James M Snell wrote :
Thomas Broyer wrote:
How would you use these link relations for feed state reconstruction
(that is, automatic handling by the Atom processor, without user
action –except probably the "please reconstruct this feed's state"
action) if you can't know what's pointed at?
How would you navigate through a paged search result (e.g. OpenSearch
result feed) if previous/next might point at previous/next queries
(e.g. history of what people searched for before you) or
previous/next pages of the result feed?
How would you disambiguate between two "next" link on pointing at the
next chunk of a paged feed and the other pointing at the next
archived feed? (e.g. "August Top 100, page 2" and "September Top 100")
I thought we all agreed that we need a specific definition (related
to paging) of these relations and were previously voting about how to
call them (previous or prev-archive)
I wouldn't disambiguate them as I don't believe it's necessary to.
What's needed is a way of identifying the role|purpose|intent of a
feed; not a way of disambiguating the links between them. An
"archive" feed should be treated differently than a "search" feed
which should be treated differently than a "subscription" feed.
So you are OK with these feeds:
GET /top50.atom HTTP/1.0
Host: music.example.net
200 OK
Content-Type: application/atom+xml
Content-Location: /top50/2005/w42/1.atom
<feed xmlns="http://www.w3.org/2005/Atom";>
<title>Week #42 Top 50</title>
<subtitle>Top 50 best selling singles</subtitle>
<link rel="self" href="http://music.example.net/top50/2005/w42/1.atom"; />
<link rel="subscribe" href="http://music.example.net/top50.atom"; />
<link rel="previous" title="Week #41 Top 50"
href="http://music.example.net/top50/2005/w41/1.atom"; />
<link rel="next" title="Week #42 Top 50 (11th to 21st)"
href="http://music.example.net/top50/2005/w42/2.atom"; />
<fh:incremental xmlns:fh="…">false</fh:incremental>
…
GET /top50/2005/w42/2.atom HTTP/1.0
Host: music.example.net
200 OK
Content-Type: application/atom+xml
<feed xmlns="http://www.w3.org/2005/Atom";>
<title>Week #42 Top 50 (11th to 20st)</title>
<subtitle>Top 50 best selling singles</subtitle>
<link rel="self" href="http://music.example.net/top50/2005/w42/2.atom"; />
<link rel="subscribe" href="http://music.example.net/top50.atom"; />
<link rel="previous" title="Week #41 Top 50"
href="http://music.example.net/top50/2005/w41/1.atom"; />
<link rel="previous" title="Week #42 Top 50 (1st to 10th)"
href="http://music.example.net/top50/2005/w42/1.atom"; />
<link rel="next" title="Week #42 Top 50 (21st to 31st)"
href="http://music.example.net/top50/2005/w42/3.atom"; />
<fh:incremental xmlns:fh="…">false</fh:incremental>
…
GET /top50/2005/w41/1.atom HTTP/1.0
Host: music.example.net
200 OK
Content-Type: application/atom+xml
<feed xmlns="http://www.w3.org/2005/Atom";>
<title>Week #41 Top 50</title>
<subtitle>Top 50 best selling singles</subtitle>
<link rel="self" href="http://music.example.net/top50/2005/w41/1.atom"; />
<link rel="subscribe" href="http://music.example.net/top50.atom"; />
<link rel="previous" title="Week #40 Top 50"
href="http://music.example.net/top50/2005/w40/1.atom"; />
<link rel="next" title="Week #42 Top 50"
href="http://music.example.net/top50/2005/w42/1.atom"; />
<link rel="next" title="Week #41 Top 50 (11th to 21st)"
href="http://music.example.net/top50/2005/w41/2.atom"; />
<fh:incremental xmlns:fh="…">false</fh:incremental>
…
How do you expect a newsreader to *automatically* download this week's
50 entries without downloading last week's entries instead?
(and show you links/buttons for you to ask download and display of
previous/next week's Top 50)
--
Thomas Broyer
