Dave Winer: I read a piece yesterday about SixApart and their standards compliance. Interesting, but they do RSS in a funky way.
I don't get it. I've looked at MovableType's RSS 2.0 template. It is valid. It provides a considerable amount of useful information about each item.
If someone can enlighten me as to what is funky about it without finding the need to resort to questioning Dave Winer's parentage or Ben Trott's ulterior motives, I would appreciate it.
Dave is speaking from his MT experience with MT 2.62, which used RSS 0.91 as the format for index.xml. His feed for that version is here:
http://static3.userland.com/mtweblog/index.xml
The test weblog itself is here:
http://static3.userland.com/mtweblog/
From what I see, all the problems went away with the release of MT 2.64 where RSS 2 is the format for the index.xml feed.
Randy, many of those "standard" RSS 2.0 elements -- item/pubDate, item/source, item/author -- are simply reformulations of Dublin Core elements. Dublin Core is a worldwide standard for metadata (it's actually an ISO standard now) and is used in many many HTML, XML, and RDF vocabularies.
Furthermore, Dublin Core predates the introduction of those elements into the 0.9x branch of RSS, so one could make the opposite case that Dave's RSS feed is the one that is "funky", and does not respect ISO standards.
Regardless of how you look at it, Sam's assertion is correct. The feeds produced by Movable Type are valid and conform to every aspect of the RSS 2.0 spec, as published at http://backend.userland.com/rss . The spec specifically states that any RSS 2.0 feed may use any element from any namespace for any reason whatsoever. It does not define what is or is not "funky"; that's just one vendor's opinion.
Mark, To be fair, we have yet to hear that it is UserLand's opinion, or even Dave Winer's opinion.
I am still curious as to what is Dave Winer's opinion. For now, all I have to go on is "funky".
Mark,
From the start of the RSS 2.0 specification:
RSS is dialect of XML. All RSS files must conform to the XML 1.0 specification, as published on the World Wide Web Consortium (W3C) website.
...and from the XML 1.0 specification:
Violations of well-formedness constraints are fatal errors.
...and:
Once a fatal error is detected, however, the processor must not continue normal processing
There is no wiggle room here, if what Sam states is true, and the feeds are not well-formed, the result is not valid XML 1.0 or RSS 2.0, and should not be processed. Anything else is in direct violation of the specification.
Jim, my guess is that you have two conversations tangled. Mark and I would agree that the feed that UserLand's RSS validator has no problem with is ill formed, and therefore not valid RSS 2.0.
Brad postulates that that is what Dave Winer is referring to as funky. If so, that problem has long since been fixed, as Brad also pointed out.
Randy speculates that it is the use of dc:date. But at this point, that is just speculation as to what Dave Winer meant.
Randy, fair enough.
The merits of pubDate vs dc:date were debated at length, and it appears that a majority of those that expressed an opinion preferred dc:date (though I respect that you differ on this topic).
It's possible that a public apology might be in order.
It's possible that I feel another parody coming on. Yes, that seems more likely.
Hugh: Yes I have and it hasn't been lately.
I'm curious to how many aggregators support <guid isPermalink="true"> over the mainstay <link> tag. It would be useful to know in light of the concern for breaking aggregators with an optional namespace.
Daniel (a.k.a. Archipelago): I did... via my weblog. ;-)
P.S. Since MovableType basically just adopted the template that is provided with the RSS validator, I can't help but suspect that any "funkyness" that is there was due to Mark or myself, and not due to any nasty commercial interests of SixApart.
In fact, I can't help but wonder if Dave ever tried to contact Ben&Mena, and explain what about their feed could be improved for the purposes of interop. As far as I know, no user of Radio UserLand's aggregator has every complained about Mark's RSS 2.0 feed... until today.
1. Does someone have a current unmodified MT installation for me to look at. So much confusion about what you get when you install it.
2. Sam, I have been emailing with all the people at SixApart for quite a few weeks on this subject. An alternative to #1 for Anil, Ben, Mena or Joi to write me offlist and say what they changed.
3. Why did Mark Pilgrim hack up his RSS feed? I don't get it.
4. My Internet connectivity is very intermittent. Sorry.
(BTW, my hacked RSS feed has nothing to do with this thread, and people who claim otherwise have simply jumped to the wrong conclusion. It arose out of a discussion with Joe of http://bitworking.org/ about his new comment posting system.)
The information on this thread is correct, we're using the MT template from the RSS Validator page, except with the default to excerpts instead of full entry bodies.
I'm unsure what confusion there is about what you get with an MT install.
Anil, since you're here -- and if you want to do this offlist that's fine -- what is SixApart's policy about RSS, going forward. In other words, what format is the default for new MT installations?
Dave :
http://www.movabletype.org/default_templates.shtml#rss_20_index
MT is including both an RSS 1.0 and an RSS 2.0 template by default, and as far as my knowledge goes, they are both automatically built.
In the default template MT links to the RSS 1.0 files.
