Milestone crossing for the official bitcoin client

[iCEBREAKER]

This is it : blk0002.dat is here.

Well then.  That settles that.  Bitcoin is dooomed!  Because mcorlett said so.  And when pressed, endeavored to prooove (^^) it.

It's been a lot of fun and I'll miss you guys, but now there's no way that anybody will ever again bother to download such a massive blockchain.

Over 2 GIGAbytes?  How could that much data possibly be moved?  Maybe by sneakernet (forklift) on tape drives?  Lord knows our 2400 baud modems won't do it.

Not to mention that no operating system can handle such large files, what with them being limited to a 16-bit address space.

And don't forget, this is ALL the fault of SatoshiDice.  Bad, bad, terrible, evil, wicked SatoshiDice.  Why, oh why, did they have to go and ruin Bitcoin?

/so much FUD so little time

[mcorlett]

This is it : blk0002.dat is here.

Well then.  That settles that.  Bitcoin is dooomed!  Because mcorlett said so.  And when pressed, endeavored to prooove (^^) it.

It's been a lot of fun and I'll miss you guys, but now there's no way that anybody will ever again bother to download such a massive blockchain.

Over 2 GIGAbytes?  How could that much data possibly be moved?  Maybe by sneakernet (forklift) on tape drives?  Lord knows our 2400 baud modems won't do it.

Not to mention that no operating system can handle such large files, what with them being limited to a 16-bit address space.

And don't forget, this is ALL the fault of SatoshiDice.  Bad, bad, terrible, evil, wicked SatoshiDice.  Why, oh why, did they have to go and ruin Bitcoin?

/so much FUD so little time

I see that my views give you cause to lash out at me. I'm not going to argue with you.

Have a nice day!

[iCEBREAKER]

This is it : blk0002.dat is here.

Well then.  That settles that.  Bitcoin is dooomed!  Because mcorlett said so.  And when pressed, endeavored to prooove (^^) it.

It's been a lot of fun and I'll miss you guys, but now there's no way that anybody will ever again bother to download such a massive blockchain.

Over 2 GIGAbytes?  How could that much data possibly be moved?  Maybe by sneakernet (forklift) on tape drives?  Lord knows our 2400 baud modems won't do it.

Not to mention that no operating system can handle such large files, what with them being limited to a 16-bit address space.

And don't forget, this is ALL the fault of SatoshiDice.  Bad, bad, terrible, evil, wicked SatoshiDice.  Why, oh why, did they have to go and ruin Bitcoin?

/so much FUD so little time

I see that my views give you cause to lash out at me. I'm not going to argue with you.

Have a nice day!

It's just a little sarcasm.  Harmless fun, yet effective at making the point that your FUD was transparently ridiculous.

That you would mischaracterize so mild a rebuke as 'lashing out at you' indicates just-a-little oversensitivity on your part.

Perhaps that's because the blockchain crossed the 2GB "milestone" and yet, almost nobody noticed, was effected, or indeed cared very much at all.

Except you, who felt the need to continue the argument while claiming 1) to be the poor victim of my 'lashing out' and 2) that you weren't going to argue with me.

Performative contradiction much?  Or only when your FUD amounts to nothing, except for inviting some gentle mockery?

[bc]

Overlaying morality on top of bitcoin doesn't make it more resilient to attack.

When government-run water supplies run low, fines and invective are used to punish and embarrass those who might water their lawns or wash their cars. A free market solution is to simply raise the price of water - incentivising conservation. It sounds like that's what the miners have done (de-prioritizing SD txs). Berating SD doesn't seem to have accomplished anything.

Satoshi Dice has helped make it clear how easy it is to bloat the blockchain. If that slows adoption today, that's fine. Slow and steady growth is better than the boom and bust of publicity and exposed-flaw.

Lastly, it sounds like they're filling a market need. Apparently lots of people are "rolling the dice" because they want to. This is what P2P transactions are supposed to be about. No one should be able to stop anyone from sending bitcoin to anyone. This scenario was possible ever since the genesis block. If we've got a "tragedy of the commons" in the blockchain, then we've got to solve it in the model of P2P - not by name-calling.

[deepceleron]

Welcome to my hard drive, blk0002.dat.


We are 1/3 of the way to the blkindex.dat file also being >2GB, and there are several references to this file by name in the source, so I don't think provision has been made to split this file or use multiple databases if the index file grows beyond the filesystem file size limit (such as 4GB in FAT32).

[iCEBREAKER]

Overlaying morality on top of bitcoin doesn't make it more resilient to attack.

When government-run water supplies run low, fines and invective are used to punish and embarrass those who might water their lawns or wash their cars. A free market solution is to simply raise the price of water - incentivising conservation. It sounds like that's what the miners have done (de-prioritizing SD txs). Berating SD doesn't seem to have accomplished anything.

Satoshi Dice has helped make it clear how easy it is to bloat the blockchain. If that slows adoption today, that's fine. Slow and steady growth is better than the boom and bust of publicity and exposed-flaw.

Lastly, it sounds like they're filling a market need. Apparently lots of people are "rolling the dice" because they want to. This is what P2P transactions are supposed to be about. No one should be able to stop anyone from sending bitcoin to anyone. This scenario was possible ever since the genesis block. If we've got a "tragedy of the commons" in the blockchain, then we've got to solve it in the model of P2P - not by name-calling.

All of those rational points don't mean anything because I can gainsay each one of them in a tit-for-tat manner.

You're just lashing out.  Ouch, stop lashing us!  I'm going to equate your rational argument to violence so I have an excuse to run away from the debate.  Just like that other guy.

