Namecoin Forum

latest discussion: http://dot-bit.org/forum/viewtopic.php?f=5&t=1163

Current size is 520 bytes due to a bug, should have been 1k

According to virtual_master the maximum expected key size is 2k

Note that the size is NOT directly related to the size of the blockchain as it is / will be cheaper to create two small txs than to create one large one (though it might make spamming slightly faster if you only run a single instance). The main thing this will do is allow for easier handling of large keys, many signatures, etc.. I am so sick of this discussion that if you post in this thread that a large value field will bloat the blockchain I will delete your post :p

http://dot-bit.org/forum/viewtopic.php? ... 3&start=40
This was the latest thread where we agreed on 5k to enable the maximal sized GPG key which is 4k.

phelix wrote: Consensus on size is still 5k.

virtual_master wrote:http://dot-bit.org/forum/viewtopic.php? ... 3&start=40
This was the latest thread where we agreed on 5k to enable the maximal sized GPG key which is 4k.

phelix wrote: Consensus on size is still 5k.

I agree with our previous consensus :p

/voted

It's pretty easy to fit a decent resolution picture of a child into 7K. Moving from 1K to 5K makes that 3.5x cheaper.

indolering wrote:It's pretty easy to fit a decent resolution picture of a child into 7K. Moving from 1K to 5K makes that 3.5x cheaper.

Agreed, I have no problem with a larger size, but I think we should implement encryption before touching the value size. ryan-c is making progress on encryption if I understand correctly.

(And yes, I know it's theoretically possible to fit it into multiple names... but some people aren't that motivated, and perception/intent matters for both PR and legal purposes.)

indolering wrote:It's pretty easy to fit a decent resolution picture of a child into 7K. Moving from 1K to 5K makes that 3.5x cheaper.

biolizard89 wrote:(And yes, I know it's theoretically possible to fit it into multiple names... but some people aren't that motivated, and perception/intent matters for both PR and legal purposes.)

We had evoked exponential costs for big data. Isn't it sufficient for the perception ?

khal wrote:

biolizard89 wrote:(And yes, I know it's theoretically possible to fit it into multiple names... but some people aren't that motivated, and perception/intent matters for both PR and legal purposes.)

We had evoked exponential costs for big data. Isn't it sufficient for the perception ?

Reality and perception need to align and an easily accessible image of a child being raped is bad, even if it would be easier and cheaper to store said picture on Freenet.

I'm being serious here: we shouldn't increase the value size past 2K unless we can clamp down on the content being stored. Correct me if I am wrong, but since a key is already just highly compressed data we can't do anything to restrict the content. If someone needs a keysize past 2K they should link to it and stick a hash signature in the value field. Everyone is already switching to ECC which will never need to get above .5K.

I'm also not worried about increasing the size in the future, after we have established credibility in the public eye. But if the first news report most people hear about Namecoin is that it's a new distributed datastore for child pronography ... well, consider Freenet vs Tor.

indolering wrote:

khal wrote:

biolizard89 wrote:(And yes, I know it's theoretically possible to fit it into multiple names... but some people aren't that motivated, and perception/intent matters for both PR and legal purposes.)

We had evoked exponential costs for big data. Isn't it sufficient for the perception ?

Reality and perception need to align and an easily accessible image of a child being raped is bad, even if it would be easier and cheaper to store said picture on Freenet.

I'm being serious here: we shouldn't increase the value size past 2K unless we can clamp down on the content being stored. Correct me if I am wrong, but since a key is already just highly compressed data we can't do anything to restrict the content. If someone needs a keysize past 2K they should link to it and stick a hash signature in the value field. Everyone is already switching to ECC which will never need to get above .5K.

I'm also not worried about increasing the size in the future, after we have established credibility in the public eye. But if the first news report most people hear about Namecoin is that it's a new distributed datastore for child pronography ... well, consider Freenet vs Tor.

I agree here - personally I'm not sure whether we really need such large sizes at all (just store key fingerprint), but if we change it, we should first implement value encryption. But I think this is already the plan, right?

4k file size is very dangerous. I don't understand how they came to the evil idea in 1980 to develop FAT12.
In a porno video of this size 2 pixels could move around and that could be interpreted as a small child.

What about a compromise of 2.5k? That way we could still fit in a largish key with some additional data. It should be sufficient for most applications so the benefit from anything more is small.