RE: draft-reschke-webdav-search-05 - a few questions on the draft

> > Thanks Julian.
> >
> > Your responses were very helpful. I would be curious to know what 
> > other issues you forsee with multiple scopes.
> >
> > I had one final question on this issue. From what I have 
> heard so far, 
> > it seems like URIs of different schemes (basically ftp, 
> http, etc) are 
> > issues. In which case I am wondering if there is any issue 
> if all the 
> > URIs are of
> a
> > particular scheme lets say http?
> A SEARCH arbiter *usually* will not use HTTP to actually access
> (PROPFIND...) the resources in scope. Instead, it will just 
> go though whatever internal APIs it has to the resources. If 
> there is more than one scope, it's just more likely that it 
> won't be a single API, or that it is API but it behaves 
> differently for the various scopes.

Well, yes and no.  I agree that the server usually won't use HTTP
in real-time to trawl through the files & answer the SEARCH request.
However, it's very possible that in the past the search arbiter
did do so.  That's the case for search engines that use spiders to
index Web or FTP pages on other sites.  So for that kind of search
arbiter, there can be one scope or many, but it *could* be a single
API to look at the index that had been compiled for those scopes.

Most of DASL was designed to accomodate these kinds of Web (or 
non-Web) indexing engines as well as DAV repositories.


Received on Monday, 6 October 2003 13:19:42 UTC