Dave: the correct forum for opinions about my RSS feed are here:
http://www.intertwingly.net/blog/1462.html
or here:
http://diveintomark.org/archives/2003/06/12/how_to_consume_rss_safely.html
or on your own weblog.
I did not test a security vulnerability today; I did not expose anyone's computer to harm. I added styles to my feed, CSS styles, which several programs displayed (as they are designed to do). It is true that this led to a larger discussion of the potential to use similar means to exploit security vulnerabilities, and that's an important discussion to have, but I did not in any way put computers at risk (as others have suggested in other forums).
Please, follow up in one of the above forums.
An example. George Clinton is funky. Imagine what he would look like in a suit and tie with a haircut. So are the Neville Brothers.
Hey Pocky Way: "The Neville Brothers certainly qualify. It's even better if it has a funky beat, because that puts the smile on your face right away. The funkiness is the smile... Get it?"
BTW, do you know how funny it is to say my involvement here is hit-and-run. Here let me laugh. Hehehe. I've been dancing to this beat for seven g'dd'm years Christian. What were you doing in 1997? I was doing this stuff.
Dave, I for one don't get it. I said so in the blog entry that started this thread. Two days and 42 comments later, and I still don't get it. Why not identify what is funky? We have two or three different theories... why make us guess?
Flying back across the country, I had some time to think. I'm a bit troubled by the lack of forward motion lately in RSS. Profiles? Stalled. Namespace? Stalled.
Re: "...look back a couple of years from now and point out where a little working together would have made the world a better place..."
Dave, you are absolutely correct. I hope someday we will all look back on these days and see the missed opportunities for cooperation, and learn from them.
However, for the benefit of people who don't realize it, I feel it is important to point out that your idea of "working together" is "everyone doing it my way". This is not a flame; it is a quote:
"If they want respect for the formats and protocols they implement, they [Blogger and MT] must do RSS exactly as UserLand does."
source: http://blogs.law.harvard.edu/bloggingFormatsProtocolsMay2003
I dont claim to know Dave very well, but it seemed to me from the start that he was just waxing centrist and really wasnt making a diehard claim about MT not supporting RSS standards. Maybe Im wildly off base, but that seems to happen here and there. Whatever, its a sideshow.
MT currently generates compliant RSS out of the box. What more is there to ruminate over, whats to be gained?
Sam I don't want to argue about it, and getting into the details certainly would get into an argument. Today is the first anniversary of my almost dying of heart disease, and it probably had a lot to do with exactly this kind of stress. So I'm going to insist that you cut me some slack this one time. Thanks.
And as I said, the info is on Scripting News, frankly I don't see why you asked the question at all, just go read the site and you'll find out what I think.
Mark, you found the correct reference, but you omitted the context. I know you're really smart and you know how to read, so you must be making some kind of argument, and a personal statement, not one aimed at peace, love and understanding. What you say isn't true, obviously, because I implemented the Blogger API and Trackback according to the lead set by Evan and B&M. Now I say to them, why aren't you doing the same, and how would you feel if I implemented something named after your creation without following your lead. I don't expect you to understand this Mark, it's subtle, and would involve seeing the world from someone else's persepective. As I've gotten to know you over the years I've learned this is not one of your strengths. So be it.
But looking at the bright side, I remember you're the guy who didn't like what I said about CSS without reading what I said (you went by what Zeldman said, and he didn't bother to read it either, duh), so I guess this is progress. This time it appears you actually did read what you're criticizing. Mazel tov and many happy returns.
Have a nice day.
Dave, a number of people are guessing. Daniel Berlinger suggested that I ask you. I still don't know what you mean.
I am going to find a way to provide some forward motion in this space.
I wish you well.
Saturday, June 14, 2003
Why I called Movable Type's RSS support "funky"
by David Winer
Dave has mentioned that a read through the syndication mail list will illuminate.
To that end, I have collected all the links I could find related to the development of the RSS specification, from the release by Netscape in March 1999 through the release of the RSS 1.0 Proposal in August 2000. The "beginnings of the RSS fork" that covers the time before Mark Pilgrim's History of the RSS fork.
The more substantive links are arranged by date, and the whole collection as plain-text files are a available as a .zip or .tgz file, released into the Public Domain.
OK, I've returned home. I've read Dave's Why I said Movable Type's RSS support is "funky".
Despite what Dave apparently told Brad offline, I'm coming to the conclusion that the issue is the use of dc:date instead of pubDate. As near as I can tell this produces no incompatibilities with any Radio product, so I don't understand the issue. Meanwhile, I agree with Mark Pilgrim that ISO dates are more consumer friendly.
Given that MT based their template off of the one that the RSS validator uses, it seems more likely that what Dave should have issue with is the RSS validator than SixApart.
Summary:
I read Dave's note and got the impression it was because it defaults to RSS 1.0 not 2.0, not a dc:date vs pubDate issue.
