Topic: Wall Observer BTC/USD - Bitcoin price movement tracking & discussion - page 20689. (Read 26623416 times)

I only own about $2 worth of bitshares, but I'm keeping an eye on it. I know how annoying altcoin pushers are, so i'm sorry if that's what I'm doing. I thought it was tangentially related.

Hey BJA. Didn't you mean to say Bitshares Bitshares Bitshares Bitshares Bitshares Bitshares Bitshares Bitshares Bitshares Bitshares Bitshares Bitshares Bitshares Bitshares Bitshares Bitshares ??

Or something about Bitshares maybe?

Coin
Explanation

+1

We are being played against each other.  Rough consensus attack:


http://www.metzdowd.com/pipermail/cryptography/2015-August/026151.html

There is also a cap on the number of sigops per block that more directly addresses this concern, but for some reason that seems to never get mentioned.

Are we on rally now?

Their Bitcoin-related solutions might benefit from what artificially handicaps (without killing) Bitcoin. No need to distill it any further. I know you are fully aware of, and have summarily dismissed this possibility, but others are less trusting.

I have no issues with Blockstream itself, nor their work, but I tend to squirm when I see a glaring conflict of interest that not only can, but potentially has already directly affected the development of Bitcoin itself.

Coin
Explanation

Obviously there can be several takes on what does that mean in practice. If they mean "oppose any progress, change or improvement" then obviously I'm against that but then again who isn't? If the statu quo (more correct Latin grammar is statu quo not status quo) means "the system works let's keep it working", then I'm all for it, as opposed to breaking it with absurd decisions.


----

Jorge, we already had this exact same conversation, I'm not going to have it again.

There are several hypothetical attacks that could be called "spam attacks".

One type (1) is what we have seen since July this year: issue hundreds of MB of cheap transactions, to fill the queues and maybe crash the relay nodes (and generally cause trouble for software that deals with the queues.  That attack does not affeact transactions that pay fees higher than the attacker pays.  It will delay zero- and low-fee transactions.  The delay depends on the clearance C - T between the average capacity C of the network (currently ~0.80 MB/block) and the average non-spam incoming traffic T (currently ~0.45 MB/block).  Currently, with less than 50'000 USD anyone can launch an attack that clog the queues and delay low-fee transactions for 2-3 weeks or more.

Another type (2), that we have not seen yet, tries to delay some of the high-paying fee too.  To hold up 50% of the legitimate traffic, for example, an attacker of this type would keep issuing transactions with suitable fees to ensure that the top 0.80 MB of the queue always included at least 0.58 MB of his own transactions.  That is, he would have to issue C - T/2 transactions per block, with enough fees to keep the low half of the legitimate traffic perpetually out of the next block.  I don't know how much this attack would cost per hour; but I guess that an attacker with 50'000 USD budget would be able to keep it up for at least half a day.  

Yet another type (3) of attack may be a miner creating and solving a large bock to cause relay nodes and some clients to crash.   Rumor is that the 1 MB limit was introduced to prevent hypothetical type (3) attacks, or some similar one.    But there was no such attack during the 20 months or so in 2009 and 2010 that bitcoin operated with a 32 MB block size limit; and it is not clear whether such a rogue miner could do much damage except to himself.

On the other hand, presently it is the 1 MB limit that makes attacks of type (2) viable, and causes type (1) attacks to delay low-fee transactions by weeks.  

If the block size limit were to be raised to 8 MB, the effective capacity C would rise from 0.80 MB/block to maybe 6.00 MB/block .  The clearance C - T would then increase from 0.35 to 5.65 MB/block.  That would speed up the recovery after type (1) attacks some 15-fold, so a backlog that now takes 2 weeks to clear (like the one that exists now) would be cleared in 1 day.

If the block size limit were to be raised to 8 MB, a type (2) attacker who wanted to hold up 50% of the legitimate traffic would have to issue 5.77 MB of transactions per block, instead of 0.58 MB/block; but he would need to pay the same fees in order to keep the half of the traffic out.  Therefore the cost to the attacker, per hour, would be ~10 times larger.

Far from being an argument against a block size increase, spam attacks are one of the strongest reasons why the increase should be a no-brainer.  In fact, deterrence of and resilience against spam attacks demand a large increase rather than a small one.

Coin
Explanation

No, that's not what's being talked about. We don't know what the decision will be, nor how it will play out technically. The "what if" scenario just presumes that whatever the decision is, it doesn't work. It's the decision process we're talking about.

We also have a valuable limit used to avoid spam filling up blocks, I'd also like to keep it that way.

This works only under the presumption that Bitcoin competes only on the basis of transaction throughput which it does not.

It depends on how you define the status-quo.

I would define it as we currently have more than ample space in a block for genuine transactions. I'd like to keep it that way.

If it can't offer enough people usability then other systems will and it'll die. It's a trillion miles away from being 'too big to fail'. I can see why they'd grind to a halt but it doesn't exist in a vacuum and the world is a fast moving place that don't give a shit about Bitcoin's nobility.

Then again, how exactly is status quo an irrational stance?

Oh please 

I don't have any idea what his opinion is about this.

In other subjects in the past he's been completely right about the subject matter, excessive in his language. Something that has happened also to Luke. This doesn't make them any less right when they are right.

The Australian government also has a bunch of seized coins (some 20'000?) that it plans to auction.  And the Mt. Gox trustee still hasn't defined whether he can return the ~200'000 BTC as BTC, or will have to aution them for JPY.

(In fact, the Mt. Gox trustee is still trying to determine which of the former victims' claims are valid and which ones are bogus.  It seems that some of them are quite outlandish...)