If we have learnt anything, it is that you never connect something that is vital to the running of the country to the internet. That makes using Bitcoin as a national currency, problematic.

Great.
I was able to download the data for the block chain size, but the reachable nodes had no such facility as far as I could tell.

The nodes graph seems to be all reachable nodes, which is useful, but not as useful as knowing full nodes which require the full block chain. Is there a way to distinguish a full node from any old node or is the user agent sufficient?

Can any point me to graphs or data showing a relationship, if any, between the block chain database size on disk  (currently at about 40ish GB) for a full node and the number of full nodes over time? Is there a correlation?

I don't mind if I have to create them from raw data. Just point it out for me please?

That's even worse. A group doesn't even have to takeover the network, only the guardian of the button. The benevolent dictator is not a solution to this.

I think the vehement opposition to admitting the issue is because when the dust settles, the solution will be that there is no safe economic model that allows miners to do anything other than mine for bitcoin. The idea that miners will gradually switch to fees to continue making money to pay for their investment will be the undoing since the alternative is for all the miners to become obsolete once all coins are mined or all clients to be able to collect fees.

This is the crux of the issue that I have real concerns about but  I don't think it is as trivial as wait until miners go out of business. That's just a boom/bust cycle waiting to oscillate.

I also think the current lack of real opposition to Bitcoin is down to the old financial institutions seeing that the current trend is pushing Bitcoin into a centralized system. They will just step in at some point and start buying up the infrastructure since whoever controls the miners; controls confirmations and therefore the money flow. How would users fare with 3 of the largest mining pools all owned by Goldman Sachs saying they want to make bitcoin inflationary? How would you stop it?

Wow. You have a voice like chocolate.  Hello Microsoft Windows Developers

Great videos and threads. I went back to the component building since I am interested in building things like Yacoin rather than using it a set piece so a lot of what you talked about was a step-aside from where I need to be. It all went great (sort of) then you got to the LevelDB  which wouldn't have been so different if it were a DLL rather than static, eh?

Can we continue via PM? I have a lot of discussion points that I'd like to cover about static and dynamic linking (why didn't you use the OpenSSL DLLs for example) which isn't really on-topic for this thread and you seem to have the patience to mentor rather than preach.

You know what? I think they should - before some government does when we can't do anything about it. At least for a 48hr period to demonstrate the consequences and it is a real threat that needs to be addressed. Lock the main devs bitcoin addresses for a couple of days to focus their attention; first on the testnet then on the real net. Put out a public warning a week before and goad the devs to stop them if they can. Treat it as a white-hat DOS test with full disclosure.

I haven't, but I will look (long, long threads so will take me a while to read through). Looks like you are using VC++. You have VSIX packages?

No. You didn't FTFY. Windows devs are generally application developers. In the same way that web developers don't compile Apache or PHP, Windows devs don't compile reuse libraries . Linus Torvalds gets it but few from Linux seem to listen to him.


That's just nonsense.

[sarcasm]
Bandwidth's pretty much cheap now; I have unlimited 100 mb/sec. Lets increase the required bandwidth because its only waiting a couple of minutes longer.
[/sarcasm]

I get fed up with the "I'm alright Jack cos I have a 10 tonne tower of hard drives." attitude. It's not a problem FOR YOU and it's cheap FOR YOU and you can do anything you like to your computer because you don't have an IT department in your bedroom.

If they just compiled libbitcoin binaries for Windows devs they would see an explosion of applications for bitcoin on Windows. You won't get Windows devs looking at bitcoin software until you supply these binaries in the same way people do for OpenSSL or SQLite. Then you would see header translations and workspaces for Codeblocks, Lazarus and Visual Studio.

We are not going to go through 500 pages of Linux speak, install Linux or a Linux emulator/VM and search far and wide for esoteric packages that noone has ever heard of just to compile your library, which will probably be unsuccessful after 2 days of getting it set up because you already had a dependency that you didn't notice!

There are a number of issues that I think get rolled into one "its too big" which really shouldn't be.

A client doesn't require a lot of CPU or memory or disk space to confirm transactions or signatures. The hardware requirements are pretty much the same now as they were when Satoshi had his brainwave. This aspect can be achieved on Raspberry Pi and even some micro controllers. You don't need huge clusters of supercomputers either now or in the future to confirm transactions and signatures.