Unless Dave is going to clearly indicate what he doesn't like, I'm just going to call it FUD and move on. YMMV
Could be dc:date versus pubDate, could be autodiscovery and link going to RSS 1.0, could be the existence of RSS 1.0 at all, could be the assumption that TypePad will prominently feature RSS 1.0, could be content:encoded versus encoded content in description, could be excerpts versus full posts, could be an attempt to present a united front against the new version of Blogger Pro producing RSS 1.0, could be meat, could be cake.
I've been in relationships before where I was supposed to guess what the other person wanted me to do. Past tense.
Re: "could be meat, could be cake."
Bwahahahaha. I doubt if 3 other people get this reference.
Apparently Dave Winer thinks of himself as Don Quixote fighting for unity against Microsoft rather than as Don Corleone making offers that other RSS developers can't refuse. Whatever his intentions, premature standardardization of weblog and RSS technology in the name of interoperability is more likely to lead to commoditization. That makes it all the more likely that large vendors would dominate if they saw a chance for profit, not less likely as Winer professes.
Your strength in the weblog / RSS world is how fast you can move, not how close you can stand together. Sheesh, any big company could buy all the players in this niche with a couple hours' worth of revenue! The more this stuff is a commodity, the more easily it can be bought and sold, or ripped and replaced, or simply made irrelevant by new features in Windows, Office, IE, Google, Java, or whatever.
You folks just keep up the good work of innovation and experimentation! If you keep things reasonably simple, interoperability will take care of itself if there is the will to do so. That lets good ideas like the Comment API Sam discussed recently prolferate, and bad ideas like [oops, no flames :-)] die out, unhindered by gatekeepers, whether they be standards committees or Godfathers.
Based on this comment, it seems as though the 'the three bits of funk' refer specifically to the three namespaced elements.
So, 'funky' apparently means 'uses namespaces'.
The operative word is "apparently".
What this means is that in order to be non-funky, I need to disadvantage tools like NewsGator which benefits from my use of dc:creator.
Lovely.
Yes, "apparently", but I don't see any other candidates for 'three bits of funk' other than the blogChannel namespace declaration plus the two namespaced elements blogChannel:blogRoll and blogChannel:mySubscriptions. Do you?
If I'm correct, then Dave is not only calling elements that replace roughly equivalent optional RSS 2.0 elements (like dc:creator vs. author) 'funky', but any extension of RSS using namespaces, even an extension he originated.
None of this gives me warm fuzzies. This feels like the sort of trial balloon a BigCo sends up to check whether they can get away with something. I don't think that's what Dave is doing, but it still feels that way.
Grant, we don't have to state it; Dave has already stated it.
http://blogs.law.harvard.edu/bloggingFormatsProtocolsMay2003
"""If they [Blogger and MT] want respect for the formats and protocols they implement, they must do RSS exactly as UserLand does. The thing that Blogger and MT currently call RSS is not only not what UserLand does but it isn't even an improvement over what UserLand does."""
I'm sick of this. I'm sick of RSS. I'm sick of this debate. Really sick of it. The minute a viable alternative to RSS 0.9x/1.0/2.0/3.0/4.0 emerges, I will support it exclusively and deprecate every single one of my RSS feeds. There will be breakage. Serious breakage. But it will be worth it if it means I never have to talk about RSS again.
Mark: hear hear. As a long-time follower of developments in the RSS arena, the whole show is a shambles.
Or perhaps RSS could stand for Really a Set of Sub-standards.
I'm not hopeful that a viable alternative to RRS will emerge, unless the W3C or a similar body takes it on board and puts energy into creating momentum for the new standard.
Sadly, as RSS becomes more widely used, inertia comes into play, and I can understand that some groups and individuals are reluctant to change to something else. But with four dialects to support (and more on the way I suppose), what the heck, you might as well support another one, right?
At least the One True Syndication Format offers the hope that one day we can turn off RSS, add an entry to the Great Book of Screwups and Lessons Learned, and move to greener pastures.
Serious breakage would be a blessing at this point. It certainly couldn't make the situation much worse.
Bah!
Mark, why throw the baby out with the bickering community?
Its a shame that innovation in RSS has been totally killed by the political strife, RSS 1.0 had real potential to be an interesting format if it hadn't been sabotaged.
But if you're sick of the conversation, disengage, stop paying attention to all the various RSS conversations. Check back a year or two from now, I promise your RSS will still be valid, and you'll still be on the cutting edge.
It is sort of like any other forms of activism, you've got to disengage to step back and look at all you've accomplished. And by any sane standards, RSS (and your and Sam's contributions to RSS) is a success. You just don't see it when you're on the front lines.
What I wrote was part a joke (I'm not likely to give up using RSS any time soon) and part a statement of my dislike of all this funky/non-funky stuff. (I also wasn't paying attention and didn't expect the except to appear here; I didn't trackback you as I didn't really think it justified it).
Anyway, I am likely wrong because Dave Winer hasn't explained exactly what funky is, but I get the impression that the use of namespaces is what is funky, so by that definition, you can add RSS 2.0 to ESF and Aaron's RFC822.