Actually it relates the width of the perimeter of a circle to the width of its diameter.  (If you are going to get it experimentally, be sure to draw the figure with a very sharp pencil, and squint your eyes when measuring the widths.)

In mathematics,"ip" is the irrational number that has the same digits as "pi", but in reverse order.

Is a caravan a car or a van?

The George Foreman BATM.

Should't you first ask the FTC whether they did hire ICFI to do a survey?

What will they offer that itBit doesn't?

Headline: "The French bitcoin revolution: BNP Paribas testing crypto on its currency funds"

Columnist's interpretation:


Actual statement by the "source at the bank":


"Blockchain technology" is not bitcoin, and does not have to be bitcoin's blockchain, or any cryptocurrency blockchain.

I don't know what "post trade processes" mean, but it is definitely not trading bitcoin...

Most metals form carbides, but tin and lead are two notable exceptions.  Perhaps their carbides exist too, but decompose to the separate elements at fairly low temperatures, below the melting point of the metals; and are also decomposed by water, like the calcium carbide of carbide lamp fame.  That may explain the failure of all attempts to prepare them by direct reaction of the elements, or by reacting aqueous solutions of salts of those metals with acetylene or other carbides.  (I don know whether the latter experiment has been tried with a polar but aprotic solvent, like propylene carbonate, in place of water.)  There is however an isolated claim of a micron-thin film of PbC₂ having been detected on the inner surface of a sealed graphite crucible that was used to melt a Pb-Bi alloy and allowed to cool down very slowly.  If the report is correct, perhaps the carbide formed by reaction of the graphite with Pb vapor, when the temperature was too low to decompose it.

That is true... however:

In the 2 years before the last halving (2010-11-27 -- 2012-11-27) the price rose from ~0.28 to ~13 $/BTC (~45 times)

The price was practically flat around the halving (on 2012-11-28) and for the next 5 weeks.

In the 2 years after that halving (2012-11-29 and 2014-11-29) the price rose from ~13 to ~330 $/BTC (~25 times)

These numbers do not seem to indicate a significant correlation between the block reward and the price evolution.

Moreover, the rallies in 2013 have other explanations, not related to the reward:  Bobby Lee took the command at BTC-China, OKCoin and Huobi opened in Mailand China, and bitcoin was "discovered"  by some big investors like Fortress and marc Adreessen, and by the legion of Chinese amateur commodity traders. 

By the way: the crash of Dec/2014 was clearly due to developments in China, so the preceding rally must have been born there, too.

It is probably not about the consistency of the pulp of the berries of the Eastern Mistletoe or North America, unfortunately.

Three blocks down the street, then turn left, then one and a half blocks more.  If you are lucky, there will be a pile of money on the sidewalk waiting for you.  If not, then I don't know...

In that previous report that I linked to, they say that they had aborted the test for technical problems, they were about to retry the full test, and that their goal was to get 200 MB of backlog.  Which they did.  I see no reason to think that the test was done by some other entity.

On the other hand, no one knows who "coinwallet.eu" actually is.

After the first aborted attempt, Coinwallet.eu posted a report of the method and effects on reddit.  Hopefully they will do the same for this one.

There is no free electricity, of course; only electricity that someone else does not know he is paying for.  

Thanks! (I didn't realize that those cases were part of @BurtW's operation; I thought that they were just similar cases presented for comparison.)

Other people were charged in the same operation.  What happened to them?

Oops!  Indeed. "Nobody expects the Spanish Inquisition!"

Some nodes "censor" what they consider "spam" transactions in an attempt to cut down their own load and prevent them from being included in the blockchain.  This site still shows ~18 MB of backlog (bottom plot), which is clearing at the rate of ~10 MB/day.  

The backlog now seems to consist mainly of small-fee transactions, so the fees required to get confirmed in 2-3 blocks are back to almost the pre-test levels (when there was no backlog, so higher fees should not have bought faster service).

Coinwallet.eu, the entity ostensibly running this test, claimed originally to have a budget of 5000 euros (~5000 USD), and the goal of creating a 200 MB backlog of unconfirmed transactions.  They reached that goal on 2015-07-11, and the backlog has been clearing since then.  As predicted, it would take about a week to clear.

The incoming traffic rate reached a peak of over 110 tx/s on 2015-07-07. The average pre-test traffic was 1.4 tx/s, and it rarely went above 2 tx/s.  The network's effective capacity is less than 4 tx/s.  There are still occasional bursts of incoming traffic at 20-20 tx/s, and a background rate of ~3 tx/s. The extra traffic may be other people collecting their gazillion outputs that they sent to donation and public addresses.

Transaction fees added to about 20 BTC/day, on average, before the test.  This plot Shows that the miners collected so far 9 + 25 + 30 + 22 + 19 + 17 + 26 + 14 + 8 + 8 + 7 + 5 = 190 BTC of fees, above the normal amount, as a result of this test.  At today's price, that is ~52'000 USD of extra revenue for them.

According to statoshi.org, there is still an excess traffic, ~3 transactions/second (compared to the average of 1.4 tx/s that was the norm before the "test").  

Yet the backlog is being cleared: from its peak of ~200 MB iy is down to ~35 MB.  The slow rate of clearing is due to the 1 MB block size limit, coupled to the unwillingness of many miners to mine full 1 MB blocks.  

On the other hand, most of those transactions in the backlog are paying low fees, so the fees that are needed to get prompt service are almost back to the pre-test levels.

Definitely not.


Well, if the block size limit had been 8 MB instead of 1 MB, with the same budget the "testers" would have achieved a much smaller backlog, and it would have cleared much faster.  And clients would have got faster service by paying the same fee.

There are two issues in the block size question, that must be considered separately: how will the system behave under natural growth of its "normal" traffic, and how it will behave under a malicious spam attack.  As far as I can tell, in both situations 8 Mb will be much better than 1 MB.  

An increase in the minimum fee (and making it really minimum) would also help, in both scenarios.


Agreed.  As a test, it should have had a more useful goal than just "create a 200 MB backlog".  For this goal, it would not have been necessary to use more than the minimum fee.  Since they used higher than minimum fees, they must have had some other goal, even if unconscious.  Then they should have stated this goal explicitly and devised a fee strategy to achieve it.


Actually, if this theory is right (and I have no evidence that it is), it may have been profitable -- meaning that they could keep running the test for as long as they wanted, even with increasing fees, without running out of money...