To be able to confirm transactions and sigs, however, you need to be able to trust your current information about the blockchain. This is where things are ballooning.

The Bitcoin Core software is naive, in software terms, in that assumes the host computer is for the sole purpose of running the application. It requires this, not for operation, but for pre-processing the block-chain only. It will max out bandwidth trying to catch up, hammer the CPU to verify blocks and use a huge chunk of a disk drive without options to move the data or offload any or all of those facets. The software needs to be more cooperative and smarter in the way it stores and reassembles the block chain and it is my opinion that bittorrent, as has been adopted by the games industry for large scale distributions for this exact reason,  is the way forward there where the data can be downloaded directly to a NAS, piece by piece in reverse order with the option of how far back you want to go.

I also see that there is a real need for a ACID compliant database - after all, this is money we are talking about  - which would relegate to the history books, the godawful verifying every time you start the application (SQLite is the first choice here, I think). The latter would also enable adhoc verifications rather than the completely linear one of the hash-table method. (e.g. pick tx 1, 9 and 50 produce a hash then ask another client what they make it).

I would like to see these kinds of things. Perhaps not exactly, but in that general direction. All I see at the moment though is less and less hardware being capable of running the software and more and more centralization to those who own the infrastructure. The end result will be online web services that you access through a browser and the banks will be those services.

I agree. The mindset that tries to fill technology as it arrives should be resisted. We should be viewing this as a minimal resource system so that you don't necessarily need expensive 1st world infrastructure to participate.

My opinion is probably on the other extreme to those that think throwing TB of disk space is a non-issue but my sentiment is valid, I think. We should be trying to put our energies into making Bitcoin resource agnostic, as far as it can be, and have design goals around low footprint, low bandwidth with the aim of being able to run full nodes over things like radio. The current mindset of a "huge blockchain is OK" is an anathema to a decentralized system pushing more and more control into infrastructure-rich people and societies and I would like to see a reversal of this.

I would go further. I would advise not using online wallets at all otherwise you might as well use a bank.

Most people won't mine it unless you go back to including the miner in the client and it doesn't use the CPU 100%. That would be a good solution if you can convince the virus scanner people that bitcoin mining software is not a virus anymore. Relying on unproven psychological incentives is a political solution not technological and I prefer relying on the maths.

Check your privilege.

+1.

I have come in pretty late to this thread but your post is pretty much my view. Bitcoin is peddled as a decentralised payment system and all the latest decisions seem to be pushing it closer to centralised. The proliferation of web wallets and exchanges means that you don't need a client to *use* bitcoin - just a browser - but you are forced into trusting a 3rd party so you are pretty much back to centralised banks. The block chain is getting so large that most home users are running out of drive space on their netbooks, laptops and phones, and the verification time when you start the software is enormous unless you have a huge SSD. So again we are coming to rely on others to be guardians of the ledger, verify transactions, commit to the ledger and mine new bitcoins. That can all be done by one company!

I don't see this as a good trend. It was originally pitched there were no transaction fees. Well. Now there are and what they are is up to a very small group of people. The protocol should be thinning, IMO, not clumping into resource hungry clusters of services.

I don't think the "cost" is the main issue with storage either. I personally am annoyed at verification times for large blockchains but the real problem, as I see it, is all of these solutions to issues have the effect of centralising operation of the network into large, resource rich, enterprises.

We are already in the situation where miners are dedicated large organisations (some even have more than 51% hashing power) with enormous resources. The barrier to entry for mining is significant and has been for some time. We are also discussing large scale storage and having "Data Centers" that have the full block chain and everyone else having thin clients. This is centralising the network and at some point the miners will merge with the data centers and you will have a small number of large commercial organisations that can hold the entire network to ransom and dictate terms. Sound familiar?

That article is inconsistent (look at the table to the right hand side).
https://en.wikipedia.org/?title=Binary_prefix

Oh, this. This x1000. This x1,000,000

I keep hearing about these bodges in the protocol. It's like listening to the ECB kicking the can down the road for Greece. This coupled with some mining groups clearly achieving enough hashing power to subvert the network when, only a year or two ago, it was being poo, poo'd that it was possible; means these issues need to be put to bed once and for all - sooner rather than later.