> Freenet isn't any good if you dont have an entry point. Which is why people will have the incentive to go and find a local entry point (or create one of their own)! If we provide a fallback mechanism for those too lazy to do this, then we are shooting ourselves (and Freenet) in the foot. Did email fail to catch on because people need to find a mail server? Did DNS fail to catch on because people needed to find a DNS server? I think the whole news server in a URL was a bad idea from the start, and it has put nasty ideas into people's heads about Freenet.
> Your key is > basically useless unless the client can find a server that is close enough > to the data to retrieve it. Encoding a node's information in there helps > the client find data. The only difference between this and > http is that ours is a suggestion, not a demand. Most of the time users > wont make suggestions. Sometimes they will have to in order to get what > they want. Actually having this fallback mechanism will encourage people to be lazy and not to bother finding a local Freenet node. I really don't want to do this. > Hopefully this will eventually become obsolete. But for now I think there > needs to be a mechanism to suggest where to enter freenet. Designing something *in a protocol* which is expected to become obsolete frequently indicates that you are making a mistake. We didn't create Freenet so that people could use it as if it was HTTP (even if only as a fall-back). If someone wants to provide a fallback then they can have two links for the data, one which gets it from Freenet, and another which downloads it via FTP (much as you often see an option to get something via HTTP or FTP). Ian. _______________________________________________ Freenet-dev mailing list Freenet-dev at lists.sourceforge.net http://lists.sourceforge.net/mailman/listinfo/freenet-dev
