afford to spare any chips at all) but it would be good if someone with knowledge would confirm whether the latter or the former is the case to help quell some of the speculation here.

roy

Right, whatever the natural stable value of Bitcoin is (if there is one)... BTC will almost certainly enter a huge bubble and go way above it's equilibrium value, then crash below, before stabilizing.

Doesn't matter if you really think the end game is $1,000,000 per coin (which I personally doubt, but who knows?).  If the end game is $1,000,000 then it's inevitable that the final bubble will take us well above and then crash well below....  Imagine peaking at $10,000,000, crashing down to $300,000, before finally stabilizing at $1,000,000.

I'm plucking numbers out of the air (and I personally don't believe the numbers will be anywhere near that high) but however bullish you are about the fundamentals, it's pretty much inevitable this will end in a bubble.  The idea of a soft landing, where it keeps going up until it just levels off is, sadly, a fiction.

Of course, you have no idea where you are in the process except with hindsight.  Maybe the end game in $100, and $1,000 is the bubble that preceeds the stable price.  Personally I don't believe that, for all sorts of reasons based on the market and the fundamentals, but who knows?  

My Bitburner Furys don't seem to undervolt.  Even though cgminer lets you set the voltage down to 800mV 700mV, setting below the default of 900mV doesn't seem to have any affect, at least judging by the reported voltage in cgminer.  I guess that's a firmware limitation, so your mileage may vary.

Also, as it stands, cgminer doesn't allow clocking below the stock Bitburner Fury speed of 256, unless that's changed very recently.

You'd have to have some pretty good chips in your board to want to undervolt without underclocking, anyway, or you're just going to get HW errors.

roy

If this was a video taken in a third party production facility, then most obvious reason to take it down would be that they didn't have the proper permissions to shoot it in the first place - I gather from other projects that some facilities have fairly strict rules on photography.

Why would they bother to try to crack it?  The same malware that steals your wallet will just steal your passphrase next time you use it.  And even if there isn't a passphrase stealer, you're assuming you are anywhere near as good at picking strong passphrases as you think you are (the fact that you say 'password' rather than 'passphrase' is not a good start).

Using Armory is good advice, but if you have a non-trivial balance you should be using cold storage, which is where Armory comes into its own.

roy

At the risk of stating the obvious: I presume you don't actually have it over two lines like that in the batch file?  It needs to be all on one long line, otherwise the options on the second line will be ignored.

roy

In the contexts I'm familiar with, remittance refers to services in western countries designed to help expats send money back to their home country.

If that's what it refers to here, then it's not clear how much help it is to most people in China, as it would require them to have funds in the west to start with (or the ability to transfer funds from China to a western country).

roy

Localbitcoins?  (Or a local equivalent.)

Cash trades aren't banned right now, I think

roy

I always assumed this was because the chips from the first run didn't work, and they had to do a respin but chose not to go public about it.


I see no reason to disbelieve that they expect to receive some chips on 9th January.  Whether those chips will work, or need some modifications and another run is another matter.  But that's just the nature of chip design - these things don't always work first time.  And from what I've read pretty much all the Bitcoin ASIC vendors are producing their chips on compressed timescales which are quite aggressive by industry standards, so I would imagine that  increases the risk of failure.

So yes, it probably is the same as November 2012 in that no one knows yet whether they will be lucky and get it right first time.  But I think that applies to every ASIC vendor who doesn't yet have sample chips in hand.

roy

Right.  Josh writes in that thread:


I don't think anyone is claiming 4 weeks from tapeout to wafer.  Simply that BFL claim they will have wafers on 9th January (i.e. approximately four weeks from now...)

Clearly they must have already taped out for that claim to be credible.

Somehow I thought Android Wallet was a deterministic wallet?  Is it not, then?

EDIT: Ah, nm, I found the answer in another thread.  I obviously made some incorrect assumptions.

Order number: 161x
Queue Position: 67
Date of payment: 15 Sep
Method of Payment: Bitcoin
Product: Desk
Hashing Power: 400 GH/s
Destination: UK

That's true, but my point is broader than that.  Even if you're in a currency zone with a relatively liquid XBT market (the EUR countries spring to mind) then it still doesn't make sense.  If trading on a USD exchange gives a better rate than an EUR exchange why would you not take advantage of that?

You can argue that other-currency markets are less relevent because they're more expensive to access - but I'm not convinced that 'same currency' is a good proxy for identifying which markets are cheaply accessible from a particular country.  And once you get into wire transfers, fees aren't necessarily that strongly correlated with whether or not there's a currency conversion involved...

roy

Good work, I have to say that the global index is the way to go.  I really question the usefulness of currency-specific indices, at least for freely-convertable currencies.

I'm in the UK.  If I want to buy XBT, I would probably buy on Bitstamp.  Even though this is an EU-based exchange, they run their market in USD.  They accept SEPA transfers (the internal EU bank transfer mechanism) which are much cheaper than wire transfers, but only available denominated in EUR.

