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

Wishlist Tally, commentary

From: Sandro Hawke <sandro@w3.org>
Date: Mon, 21 Apr 2014 08:51:11 -0400
Message-ID: <5355143F.3030606@w3.org>
To: Linked Data Platform WG <public-ldp-wg@w3.org>
I tallied up the votes, and here they are sorted:

9.5 PATCH
9.5 Efficient delivery of changes
9.3 Access Control
8.5 Linked-Data Authentication
8.0 Filtering collections
8.0 Embedded representations (inlining)
7.0 Client-Controlled Container Filtering
6.5 Push
6.5 Client-Controlled Inlining
4.7 Containers as SPARQL endpoint
4.7 Connection to SPARQL datasets
3.5 Client-Controlled Page Sorting
2.6 Snapshots
2.6 Provenance
2.2 Client-Controlled Page size
2.1 Post Multiple
2.0 PATCH Multiple
2.0 Datasets
2.0 Container-Inclusion
1.2 Richer Containers
1.2 Guidance for PUT-for-Create
1.2 Client-Initiated paging
1.2 Back Links
-0.5 Crowd CDN

It's nice that the closely related items got closely related votes.

I edited the minutes to make them link to the wishlist as it stood at 
the time of the meeting, so we can keep editing and re-arranging it.

I'm trying to understand some of the low votes....   Here's some 
commentary and questions:


2.6 Snapshots

Not sure if people don't see why clients would want this, or think it's 
too hard to provide.

2.6 Provenance

To me, exposing provenance is the path the data quality.     The 
downvotes might be due to confusion about what's still needed here, 
given the existence of PROV.

2.2 Client-Controlled Page size
1.2 Client-Initiated paging

Interesting, given the next day we resolved to include these in the spec 
right now.  :-)
https://www.w3.org/2013/meeting/ldp/2014-04-17#resolution_4

2.1 Post Multiple
2.0 PATCH Multiple

*shrug*

2.0 Datasets

I think I didn't explain this well.   It's necessary for the SPARQL work 
which was better received.

2.0 Container-Inclusion

I'm kind of shocked at this low vote.  I can't imagine folks working 
with containers for very long without wanting divisions within 
containers.   I guess you can do it with filtering (if we had 
filtering), but to me it seems very natural to use sub-containers. I 
certainly do it all the time in the filesystem, putting things in 
sub-directories, even though they are all the same basic type, and then 
recursively accessing all the things in every subdirectory.

1.2 Guidance for PUT-for-Create

Interesting vote for a feature that TimBL seems to consider mandatory.   
But that was before we really had POST-for-create, so maybe he'll be 
okay with changing.

1.2 Back Links
-0.5 Crowd CDN

Complex concepts, for other groups and/or research, sure.
Received on Monday, 21 April 2014 12:51:19 UTC

This archive was generated by hypermail 2.3.1 : Monday, 21 April 2014 12:51:19 UTC