The Feedmesh after school meeting last night was the big buzz of the evening last night. I had four or five folks come up to me and thank me for coming to the event to give the publishers perspective on the FeedMesh project.
So, what is the FeedMesh?
OK, here is the basic idea: When a blog is updated it pings—sends a signal—to a number of services to let them know that the content has been updated. This method, of the blog pinning the search engine, is much more efficient then the blog search engines check the blog every day, hour, or minute to try and see if there has been an update.
Right now blog software, and bloggers, ping dozens of sites including pingomatic, weblogs.com, blo.gs (now part of yahoo), and the individual services (technorati, feedster, etc). So, Bob from Pubsub suggested that there be a “cloud” which has all the updates in it that the blog search engines can tap into. One service to unite them all, if you will.
This Feedmesh as you can imagine, is just starting to become powerful, and with great power comes great responsibility—and controversy.
My first question was “who owns and controls this?” I might as well have thrown a grenade in the room. This was a room full of tech people mainly—the people who do the work. The CEOs of Feedster and Technorati were not in the room, for example, but their tech guys were. That
question of who owns this lead to a huge debate which will probably
result, IMHO, with a non-profit type org with some basic rules for the
use of this valuable service.
Now, what this cloud does is make it so anyone can tap into the live
web. From what one insider told me Technorati had their huge jump in
number of blogs when they tapped into the FeedMesh.
The second big question I had was, “I understand everyone is now
listening to this, but who is giving their data on new blog posts into
it?” The answer to that was dead silence. Turns out only Pubsub is
putting their data into it--Technoati and Feedster folks are way to
smart to give their data (i.e. their entire business) into the cloud.
You see, Pubsub was the third place player so they suggested—under the
guise of the good of the community—that everyone share the blog updates
and compete on the services they wrap around them. That’s a great
position to take when you’re the #3 player, of course Technorati and
Feedster have a lot of data coming to them that if they put it into the
cloud would make their service on par with Pubsub and new competitors
(think Google, Yahoo, MSN, VC-backed startups).
I’m sure the folks from Pubsub will take exception to my business
analysis of this and I respect them for defending their “good of the
community position” even if I don’t buy it.
The third big question/issue that came up was the problem with ping
spam. Pingomatic is getting 70% ping spam, and having this cloud means
that the blog-spamming scum have a clearing house for their efforts.
For example, you could create a blog that links all the new blog posts,
or scrape them and put that data onto your blog with Google Adsense
around it. You would get away with it too since Google isn’t being
aggressive about shutting these RSS-scraper down—something I’m very
disappointed with Google about. However, they are working on the issue
with me and I have faith that they will see that their “not our
problem” stance is absurd: people killing your publishers IS YOUR
PROBLEM Google.
My feeling is the Feedmesh should be a private effort by a group of
services and publishers, organized by a board, and unlimited access to
the cloud should be limited to members. Limited access (i.e. for a
certain amount of time/bandwidth) per day should be open for people to
experiment. Open access to this is just too powerful for everyone—i.e.
spammers—to have access to it. There should be some controls.
However, I give this concept a 50-50 chance of going forward because if
you were an established player like Feedster and Technorati you’ve got
zero to gain by putting your data in. If the big players only take (as
they currently are) and don’t give this service is only going to be
60-80% of the picture. Truth be told if I was CEO of Technorati or
Feedster there is no way I would ever dump my entire business into an
uncontrolled cloud. Forget it.
FeedMesh 101 (or why it will probably fail to be the one service that unites them all )
Reader Comments
(Page 1)2. I doubt a closed Feedmesh could succeed. The gamble has to be that Technorati will eventually join because the benefit outweighs the cost.
Posted at 6:38PM on Jun 25th 2005 by Randy Charles Morin
3. Closing it to non-members is just going to create an incentive for smaller players to create an alternative infrastructure. Sure spammers will get into that, but closed systems usually get trumped by open ones. So - I disagree with the concept that unlimited feedmesh access should be closed to non-members. That is a formula for the "big" players to try and create lock-in, which would inevitably happen under such a system.
4. I was one of the Technorati tech guys in the room.
You wrote: "Technorati had their huge jump in number of blogs when they tapped into the FeedMesh."
Technorati is not currently tapped into the FeedMesh. Perhaps you are confused with Scott Johnson of Feedster's comments about their service's recent changes?
Posted at 2:22AM on Jun 26th 2005 by Niall Kennedy
5. The problem with this "group" and "board" is that it quickly becomes a cartel and prevents new players from entering the industry.
Which leads me to the point of Technorati and Feedster, the should be all over your idea of a closed industry group running the Feedmesh. It essentially gives them control of who has access to the Feedmesh and let's them determine who their competitors are. So while they'd have to share their databases of information with each other and compete on services, but if it means potentially keeping Microsoft and Google out of the market and making them buy your company to get into it, you do that deal instantly.
6. You wrote: "Pubsub was the third place player so they suggested—under the guise of the good of the community" This is simply not true. While others may have more users than we do, PubSub clearly led in the number of feeds monitored *prior* to our deciding to support the FeedMesh. We have only saw others approach our numbers after our data was made available via the FeedMesh.
The FeedMesh was proposed at Foo Camp by folk from IBM, Technorati, Feedster, etc. PubSub didn't even have anyone at the meeting.
bob wyman
7. I've posted a "FeedMesh Primer" here:
http://www.johnkeegan.org/blog/_archives/2005/6/27/978881.html
for anyone who wants to learn a little more about the concepts behind the FeedMesh.
Posted at 10:31AM on Jun 28th 2005 by John Keegan
9. This reminds me a lot of the myriad problems of peering between ISPs (no wonder Sam Ruby called FeedMesh a peering network). Even though the point of the Internet is to connect to share information, larger ISPs hold access to their content ransom, making smaller ISPs pay well over odds. Being in Australia as I am, our whole industry suffers continually on pricing by the US-based ISP cartel's decision not to peer with Telstra. I hope that blog aggregators don't fall victim to the same evil, which produces a small number of lazy incumbents and stifles innovation.
If this cartel does come to pass, how much will it cost to join? What will be the barriers to entry? How dominated will it be by American companies? How could any of its members sneer at Microsoft et al for not "getting it" about open systems when they belong to a closed shop?
Posted at 1:40PM on Oct 8th 2005 by Paul Montgomery
12. "You see, Pubsub was the third place player so they suggested—under the guise of the good of the community—that everyone share the blog updates and compete on the services they wrap around them."
Wait a minute, Bob clarified that the Feedmesh concept was not originated by PubSub. Jason should ammend his post to clarify this piece of misinformation.
Posted at 5:17PM on Dec 16th 2005 by Cameron Barrett
1. What you stated implies that the "entire business" of Technorati and Feedster is collecting and processing pings. Of course, that's silly.
Posted at 6:31PM on Jun 25th 2005 by Gabe