So what would happen is that I would instruct my bank to send funds from my GBP denominated bank account by SEPA to Bitstamp.  My bank will convert my GBP to EUR and make the transfer.  Bitstamp, on receiving the transfer, will convert the EUR into USD and deposit it in my account, and finally I would trade that for XBT.

In the past when I used Gox, I might have made a similar transfer to Gox's European account.  Again my bank would have converted my GBP to EUR and sent them to Gox, who would have deposited the EUR to my account.  I would then place a buy order on Gox.  But the way Gox's multi-currency orderbook works, my EUR buy order might well (for a small fee) get matched against a USD sell order, with Gox doing the currency conversion.  So again, multiple currency conversions.

The fact of the matter is, it makes no difference to me whether the exchange is denominated in GBP or EUR or USD or JPY - these are all freely convertible expressions of fiat, and it's often the norm that the fiat will be converted between currencies multiple times. It's just a units change (albeit one that incurs a small fee).  So the idea that anyone might care about what currency the exchange's orderbook is actually denominated in has always seemed bizarre to me.

What I care about is exchanging fiat for coins (and vice versa) and I'll use whatever exchange is best for me in terms of costs, convenience, reliability, etc.  I care not one jot what units the fiat quantities on their orderbook are quoted in.  This is one thing that Gox has understood for a long time - their multicurrency orderbook essentially regards fiat as a single market.  You can pull currency-specific datafeeds off Gox if you really want (and people like you and Bitcoincharts do).  But it's not really that meaningful because Gox will happily match buy and sell orders that are denominated in different currencies.

So I encourage you to continue to shift the emphasis away from indices that select a subset of exchanges based on the fiat currency their orderbook happens to be quoted in to global, or at least regional indices.  What we need are global (and possibly regional indices).  But what fiat currency we choose to express the index in should have no impact on how the index is calculated.

It does make sense to construct regional indices based on access to exchanges; e.g. BTC China's rate is not of much relevance to me because I have no access to that exchange.  Even though CNY is somewhat convertible these days, BTC China is only open to people with Chinese bank accounts.

But mainly - global is the way to go.... I'd put much less focus on currency-specific markets.  Oh, and any chance of history for the global index? :-)

Access to historical data is going to become an increasingly common requirement in order to provide valuations for accounting or tax purposes....  Currently we have the Coindesk BPI, but the more indices to choose from, the better.

roy

The usability warning on https://bitcoinarmory.com/download/updates-latest-commits/ probably needs removing/updating now that 0.90 is the default download.

roy

I dunno, I'm in the pre-order queue.  No idea if I'll ROI but as long as I cover the bulk of the cost I don't really care; I mine for fun and to support the network.

But I'm in the pre-order queue because I always figured you'd have to preorder to be in with a chance.  This idea that you could be in this reservation queue that was behind all pre-orders ever was only ever going to work if pretty much no one pre-ordered.  And despite pretty much everyone saying they'll never pre-order anything from anyone ever again, we all know that's not how it will pan out - the pre-order queue will be full, and the reservations won't get a look in.  Or at least, that's how it might pan out - but who knows, maybe the market is (at least temporarily) saturated right now...

Not saying Bitmine didn't handle things badly, but not sure if there was really a right answer.  They had to offer the reservation model because the community was so vocal on being anti-preorder, but the reality (I think) is that enough people have/will preorder.

roy

Out of curiosity, do they [China] block Cloudflare port 80, too, or only Cloudflare port 443?

https://bitcointalk.org/index.php?topic=354565.0 and also https://bitcointalk.org/index.php?topic=354365.0 (EDIT: actually probably got those the wrong way round - the second thread there has more info.  There may well be other thread in the Meta forum. EDIT: Also https://bitcointalk.org/index.php?topic=354570)

The main story is that someone compromised the bitcointalk.com account at the domain registrar, and then changed the nameservers so they could mount a man-in the-middle attack on the forums.  (I haven't seen any details of what they actually did, if it is known, but my personal guess would be that they intended to harvest passwords in the hope that some people would use the same credentials on online wallets or exchanges.)

The interesting detail is that the attackers used Cloudflare to host their mitm site.  Cloudflare does some (slightly controvertial) SSL hackery whereby they automatically get GlobalSign to issue a cert on your behalf when you sign up with them (techically they include your domain in the subjectAltName extension of one of their server's certs).  They'll present your site as SSL to the web at large using that GlobalSign cert even if you don't bother to implement SSL yourself (or if you only use a self-signed cert).

No doubt you agree to this in the small print when you sign up with them, and of course, once an attacker controlls your DNS, they can trivially obtain a cert by hand in half an hour or so, using any of the dozens of providers that just validate domain ownership by checking you can respond to a mail to admin@domain.com or some such.... But the Cloudflare solution makes things very easy for lazy hackers wanting to spoof an SSL site :-)

roy

You might want to clarify on the site whether your Nexus 5 phones are the "North America" version or the "Rest of World" version.

Although for me (UK) it's cheaper for me to buy from Google.