Bitcoin Forum
June 22, 2024, 10:35:24 AM *
News: Voting for pizza day contest
 
   Home   Help Search Login Register More  
Pages: [1]
  Print  
Author Topic: Scaling BTC - transaction fees and times cf demand  (Read 981 times)
davidpbrown (OP)
Sr. Member
****
Offline Offline

Activity: 531
Merit: 260


Vires in Numeris


View Profile WWW
May 05, 2013, 09:19:36 AM
 #1

Are we confident yet that BTC can scale to handle large volumes of transactions?
Can each block and each miner accept enough transactions that the market doesn't stall?

I'm seeing really slow transactions.. 20mins for the 1st and 30mins for the 2nd then upto an hour for the 6th. It's still magic but cf real world I wonder how practical that is.
Perhaps I'm wrong to have set transaction fee as 0.005BTC but then the promise is "Low or zero processing fees" and miners are making their money already.

Should all transactions be 0.01BTC and if that's the case now are we heading for WU rates in the future!.. or will miners be able to accept more transactions into each block and the fees perhaps will stay low.


On that thought, can any of the Alt-TC maintain an advantage by having faster confirmations.. or will they face the same issues later?

฿://12vxXHdmurFP3tpPk7bt6YrM3XPiftA82s
bitcork
Newbie
*
Offline Offline

Activity: 13
Merit: 0



View Profile
May 05, 2013, 09:39:03 AM
 #2

When did the offending transactions occur?
If recent, I was pondering the same issue today, especially now that the Chinese have overtaken the rest of the world in wallet downloads on Sourceforge. Ever since a documentary on BTC aired a couple of days ago on CCTV, Sourceforge has been slower than usual.
I reformatted my PC today and downloading the blockchain seemed to take longer than usual too.
It wouldn't surprise me if all of this is due to the new Asian burden placed on blockchain, after all, miners subsidize transactions right? And there certainly couldn't have been a dramatic increase in Chinese miners..
My guess is that this will be a transitory problem now that ASICs are hitting the market and difficulty will decrease substantially.
Although we're still waiting for India to join the party...
davidpbrown (OP)
Sr. Member
****
Offline Offline

Activity: 531
Merit: 260


Vires in Numeris


View Profile WWW
May 05, 2013, 09:54:20 AM
 #3

This was an hour ago.. took 40 mins to get six confirmations.. but that was waiting 20mins to get one. Even with green address then I wonder how practical that is.

Having seen what happened to TRC a couple of weeks back, I wonder whether BTC can become starved of oxygen too.

Perhaps it is inevitable that there will be fluxes in all factors while BTC and other find their feet. However, it would be good to know that we are confident that there won't be a problem; or otherwise at least that we have some sense of where the limits lie. Perhaps if we understand the potential issues, we can be more prepared with flexibility and additional capacity even.

฿://12vxXHdmurFP3tpPk7bt6YrM3XPiftA82s
edmundedgar
Sr. Member
****
Offline Offline

Activity: 352
Merit: 250


https://www.realitykeys.com


View Profile WWW
May 05, 2013, 10:21:20 AM
 #4

I'm seeing really slow transactions.. 20mins for the 1st and 30mins for the 2nd then upto an hour for the 6th. It's still magic but cf real world I wonder how practical that

Do you mean 20 mins to get to the 1st confirmation, 30 mins to get to the 2nd confirmation, up to 1 hour to get to the 6th?

If so that's pretty much expected. There's a block every 10 minutes on average, but if you're unlucky you won't get that, and your transaction may get left out of a block for whatever reason and end up having to wait for the next one.

You shouldn't expect that to get better. In a lot of cases you can treat the payment as paid within a few seconds when you see the transaction get broadcast across the network, as double-spends aren't very common, especially if your transaction is revocable (eg it takes you 30 minutes to bake the pizza, so you can get started making it right away and in the unlikely event that the customer tries to double-spend you can eat it yourself.)

Note that with PayPal and credit card companies transactions are often revoked after several months, so waiting an hour or so for non-revocability isn't too bad by comparison.

Are we confident yet that BTC can scale to handle large volumes of transactions?
Can each block and each miner accept enough transactions that the market doesn't stall?

This gets into all kinds of amazing combinations of technical and political issues. First there's a technical problem caused by a software bug which is preventing the network from scaling as it should. This will be solved soon. Then there's an old anti-spam setting hard-coded in the software that needs to be fixed, but some people are arguing against fixing it for various reasons. Then finally there's the issue of whether the thing will actually scale up really big, and what will happen if and when it runs into the actual technical limits. It could scale up many times before we found that out, though.

