W3C home > Mailing lists > Public > public-ldp-wg@w3.org > November 2014

Re: Recharter scope

From: Arnaud Le Hors <lehors@us.ibm.com>
Date: Mon, 17 Nov 2014 15:44:50 -0800
To: Robert Sanderson <azaroth42@gmail.com>
Cc: "public-ldp-wg@w3.org Working Group" <public-ldp-wg@w3.org>
Message-ID: <OF0F99840F.4028E6E3-ON86257D93.00807CB5-88257D93.00827354@us.ibm.com>
Thanks Rob for your feedback. Let me try and clarify a few points.
--
Arnaud  Le Hors - Senior Technical Staff Member, Open Web Standards - IBM 
Software Group


Robert Sanderson <azaroth42@gmail.com> wrote on 11/17/2014 12:44:06 PM:

> From: Robert Sanderson <azaroth42@gmail.com>
> To: Arnaud Le Hors/Cupertino/IBM@IBMUS
> Cc: "public-ldp-wg@w3.org Working Group" <public-ldp-wg@w3.org>
> Date: 11/17/2014 12:45 PM
> Subject: Re: Recharter scope
> 
> Regrets for the next three weeks (Vacation, Meeting, Conference), so
> comments on the items inline below:
> 
> On Mon, Nov 17, 2014 at 12:24 PM, Arnaud Le Hors <lehors@us.ibm.com> 
wrote:
> 1.        How can retrieval of a container and its contained 
> resources be combined so that fewer HTTP operations are required 
> than it is necessary with LDP 1.0? 
> 
> +1. Though I'm very nervous about making everything a query: there 
> has been a long and rich history of failed interoperable query 
> syntax efforts and I have little reason to believe that LDP is the 
> group that can turn that around (no offense intended).

The intent is merely to do "inlining". Ashok described it well, although 
the details of how it's done could be different:
> If you do a GET on 
> a container you get the URLs of its members.  If you use a header, 
> say INLINE, you get the contents of the members (inlined) as well.

The key is that this alone doesn't require a query language per se. It's 
merely an optimization, allowing the server to return more data to the 
client without requiring additional round trips.

>  
> 2.        How can multiple resources be created with a single HTTP 
request?
> 
> +0.  This seems like going firmly against the current specification 
> that the entity body is a single resource, even if it refers to 
> other non-existing resources on the same server, either by absolute 
> or relative URIs.
>  
> 3.        How can a client request filtering what part of a resource
> or container the server is to return? 
> 
> +1.  And to keeping it separate from 1 and not using a query syntax
>  
> 4.        How can a client be notified when something, such as a 
> resource or set of resources, changes? 
> 
> +1. And to working with SocialWeb to see if ActivityStreams will 
> solve or already solves the problem for us.
>  
> 5.        How can a client find out whether a service endpoint, such
> as a SPARQL endpoint, is associated with a resource or set of resources? 

> 
> +0 I don't understand the question or need.  
> 
> Is it that you want to search the objects in a particular container,
> and you need to know where the search service lives?

Yes, although it wouldn't be limited to containers.

> Seems like a 
> property of the container. How else would you know where the 
> endpoint is?

That's the question! :-) If you generalize to all resources - not just 
containers - one possibility would be to have a header on every resource 
that points to the corresponding endpoint but it is not practical to 
expect a client to check every resource to find out where the governing 
SPARQL endpoint is. You really want to be able to find this out for a set 
of resources, say anything within http://example.org/mydata/*.

> If you a priori know the end point, I'm not sure that a
> machine would ever discover it and be able to follow a link to find 
> the resource set and know that the user is interested in that 
> resource set and know how to use the endpoint.  So ... yeah, I don't
> understand this one :)

I hope that clarifies it.

>  
> 6.        How can access to a resource be controlled? 
> 
> +1. This is Stanford's top priority in the list.
>  
> 7.        How can a client have greater control of how paging is 
> done (size, sorting, etc.)?
> 
> +0.  Not of interest, but doesn't seem harmful.
>  
>  8.        How can a client learn what property constraints there 
> are when creating or updating a resource?
> 
> +0 Not of interest, but doesn't seem harmful.
>  
> 
>  9.        How can we do efficient transfer of LDP resources, either
> some initial set or rolling updates (feed) of changes? 
> 
> -1 if this is "How can I get a dump of your database and then stay 
> perfectly synchronized with it on the order of 1 second latency" I'd
> like to see two independent adopters of LDP that have this need 
> other than between internal systems, where an internal solution is 
> likely to be the best one.
> Or +1 if it's just a property of a container that points to a .zip 
> file of the container's members (e.g. solution is trivial, move on)
> Either way, I would drop this from the list of issues.

This is related to 4. How can a client be notified when something, such as 
a resource or set of resources, changes?
When a client is notified a resource has changed, how can it find out what 
has changed without GETing the whole resource again and doing a local 
diff.
I haven't heard anyone talk about setting expectations on response time or 
anything like that. Just that the client be provided a way to get 
incremental updates rather than having complete download every time 
something changes.

> 
> Hope that helps!

Yes, if you see a way to edit the questions so that they effectively 
convey what is meant, please, let me know.
Thanks.

> 
> Rob
> 
> -- 
> Rob Sanderson
> Technology Collaboration Facilitator
> Digital Library Systems and Services
> Stanford, CA 94305
Received on Monday, 17 November 2014 23:45:23 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:11:59 UTC