On Fri, May 12, 2000 at 01:07:37AM -0500, Scott G. Miller wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > > Why is there a need for a different, binary protocol in a test bed? Are > > you referring to a simulation? If so, you don't need to worry about the > > protocol spec, just pass the objects around. > > > > And if you are talking about protocol translation, yes, that's a gateway. > > Hmmm.. Sorry Lee, you've lost me again. For this type of application, say > a Freenet TCP/IP<->Freenet Email or something silly like that, it would be > really really nice. Its unreasonable to expect a utility program such as > a gateway to have a rigorous understanding of the protocol in order to > merely translate messages. > > Scott
I'm not Lee :) That's what gateways do. They translate protocols, allowing transfer of content between networks that don't understand each other. If you send a freenet message over email, you are 'encapsulating' the freenet message in another transport protocol, not translating. This is often referred to as tunneling. A gateway would try to change a Freenet message into SMTP, which wouldn't be very usefull. So, if there are two Freenet message protocols, you need a gateway. You need a way to translate Freenet/binary into Freenet/text. The point being, the Freenet protocol is supposed to be 'neutral territory', it should not be making assumptions about the charactersitics of end systems, or what medium is carrying the message. Internet protocols are about interoperability, and not making assumptions about endpoints or transmission mediums. Which is why I was talking about the difference between a distributed Java app and an Internet protocol. If you're not concerned with interoperability, you can use native Java serialization and carry the result over any transport that can handle binary data. Interoperability problems are already showing up with the hex number type, and the way different programming environments represent hex numbers. Upper case, lower case, leading zero, and this is just a little taste of what's to come, and why things like ASN.1 are so huge and complicated. Nothing wrong with building a distributed Java app, but I think the project will go faster and further if a decision is made on which approach to take, it would keep people from working at cross purposes. David Schutt David Schutt _______________________________________________ Freenet-dev mailing list Freenet-dev at lists.sourceforge.net http://lists.sourceforge.net/mailman/listinfo/freenet-dev
