nk4um

Hi Micea,

Notice that styling is generally relatively linked and breaks when we relocate the resources - the designers of most web pages don''t anticipate someone mounting them!

I don''t know who wrote that! Ah actually I do, he sits next to me. ;-)

The problem actually runs deeper. We''ve thought about a solution to this for a long time now but not actually invested time in implementing it. The problem is that the mount needs to deep introspect of the various resources that it replays and depending upon at least the mime type needs to parse and translate relative links. For the web at least these might include links, images, media, css, javascript etc, etc and even then with dynamic code and such it may not be 100%. That said a partial solution would work but I think that this concept at least as far as being a web proxy is just a demo. Probably better to using a proper proxy protocol.

Cheers, Tony

Hi,

The idea to relocate the Web under an arbitrary and presumably uniform generalized addressing interface is cool. Anyway, the Mount Pattern as described in http://localhost:1060/book/view/book:tutorial:basics/doc:tutorial:basics:part3 does an incomplete job. The reason is mentioned on that page:
"Notice that styling is generally relatively linked and breaks when we relocate the resources - the designers of most web pages don''t anticipate someone mounting them!"

Try this out http://localhost:8080/mnt/nk-wide-web/www.netkernel.org and you''ll notice that even the designers of www.netkernel.org did not anticipate someone mounting them!

Now, from a web site designer perspective it is just good practice to use relative paths to refer to the site''s local resources such as style files, images etc.

Would you please provide a "fix" to the Mount Pattern for it to cope transparently with the existing Web? Not having a fix for that, the Mount Pattern as a tool to (re)map the Web is just a toy for philosophers :)

Thanks,
Mircea