1500 inputs in a single TX possible?

[shorena]

Hi,

recently someone posted in the german section about problems with bitcoin-core. We figured out that the problem is the 1500 unspend outputs from faucets transactions. They have a total value of ~0.25 BTC and bitcoin-core suggestes a fee of ~0.00232 BTC for the 232000 Byte size TX. I have never seen a single TX this big. Would that even be possible? Would miners accept a TX that big?

The other idea that came up is to make a raw TX without fee in order to fuse 3-4 unspend outputs in a feeless TX. This however will take some time. Any idea how long? How many inputs should be use per transaction?

Please spare me the "see faucets suck" post, I know that allready. And while this is a very good example why, its not the point of the thread.

Any other ideas to resolve this?

[Newar]

Size wise, for example:
http://btc.blockr.io/block/info/256961
884.45 kB

I don't think there are much more options other than to pay that 1% fee or wait (A year or so? Don't know.) until they're more mature.

[amaclin]

Would that even be possible? Would miners accept a TX that big?

Why not?
Miners are interested in fees per kilobyte.
And of course reducing the number of UTXO in blockchain is God's work

the most outputs number transaction is e951e79d2140b1844d5058bd0f609e3939b040132a1d920ec89c55a2392ef83f up to date
it has 5353 outputs.

sorry, i do not have information what is the maximum number of inputs now

hm.... what about the number of OP_CHECKSIG in block? you should not go upper the limits

EDIT:
It is better to create txs less than 100 kb. I do not remember right now where i got this number

[shorena]

Size wise, for example:
http://btc.blockr.io/block/info/256961
884.45 kB

I don't think there are much more options other than to pay that 1% fee or wait (A year or so? Don't know.) until they're more mature.

Big blocks yes, but big TX? 884KByte block with 3861 TX is still ~ 230 Byte per TX on average.
Bc.i lists "largest TX" [1] but they refer to value not size in bytes.

Would that even be possible? Would miners accept a TX that big?

Why not?
Miners are interested in fees per kilobyte.
And of course reducing the number of UTXO in blockchain is God's work

That is what I think as well. I have never encountered anything like that though, thats why I was asking for more inputs (no pun intended).


[1] https://blockchain.info/largest-recent-transactions

[zetaray]

If bitcoin core can create a tx that big, would miners include it in a block? If you pay enough fees (0.0001btc/k), I can't see why miners would reject it.

On the save side, split the tx into a few smaller ones to test.

[Newar]

Size wise, for example:
http://btc.blockr.io/block/info/256961
884.45 kB

I don't think there are much more options other than to pay that 1% fee or wait (A year or so? Don't know.) until they're more mature.

Big blocks yes, but big TX? 884KByte block with 3861 TX is still ~ 230 Byte per TX on average.
Bc.i lists "largest TX" [1] but they refer to value not size in bytes.
[...]

Right, how about this one: https://insight.bitpay.com/tx/e951e79d2140b1844d5058bd0f609e3939b040132a1d920ec89c55a2392ef83f

182160 bytes, close, but no cigar. 

[shorena]

Size wise, for example:
http://btc.blockr.io/block/info/256961
884.45 kB

I don't think there are much more options other than to pay that 1% fee or wait (A year or so? Don't know.) until they're more mature.

Big blocks yes, but big TX? 884KByte block with 3861 TX is still ~ 230 Byte per TX on average.
Bc.i lists "largest TX" [1] but they refer to value not size in bytes.
[...]

Right, how about this one: https://insight.bitpay.com/tx/e951e79d2140b1844d5058bd0f609e3939b040132a1d920ec89c55a2392ef83f

182160 bytes, close, but no cigar. 

Nice find, >5k outputs, no fee paid.

Looks like the address belongs to a big pool https://bitcointalk.org/index.php?topic=744097