I think this will probably all work out in the end, but we can't rule out ending up with WU rates for Bitcoin, whether for political or technical reasons. It's not at all clear whether transactions will end up costing 0.1 cent or 1 cent or 10 cents or $1, and obviously that affects the ways you can use it. But I'd have thought that if Bitcoin can't scale cheaply some variant of it will end up scaling cheaply. I'm not sure whether this would end up coexisting with Bitcoin or replacing it.

Perhaps I'm wrong to have set transaction fee as 0.005BTC but then the promise is "Low or zero processing fees" and miners are making their money already.

Should all transactions be 0.01BTC and if that's the case now are we heading for WU rates in the future!.. or will miners be able to accept more transactions into each block and the fees perhaps will stay low.

You could probably have gone lower than that and still got your transaction through. Some of the client software defaults are probably higher than they need to be right now, because they were set when the exchange rates were lower.

On that thought, can any of the Alt-TC maintain an advantage by having faster confirmations.. or will they face the same issues later?

LiteCoin has faster confirmations - 1 block every 2.5 minutes IIRC. I guess the question is whether this difference will turn out to be meaningful. In practice in most of the cases where an average of 10 minutes is too long, 2.5 minutes is probably too long as well.

On the Bitcoin / LiteCoin design there's a limit to how short you can make this, because you need to make sure blocks have time to get propagated across the network properly.
davidpbrown (OP)
Sr. Member
****
Offline Offline

Activity: 531
Merit: 260


Vires in Numeris


View Profile WWW
May 05, 2013, 10:40:42 AM
 #5

It was ~40 mins total. I was watching blockchain.info and local wallet, no difference between them.

> There's a block every 10 minutes on average, but if you're unlucky you won't get that
That's the bit I'm unclear on.. I don't understand exactly how that is determined relative to the network. I've read that the older chain become accepted but what does that mean for propagation and confirmations.

Is there a limit on how many transactions each miner accepts in each block? If the world was using BTC, would the part of the network less busy risk stalling and being unlucky on multiple occasions?.. I'm thinking not because it seems to be a matter of some luck which block=[set of transactions] gets accepted.. and the longer you wait, perhaps the more widely broadcast the transaction becomes.. maybe even minor prefering older transaction requests first??.. but I'm guessing and it does seem the developers are well ahead of the users.




฿://12vxXHdmurFP3tpPk7bt6YrM3XPiftA82s
edmundedgar
Sr. Member
****
Offline Offline

Activity: 352
Merit: 250


https://www.realitykeys.com


View Profile WWW
May 05, 2013, 03:27:45 PM
 #6

It was ~40 mins total. I was watching blockchain.info and local wallet, no difference between them.

> There's a block every 10 minutes on average, but if you're unlucky you won't get that
That's the bit I'm unclear on.. I don't understand exactly how that is determined relative to the network. I've read that the older chain become accepted but what does that mean for propagation and confirmations.

Is there a limit on how many transactions each miner accepts in each block? If the world was using BTC, would the part of the network less busy risk stalling and being unlucky on multiple occasions?.. I'm thinking not because it seems to be a matter of some luck which block=[set of transactions] gets accepted.. and the longer you wait, perhaps the more widely broadcast the transaction becomes.. maybe even minor prefering older transaction requests first??.. but I'm guessing and it does seem the developers are well ahead of the users.

There are two issues here: First how long it takes until a block is created, secondly whether or not your transaction is included in a given block.

Blocks are created basically at random times, but at a speed that works out to 6 per hour, or 1 every 10 minutes. So after you broadcast a transaction a block may happen to get created within the next minute or two (if you're lucky) or there may not be any for half an hour (if you're unlucky). You can see that here:
http://blockchain.info/blocks

Whether your transaction is included in any given block is up to the miner who mines it. They don't have to include any transactions at all unless they want to, but most do. The principles by which they usually decide that are here:
https://en.bitcoin.it/wiki/Transaction_fees
...although it's entirely up to them.

The thing about which block gets accepted doesn't happen that often, but occasionally your transaction may get into a block, but then that block doesn't make it into the main chain because somebody else broadcasts one first (or some weird network split thing happened). You can see some of those blocks here:
http://blockchain.info/orphaned-blocks
Pages: [1]
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!