Re: [syndication] An Open Letter of Mindless Blathering

["Dave Winer" <dave@userland.com>]

Beautiful piece Morbus, really, no sarcasm, but I have a different opinion.

Collaborative filtering is the way to go.

1. Make it easy for your users to publish their subscription lists.

2. Make it easy for everyone to find an aggregation of those lists, ranked
by popularity. This tends to make the bad or dead channels fall to the
bottom, off the list.

We're off to an excellent start with this with Radio. The hotlist and
mySubscriptions.opml, being available publicly for all upstreaming-enabled
users (not enough people yet, we're going to do something about that in the
next major release) actually gives us good info about what's available.

Add to that Jeff's work on the Newsfeeds blog, and we're starting to get
somewhere. It's just like movies, music and sports. You need stats, and you
need sources with credibility to tell you where the good stuff is.

You're absolutlely right that it's grown far beyond the ability of any one
individual or company to sort through it all, that's why it's best to let
people use their minds and vote with their mouses.

Dave


----- Original Message -----
From: "Morbus Iff" <morbus@disobey.com>
To: <syndication@yahoogroups.com>
Sent: Thursday, June 07, 2001 8:06 AM
Subject: [syndication] An Open Letter of Mindless Blathering


> Dear Mr. Blatherskite,
>
> I can't help but feel frustrations over the RSS community as it is, not
> over the naming or the cross-cultural fighting, but the duplication of
> effort (much like Mr. Binky told me when he saw AmphetaDesk and compared
it
> to Radio Userland, wondering why AmphetaDesk needed to exist). Not in
> client efforts, however, but rather in knowledge effort.
>
> (As I read over this, in postscript haste, I realize that I've uttered
much
> the same frustrations that have plagued other new technologies. A poser.)
>
> I've been thinking of the lists that contain thousands of RSS feeds. Much
> like Yahoo satisfied the need of a directory when no one knew where to
> look, RSS (currently) needs the same sort of directory. And everyone is
> jumping in with their own (even I make this transgression) [2].
>
> There are tons of lists all over the place [1]. Each ocs list has
different
> internal thinkings of how to do things. Duplicates and errors abound,
along
> with lack of loving.
>
> Duplicates abound because lists don't check against "www.cnn.com" vs.
> "cnn.com". Or they don't compare dynamically generated listings by
chopping
> off the URL after the ?, creating such wonderful duplicates where the only
> difference is the number of items outputted (o=30 or o=24).
>
> Same with Moreover. Moreover is wonderful and spits out their own .ocs
> directory. Instead of people integrating that list, they embed the list
> into their own directory (again, silly Morbus, you do this too), creating
> such wonderful duplicates of differing URL parameters, but similar
content.
>
> And even more so, the love is gone from lists. Massive robotic scripts
> merge all this crap together, without checking for dynamic user based
feeds
> (like custom searches through sherch.com or moreover.com, or created feeds
> from 10.am). Lump it all together, binky - someone has the same interest
> somewhere. The difference between giving a fish and teaching to fish. [3]
>
> It's frustrating. I compiled a large list of 2300 feeds, after checking
for
> duplicates across titles, descriptions, authors, urls, and more, I
wheedled
> it down to 1900 or so [2].
>
> Jeff Barr from Headline Viewer has an alias list [4], where he keeps track
> of sites that have different URLs but spit out the same content. I used
his
> newer, unpublished version for the AmphetaDesk list, and I found about 15
> other aliases that he didn't have.
>
> That these other lists didn't have.
> The other lists probably had aliases, but didn't know it.
> Jeff and I don't know it either. Not good.
>
> And all these lists sit in raw data. They don't move or interact with the
> common user (we do all this for them, don't we? at least, that's been the
> glorious herald of the RSS naming politics - "unconfuse the user!"). All
> this information is being given to the geeky geeks, and not to the
> "people". Silly.
>
> I keep thinking there needs to be an interface to a master list, one that
> can be "the one and only", bordering on Nazi-ism, allowing export in any
> format, allowing "show me new feeds from X days ago" or "show me updated
> feeds from Y hours". One that would allow people to advogato their
opinions
> about the feed. One that would allow "people who sub'd to this feed also
> liked...". We have shifting of data with Radio Userland, "link via" virii
> ("link via Brainrot, via Dave, via Eugene..." ad infinitum), but no
> shifting of the user.
>
> A major problem with RSS is not petty bickering, but the sheer amount of
> feeds available. People are clamoring for searches, meta searches, author
> searches, "things like this" and more. Whilst it may not be possible to
> aggregate all data into a massive google like search engine, it's
certainly
> possible to create a meta/user/love directory of "if you like this, you
may
> like this".
>
> Mike Krus from NewsIsFree.com. His site is roaring (wonderfully). At
times,
> I think, well, screw it, he has everything pretty much all set up, it'd be
> easy as pie for him to make an interface to his channel list, where we
> could query the list like "new feeds since X" or "feeds updated since Y"
> and get an ocs export.
>
> That'd solve a lot of problems, actually, with AmphetaDesk. I just don't
> have the time to devote to keeping a list perfect and updated. Hell, the
> only reason I got into RSS/syndication and AmphetaDesk was because I
didn't
> have enough time to read all the news I wanted to, and it was stealing
time
> from my other work. (Of course, Mr. Blatherskite, as you know, the
solution
> to the problem has now become the thief).
>
> But, I get worried about losing control. Ah, the ubiquitous control. It'd
> be easy for Mike to keep a list updated. But I have all these wonderful
> ideas to make a list even better and more Amazon-y in terms of features. I
> fear that he wouldn't have the time to implement them all. I fear that I
> could implement them, but not service the growing populace and power
> needed. Hell, I fear that he won't take a leap of faith, or <gasp!>
> wouldn't agree with me. It's not just you, Mike (of course), it's
everyone!
> Fervency causes fermenting that isn't easily swallowed.
>
> In a corner of my head, I say "screw control, you moron, if you pass it
off
> to Mike, then you can just be an annoying user and harass him instead of
> the other way around." Oh, quite good. But to pour a lot of fervent
beliefs
> into another's hands to mold is quite the same leap of faith some ideas
> require. (The same voice says: "he has a Paypal icon on his site! he must
> be desperate for cash! your ideas will fruit, and then rot with death!
> muahahhah!" I dislike this voice, Mr. Blatherskite).
>
> This diary entry is far too long. <skkizzzt!!>
> Being off my chest, is it any better? Perhaps.
>
> The pessimist says an interface will be built without my "valued" input
and
> it'll be all wrong. When it fails, they'll point to this message and
chuckle.
>
> The optimist says that Mike will say "hey, binky, you could have just
> called me and saved yourself some face. Sheesh. You wanna move in?"
>
> The cynic says I just got fired for wasting half an hour.
>
> Sigh.
>
> [1] http://groups.yahoo.com/group/radio-userland/message/9079
> [2] http://groups.yahoo.com/group/syndication/message/1844
> [3] http://www.disobey.com/amphetadesk/news.htm
> [4] http://www.vertexdev.com/chv_aliases.xml
>
>
> Morbus Iff
> .sig on other machine.
> http://www.disobey.com/
> http://www.gamegrene.com/
>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>