[prev in list] [next in list] [prev in thread] [next in thread] 

List:       mozilla-rdf
Subject:    Re: How does mozilla handle bookmarks HTML -> RDF?
From:       Mike Lee <mike_lee () iprimus ! com ! au>
Date:       2002-11-04 3:58:49
[Download RAW message or body]


We dealing with storage of an RDF here. Not the use of RDF, I am 
personally more disappointed at the fact RDF mail datasource was taken 
out. I can't think of any good reason why you need to store a bookmark 
in RDF. The only possible reason is lack of flexability to define 
complex relations between them. However if you look at RDF carefully 
you'll find that the main 'feature' of RDF is to solve this problem.

That being every tripple has a subject resource which point to the 
subject in question. RDF is a metadata format that annotate remote 
resources, thats the main feature and I'm not sure why people keep 
forgeting. So how do you solve this flexability problem? You create 
another file called metaBookmark.rdf and have it in RDF in all it's 
glory that annotate the bookmark with any additional metadata you want. 
Simple, in fact that is what I'm going to do instead of polluting the 
bookmark itself.

The RDF xml format is for exchange, you don't exchange bookmark.html, 
you use mozilla (with an RDF model) to do that. The mozilla profile 
directory is pretty hidden and noone should ever need to directly access 
the bookmark file (except backup blah blah). I stand my ground, the bug 
should be mark wontfix.

Regards

Danny Ayers wrote:
> Another 2 cents:
> 
> I'm not working on Mozilla, but am lurking because I'm working on an
> application that will use RDF as its primary data model/serialization
> format. I won't bore you with the details, but a large proportion of the
> data the app will be handling will be the same kind of stuff Mozilla deals
> with - web pages, links, bookmarks etc.
> My choice of model & format was really based on six questions, the following
> applied to model & format :
> 
> * can it represent all the different kinds of data adequately?
> * will it be able to exchange data with other applications?
> * is it sufficiently versatile & extendable to cover big sea changes in the
> kind of data/applications on the web?
> 
> RDF came up tops on all aspects, applied to virtually of the different types
> of data I'll be handling. I will still be using HTML for document display
> purposes (for which it is very good!), and probably SVG too - but for things
> like bookmarks HTML is way too restrictive. Ok, it may work for bookmarks
> considered in isolation/viewed through a browser, but to do stuff with those
> bookmarks - e.g. categorise them, HTML is hopeless.
> 
> Bookmarks and suchlike are just the kind of resource that RDF is designed
> for, whether for interchange or storage. The same kind of engine needed to
> work with bookmark information can be used pretty for anything that can be
> expressed through resources, there's no need to make lots of little special
> case hacks. Less work!!
> 
> The above technical points I believe apply to Mozilla, but there is also a
> slightly emotional angle which would make me hope that RDF was adopted
> wherever appropriate. Mozilla has been something of a leading-light project
> in respect of new technologies, and it would be nice for that to continue.
> Using HTML for bookmarks seems to me to be worryingly regressive.
> 
> Cheers,
> Danny.
> 
> 


-- 
Mike Lee
Website: http://www.exitspace.net/mike

["smime.p7s" (application/x-pkcs7-signature)]

[prev in list] [next in list] [prev in thread] [next in thread]