Obey my boring, snotty, white hat morality or I shall tisk-tisk at you a second time, you terrible person.

Mcorlett is right, defending SatoshiDice and advocating proactive over reactive security is exactly the same as the immoral leaking of weak, innocent Yahoo passwords.  For shame, you SatoshiDice gambling addicts.  Get some help and stop destroying Bitcoin.

/SARCASM

[Diapolo]

Welcome to my hard drive, blk0002.dat.


We are 1/3 of the way to the blkindex.dat file also being >2GB, and there are several references to this file by name in the source, so I don't think provision has been made to split this file or use multiple databases if the index file grows beyond the filesystem file size limit (such as 4GB in FAT32).

Good question and I opened an issue for this on Github, to get some core dev attention!

Dia

[kuzetsa]

...In about three days from now people who use the official bitcoin client...

blk0001.dat will be at most 2130706432 bytes, or 2080768 KiB, or 2032 MiB.

You were wrong. Both of you were wrong:

07/10/2012  11:36 PM     2,097,281,755 blk0001.dat

I actually had blk0002 before you even started this thread!!!

[Diapolo]

...In about three days from now people who use the official bitcoin client...

blk0001.dat will be at most 2130706432 bytes, or 2080768 KiB, or 2032 MiB.

You were wrong. Both of you were wrong:

07/10/2012  11:36 PM     2,097,281,755 blk0001.dat

I actually had blk0002 before you even started this thread!!!

That could aswell be the size on your hard-drive, not the real size of the file ^^. The size on your drive depends on the cluster size.

Dia

[Timo Y]

blk0002.dat  has already grown 100 MB  

I my opinion, the minimum 0.01 BTC transaction output rule (fees or no fees) should be reintroduced.  Yes it would be a clunky solution but it would just be temporary and it would give the bitcoin infrastructure time to develop to a point where it can scale.

Spamtoshidice and friends are getting a free ride.  Yes, they are paying transaction fees. But transaction fees go only to the miners, but not to the rest of us who are donating our HDs.  10-20 GB a year isn’t that much but it’s enough to discourage casual users from running full clients and strengthening the network.

[Technomage]

Seriously. I know Bitcoin needs work on the scaling but seriously. A couple of gigabytes is absolutely nothing, you can get 8GB usb sticks really cheap these days. 500GB or 1 TB USB drives don't cost much. I would understand the problem if the blockchain was hundreds of gigabytes, but it is not. It's growing but there is plenty of time to work on the scaling and on top of that hard drives get cheaper all the time.

I've said this before and I'll say it again, not everyone needs to, or is supposed to, run a full node client. If you have a computer that struggles with a full node client, don't use a full node client, use a thin client! That's the number one scaling improvement we've had to this date so take advantage of it.

[jgarzik]

...In about three days from now people who use the official bitcoin client...

blk0001.dat will be at most 2130706432 bytes, or 2080768 KiB, or 2032 MiB.

You were wrong. Both of you were wrong:

07/10/2012  11:36 PM     2,097,281,755 blk0001.dat

Pieter said "at most".  Bitcoin will not split a single block across two files, therefore blk0001.dat will almost always be smaller than precisely 2032 MiB.

[jgarzik]

The maximum block chain growth[1] is about 1GB per week.

[1] given 1MB max block size, which cannot be changed without a hard fork

[mcorlett]

The maximum block chain growth[1] is about 1GB per week.

[1] given 1MB max block size, which cannot be changed without a hard fork

This does not account for increasing hashrates or luck.

[jgarzik]

The maximum block chain growth[1] is about 1GB per week.

[1] given 1MB max block size, which cannot be changed without a hard fork

This does not account for increasing hashrates or luck.

Hence "about"

[bulanula]

The maximum block chain growth[1] is about 1GB per week.

[1] given 1MB max block size, which cannot be changed without a hard fork

This does not account for increasing hashrates or luck.

Hence "about"

Too many grammar Nazis on this forum ...

I just noticed the blk0002.dat myself and was wondering WTF is going on but it all makes sense now.

Use a good FS like ext4 and forget about Windblows trainwreck.

[Diapolo]

The maximum block chain growth[1] is about 1GB per week.

[1] given 1MB max block size, which cannot be changed without a hard fork

This does not account for increasing hashrates or luck.

Hence "about"

Too many grammar Nazis on this forum ...

I just noticed the blk0002.dat myself and was wondering WTF is going on but it all makes sense now.

Use a good FS like ext4 and forget about Windblows trainwreck.

Even if FAT32 support would be removed, AFAIK there are used functions (was it ftell()?), which seem to have a size limitation of 4 GiB - 1 Byte (I'm sure it's in one comment in the source somwhere). So next stop 4 GB ?

Edit: From main.cpp (https://github.com/bitcoin/bitcoin/blob/master/src/main.cpp#L1947):

// FAT32 filesize max 4GB, fseek and ftell max 2GB, so we must stay under 2GB

so FAT32 max file size == 4 GiB - 1 Byte
Dia

[jgarzik]

Yes, the interface is potentially limited to signed 32-bit.

[Mike Hearn]

We're working on replacing blkindex.dat with a LevelDB which uses many smaller files. So size limits on the bdb files should not be an issue by the time they arrive. Assuming anyone is dumb enough to run Bitcoin on a FAT filing system, of course.

[Syke]

As much as I hate gambling in general, SatoshiDice is providing a valuable service to Bitcoin, submitting valid transactions with fees to the network. This is exactly what we want to happen. The blockchain is growing because people are using it. 2 GB blockchain? That's nothing. I want to see a 2 TB blockchain! Bring it on!