Other clients than Bitcoin-Qt are available, not all of which require the full blockchain: dre.redmartian.org/compare.htm

Right now, there is a hard-coded limit of 1 megabyte per block.

If we assume that limit never changes, that gives:
 1MB per block * 6 blocks per hour * 24*365.25*200 = 10,519,200 MB

... or 10.5 terabytes for the maximum size of the entire blockchain over the next 200 years (somebody check my math, I'm really good at dropping zeroes).

I expect that in 200 years 10 terabytes of storage will cost a few pennies.

---

Now whether or not that 1 megabyte per block limit should go away is hotly debated, and will be debated more and more as transaction volume increases.

Then there's the option of a lite client that doesn't require downloading the blockchain.

30 years ago we had 1 mhz 8 bit processors,   maybe a whole whopping 128k of ram, 140k fancy new 5 inch floppy disks and you might have been one of the rich guys who could afford the fancy new 1200 baud accoustical modem instead of the 300 baud normally used to hook up to your favority BBS.  Hard drives were there, I think 40 meg was the high end for consumer at a price tag of about 1200 bucks if I remember properly.

Look at what we have today.

16 core, multi gigahertz processors,  128 Gigs of ram, solid state drives or you can grab a 3 terabyte HDD fairly cheap, and baud rates in the mega and giga range.  Let's not forget the 'free' internet access at many many coffee shops, mcdonalds, etc as you travel around, and libraries too.

Somehow I have a feeling that as the size of the chains grow, so will technology so it should not be an issue really.

Aaron

+1

Nice info, thanks Gavin 

Please note that the long time it takes to download the blockchain is not a download limit. It's because for each block the client check the blockchain to be sure it's ok and yaddayadda. So basically it's epic hard disk crunching.
Just downloading the blockchain would take less than a hour now, it's not even 4GB. It's the verify part that slow down it.

You can run a full node that prunes (see Satoshis paper). The only time when you need nodes that have full block chain copies is to bootstrap new nodes from scratch, which will eventually be quite rare - but that doesn't imply there are only a few nodes, just that setting one up isn't something lots of people do every day.

Also, nothing stops you acquiring a copy of the database that you know is trustable (eg, there is a hash of it in the git repository) and bootstrapping from that, you don't have to rebuild the entire DB from scratch.

Which will save, say, 70% of space. One will still have to obtain a copy of remaining 30% at least, which will be a lot, still.

I am concerned about supposed future deficiency in new installations. Most existing installations will be shut down sooner or later for various reasons. Without enough new installations, number of full nodes can drop low enough for one powerful entity to be able to control most of the remaining.

In my opinion, few will be interested in hosting huge amount of data.

Yes, I am talking about bootstrapping from trusted data. I'm not sure if full verification of the block chain will be feasible at all.