Be happy!
Matthew Ernst: content:encoded makes me sad:
People that include the full text in their RSS 2.0 feed instead of just an automatically truncated “excerptoid” make me happy. However, including the full text in a <content:encoded> while still keeping the exceptoid in the <description> makes me sad. I’m not even sure what the motivation is to having <content:encoded>, since as far as I know you can put that sort of thing in an RSS 2.0 <description> just fine.
You certainly can put anything you want in an RSS 2.0 <description>. That doesn’t mean you should. I think it’s a great thing that so many people are reading blogs in RSS aggregators these days, and as a result agitating to have more people do full-content feeds.
However, we need to not lose sight of the fact that reading in an aggregator isn’t the only use for RSS. My blog is syndicated (in the original sense of Really Simple Syndication, publishing on another website) in a couple of places that don’t really make use of a hand-crafted excerpt, but if you look at Henrik Gemal’s Mozilla-related blogs page, and hover over the links, you’ll see that some people have short ugly chopped off posts from Movable Type’s auto-excerpts, and some people have slightly longer ugly chopped off posts from Gemal’s own auto-excerpting, but a few people have a perfectly sensible, readable and understandable excerpt, because they put one in for things that syndicate, along with <content:encoded> for people who are using RSS for reading.
It’s a marvelous combination, having two elements for each of two different uses (three, if you count people who have intentionally built their aggregator to only show them a short description because they would rather read posts in the original web page), and I only hope that Radio includes Matthew’s code or something like it (just like it did with his code that now allows him to talk about those tags with HTML entities without double-decoding them and interpreting them) so that Radio users can enjoy full content in their aggregator without forcing people who are also providing their feed to syndicators to not provide them what they want.
Virtually all of RSS 2.0’s technical shortcomings are actually technical shortcomings of Radio (and Manila). No room to put both summary and full content in your blog, no interface to choose between them in your aggregator, thus the underlying format doesn’t need two fields. So we get 3 different undocumented or semi-documented ways to do it, and every now and then people yell at you for trying to innovate. Damn it, people, if Radio doesn’t need it, it should be obvious that nobody needs it! Stop innovating! Leave RSS alone!
Oh, and switch to Atom, which has separate summary and content elements for exactly this purpose.
The decision not to offer both full text and summary fields in Radio is not a ”technical shortcoming” — it’s a design choice. UserLand wanted software that could be used by a novice to create a first weblog post in five minutes.
When you consider that, the decision not to use a form as complicated as Movable Type’s entry/extended entry/excerpt form makes sense. If anything, RSS 2.0 influenced Radio, not the other way around.
I wish you could tout Atom without making these kinds of false claims about Radio or UserLand. You’re not a Radio user. You’re not a UserTalk programmer. And it shows every time you talk about them.
While my memory for dates isn’t even what it once was, my ability to search my archives is still there, and it looks like I was using Radio 8 in essentially its current form in January 2002, and arguing RSS 2.0 in September 2002, so I wouldn’t guess that it influenced Radio. RSS 0.92, maybe, dunno about that.
My vague impression of Radio’s history is that Radio 7 was some sort of MP3 blogging or serving or both thing, and that was where the ”Radio” came from, and then between 7 and 8 for a while it was going to be MUOD, a republisher of RSS feeds, and then it went from there to what it is now. So I would certainly think that RSS-as-it-was would have had quite an influence on Radio during that time, but I’m not sure what that should mean.
Phil, the aggregator in Radio today was initially developed in 1999, as My.UserLand.Com when the current version of RSS was either 0.90 or 0.91 (I could go look it up but it doesn’t really matter).
The following things developed in parallel:
1. Blogging tools.
2. Aggregators.
3. RSS.
4. Weblogs.
All were necessary to make each other.
Trying to divide it up, as Mark did, is pointless.
My policy for this episode of Syndication Wars is to assume that every time someone says something wildly extreme, that it’s an attempt to play devil’s advocate, and force people to more carefully examine a reasonable position. I hope this policy works out better than whatever I was using the last time around, since I was just rereading my archives from back then, and knowing how to translate how I write to how I’m actually feeling, I seem to have been furious at best.
