Namecoin Forum

I2P lacks advanced DNS support from the beginning. There are addressbooks. There is a private addressbook for every user and there are public addressbooks that are being consulted by other users. Different users might use different set of public addressbooks. For example, inr.i2p is rather popular (i2p.to gateway respects inr.i2p entries), but it's not enabled by default. Every I2P user has its own view of I2P web. It is not guaranteed that everybody will be able to follow any particular i2p link (except for .b32.i2p links). Address-helpers also help find i2p destination, but it is possible to inject another i2p destination into private addressbook this way. I2P is in ancient ages fully utilizing hosts.txt

Another problem is non-eepsite services. For instance, there is no way to specify MX entries. There are also no hosts and no ports in I2P. There are anonymous destinations. hosts.txt doesn't specify hosts. It lists name-to-destination mapping. We can't launch another service on another port because any particular i2p name refers to just one destination, so they are mostly eepsites. In fact, there is just 1 mailserver (mail.i2p), and there is no way to launch another one linked to mail.i2p. Well, we can agree to refer to smtp._tcp.mail.i2p to find @mail.i2p's I2P SMTP server, but a problem still remains. With regards to email and xmpp, unsynchronized view of I2P namespace is not acceptable. Centralized view isn't highly desired as well.

NameCoin feels like a breath of fresh air. I2P users are potential NameCoin users. NameCoin records are in JSON and it seems that one can put I2P destinations instead of IP addresses into DNS zone records.

http://forum.i2p/viewtopic.php?p=35652
http://forum.i2p2.de/viewtopic.php?p=35652

From my point of view there is no problem to add such addresses to the blockchain. It should IMHO be done via a proposal if you want ot use a separate namespace. But I guess it would even be possible to add it to the .bit blockchain but I guess it won't be desirable as i2p seems to be another TLD. On the other hand I'm asking myself if this address book style was used intentionally at first and that it has some reason that there is no global name resolution?

There is already a first work on supporting i2p with .bit : http://dot-bit.org/Domain_names#Value_field
It can't be used with a DNS server, but it can be used with a proxy software (same for tor and freenet). Existing Vinced proxy does not support thoses feature yet (http://dot-bit.org/HowToBrowseBitDomain ... plications).

This is what im talking about

I implemented some basic beta support for .bit domains in i2p

http://i2psupport.org/

Proposal posted to i2p developers

http://zzz.i2p/topics/1091?page=1#p4333

pent wrote:This is what im talking about

I implemented some basic beta support for .bit domains in i2p

http://i2psupport.org/

This comic is heroic.

strange thing is the dianna people never made a valid argument against namecoin...