1> Web pages are read-only
2> Web pages require a server
3> Web servers can be censored
4> Web links are unidirectional
5> Hypertext Markup Language is really Hypertext Formatting Language
6> IP4 is running out of addresses
7> The internet backbone isn’t redundant, nor is it survivable
8> The last mile is still in the hands of corporations
9> Internet Access, instead of Internet Connectivity
10> Insecure operating systems on the end nodes
These are the challenges/opportunities that we need to meet.
Subscribe to:
Post Comments (Atom)
3 comments:
1) Web pages are read-only
How about wikis? Wikipedia is a fine example.
4> Web links are unidirectional
And how would bidirectional links work?
Wikis are one hack to work around it... and I gotta say that keeping all versions around is very handy... but not being able to just open up a web page and edit it (assuming permissions were appropriate) is a major drag, years later.
Bidirectional links would work similar to trackback, but it would all happen transparently anytime someone used a link. I'd probably hack up a URL on a per server basis to point to a repository... it might even be something like http://servername.com/backlink, with a standardized set of services .
I hear you on this one -- I agree that wikis are a hack (and worse, a huge overlap of features with many other things). I may have stumbled upon an answer...
Perhaps you've seen flickrfs -- a little python FUSE script that mounts your flickr account as a volume? It serves as a pretty good model for a way to make a real 'read/write' web. I'm trying to get some traction on a project called SocialFS, and a lot of this list is fundamental to it (the first half at least)...
I'd love to hear what you think about it (especially since it was a post by you some five years ago that got my wheels spinning on this topic).
Post a Comment