Before I got tired of hearing myself type, and realized I didn’t actually know much about that period of history, ”it’s all muddled together” was where I was going. scriptingNews was (obviously) designed for what you were doing with your blogging tools, RSS moved that way for whatever reason (finger in ears, shut up shut up shut up ;)), when you built tools to work with RSS they worked with what RSS was like, when you expanded RSS you expanded it in ways that you knew tools would use, to me that seems like what you would expect someone to do. Now, if there were tools and users of those tools clamoring for separate excerpts and content, and you refused because your tool couldn’t be made to produce them or consume them, or even because you didn’t want people doing things that way, then you’re a bad bad man who frustrated me before I even knew I was going to want something. But my impression is more that you just froze it (for your reasons, and yes I remember) before I was quite done poking at it.
Eh, life goes on. My aggregator keeps dinging at me, telling me I’ve got 68 unread items, in whatever format, and I’ll read them no matter how much does or doesn’t come through.
Sorry about the lack of clarity, Phil. I was using ”RSS 2.0” as shorthand for ”the branch of RSS that’s now known as RSS 2.0.” The title-link-description model in that format inspired a lot of weblogging tools to offer a means of authoring a title, link, and description for entries. Who coulda thunk it?
If Atom causes people to be much more fine-grained in how they author feeds, putting the full plain text in one element, the encoded HTML text in another element, a plain text excerpt in another element, and so on, great. But when John Q. Weblogger is being asked to do this much work to author a syndication feed, I’m glad RSS 2.0 will still be around for people who don’t have the desire to work as hard as Mark Pilgrim.
I’m persuadable, don’t think I’m not, but I’m still not persuaded. 0.90 was just T/L, no D. From everything I’ve read, if there even is a truth to how 0.91 got T/L/D, it’s now unknowable: there’s only various people’s truths about it, no one truth. L as weblog permalink doesn’t exactly fit well in a ”RSS produced blogging as we know it” scenario, since I haven’t heard much argument against the idea that pb ”invented” permalinks for Blogger, which was quite RSS free for a very long time, and still has a certain ambivalence about whether L is permalink or is a link to the subject of the post. I’d be more inclined to say that RSS, designed for syndication on portals, turned out to fit well with how blogging was done at the time, by a happy accident, they influenced each other to some extent, and now RSS-as-it-is doesn’t quite fit as well with the way that some people want to do their blogging. That’s not a blinding truth, or a very good battle cry, but it does seem to fit with the bits of history that made it into public view, lasted long enough, and I’ve run across.
The memo on useful reasons to include both an description and a content:encoded should probably also be circulated to those producing feeds with both a description and a content:encoded. The combination of crafted excerpt in description and full text in content:encoded is so excedingly rare (I believe this site’s rss2.0 is the only one that I am subscribed to that does it, and I only subscribed to it today) that it never occured to me that the description would be used like that ”in the wild”. You’ll note that I has specifically mentioned that I was seeing the ”excerptoid” in the description… auto-generated from the full text that is included in the very same feed. There’s no information in such a description, so you can see why I was interested in moving it out of my way.
In response to the comment that reading in an aggregator is not the only use for RSS (but not necessarily supportive of my first paragraph down here), surely RSS produced spcifically for a blog should *at least* be useful for reading in a news aggregator. Mark’s theory about the shortcomings of Radio needs to be weighed against my experience that there is next to no demand in even the non-Radio audience.
In the end though I am happy because I made the damn machine do what I wanted. ;-) Although, I wonder if I could determine if the description is an auto-excerptoid, so that I could avoid dropping that one crafted excerpt…
Blosxom/Atom Entry Summaries
Extending Blosxom’s Atom support to include handcrafted excerpts.
Blosxom excerpts
My weblog software’s design is based on blosxom’s, with entries stored in the familiar title followed by body format. On my longer entries, such as this one, the first element in the body is a div tag with a class of excerpt, something that I parse ou