Rogers Cadenhead: My weblog’s Atom 1.0 feed is now
online. As an experiment, I’m routing all RSS requests to
this feed. I’m curious about whether aggregators can
handle that. The RSS 2.0 code’s still around so I can offer
both, but I’d prefer to offer a feed in one format so I avoid
the need to debug two.
I’m following Rogers' lead, and have done the same with
my remaining feeds.
If you could use absolute URI’s in your feed, that would sure help a lot (I use Sage for Firefox, which creates a local file on my disk; relative URI’s, therefore, are relative to my C: drive :/ ).
I have to second the request for absolute URIs. Newsgator Inbox (integrated with Outlook) doesn’t seem to like your relative URIs. I’ll put in a bug to the Newsgator folks, but maybe right now it would be better to use the absolute version.
I would go complain to the Sage and Newsgator developers that their Atom implementations aren’t up to spec. Atom explicitly allows for the use of relative URI’s. It’s the feed readers responsibility to handle them properly.
Did you just recently change to relative URIs? NetNewsWire doesn’t like them, but they seemed to work when you broke the internets the other day (and I was using Atom then).
I assume NNW is choking on the link element <link href="/blog/"/> because it’s trying to make links /blog/blog/2006...
Funny story, vaguely related: I was at FOO Camp last fall, and struck up a conversation with someone who (1) was familiar with the Syndication Wars, and (2) had been a regular reader of my blog in its day. He asked if there were any circumstances under which I would resume blogging. I replied — I am not making this up — “when hell freezes over, or Cadenhead switches to Atom.”
It’s an issue with Sam not specifying xml:base mostly. Most feed readers don’t correctly use the feed’s url as the base value when xml:base is omitted. NNW included. Out of curiosity, which feed readers actually do support relative urls when xml:base is left out? I know FeedTools and the UFP do, but most of the others I tried got confused in one way or another.
Bob Aman: which feed readers actually do support relative urls when xml:base is left out?
The ones that worked in my tests: JetBrains Omea, RSSOwl, Snarfer, and Thunderbird (out of 16). Of those, only Snarfer and Thunderbird got it right if the feed redirected. And if the feed has a Content-Location header, only Snarfer got it right.
I suppose I should add that JetBrains Omea and RSSOwl still won’t work with Sam’s feed because they don’t seem to support relative URIs in link elements.
As a general rule, I have no sympathy for developers who can’t be bothered to write code I’ve already written. It’s not like there aren’t lots of conformance tests. Lots of people have written code that passes them. I say Sam should keep the relative links. But then, I’m an asshole. I also supported Drunkenbatman posting an image on his home page that crashed Safari. (DO NOT CLICK THAT LINK IN SAFARI.) So in the grand scheme of things, posting some spec-compliant-but-practically-broken links isn’t really that bad.
<grumble>... I need to hurry up and get SporkFed into a usable state then, if we’re going to be “assholes” and keep omitting xml:base. This NNW brokenness business is going to get annoying fast. It’s kinda annoying that I’ve written this library that works well but since it’s still just a library, I don’t really actually benefit from it yet.
I was trying to figure out if I’d introduced a bug in RSS Bandit to make your comment count and inline comments disappear. Instead, it seems you have made your feed less useful as part of the fallout of yet another iteration of the eternal pissing match which is the XML syndication wars.
If you were really hard core, you would check out the API Manager plugin from subversion at http://code.sixapart.com/, install it Movable Type and presto, get Thread support just like that.
James, so far, I have only modest beginnings, each entry in index.atom contains a link with rel="replies" specifying the Atom feed for that entry, the current number of comments, and the datetime of the latest comment on that entry (if any).
That’s great Sam. I just checked your main comment feed and figured you hadn’t done anything yet. Maybe now that you’re supporting it I might be more inspired to finish implementing it in my aggregator.
A little later I’ll have a post indirectly related to my current work effort. Some has to be kept confidential but part of what we’re developing is work that will be released to the world at large, and that’s what I’ll be writing about. In the...
in the comments, Mark Pilgrim said: ‘He asked if there were any circumstances under which I would resume blogging. I replied — I am not making this up — “when hell freezes over, or Cadenhead switches to Atom.”’ (my thought: :-) :-))...
I’d love to be able to say yes, but it seems unlikely. The last project I worked on was carefully designed from the outset to be easily portable. Most of the base code was compilable with gcc under Linux and the GUI stuff was neatly encapsulated. I had fully intended that we port to Mac OS, OS/2, and any number of variations of Unix/Linux when we got the chance. The project turned out to be highly successful and eventually sold out for a tidy sum, but you know how many ports we did in the years we were developing it? Zip. Zero. Nil. Nada.
Tragic really, but that’s the reality of the situation. You never know what might happen, but I’m not going to hold my breath this time around.
As Josh points out, a US congressman showed up at our last meetup. Which local blognitary could it be this time? John Edwards? Maybe Mark Pilgrim will appear to announce his return to blogging. Or maybe it’ll be Dr. Jim Goodnight, who’ll ask how to...
Tim Bishop at Geodog spotted the new post at Mark Pilgrim’s weblog. It’s all about bathin' the baby. I personally liked the Beta patch in the corner--very 2.0. Whether Mark is continuing the weblog is hard to say. According to Mark in discussion at...
First, I verified that the RSS 0.91 spec is more honored in the breach than the observance. Then I switched exclusively to Atom 1.0. Time to destroy the internets again. But only for those Aggregators with an 'X" in column four in this table. ...
[more]
First, Mark Pilgrim is blogging again. He said he’d return when hell froze over or Rogers switched to Atom. Fortunately for is, hell froze over (i.e. Bootcamp) and Rogers made the switch.Second, Microsoft has kindly answered my questions about the...
Theory: Perhaps it is because your server is returning a 302 a.k.a. a temporary redirect. This means that consumers will continue to fetch the original address, and then follow the redirect. Each time. Despite having coded a permanent redirect over...
[more]
410 Just over a year ago, I permanently redirected all my feeds their Atom 1.0 equivalents. Several months later, I quietly converted all those redirects to 410 Gone. Checking back to see how effective this has been, here is a list of...
[more]
Sam Ruby has been responding to RSS requests with permanent redirects and gone statuses for a year . It appears nobody is listening. My Rmail code worked. Hundreds of Google Reader users are being denied Sam’s wit. Anybody working at Google...