TechMeme: the anti-linking engine

I’ve noticed this several times and thought I’d bring it up.

TechMeme seems to penalize bloggers who link to other bloggers. Most bloggers believe that a major part of how TechMeme decides which is the most important story is to count links. That isn’t true in following mine, and other people’s results.

I believe there’s a “linking penalty” on TechMeme. At least it seems that way after doing my own link counting.

Let’s say there’s three stories.

Story A links to B and C.
Story B only links to A.
Story C doesn’t link to anyone.

Who will be most popular at TechMeme? Often times “C” will be. But shouldn’t “A” be? Since that’s the one that has the most inbound AND outbound links?

In my experience it won’t be and that often C, who didn’t link, or get linked to, will often be the top pick.

Why is that? Because I think Gabe wrote an anti-gaming algorithm which looks for bloggers linking to each other. I believe his algorithms are penalizing bloggers who often link to each other.

Here, look here at the story about Google Gears on TechMeme.

There are currently five articles that are showing up as headlines on TechMeme (this was the order that they appeared at time of writing — being higher is better).
1. By Artur Bergman on O’Reilly Radar. (His article doesn’t link out to other bloggers who covered this story).
2. By Nick Gonzalez on TechCrunch. (Nick links to me).
3. By Martin LaMonica at CNET. (His article doesn’t link out to other bloggers who covered this story).
4. By me. (I link to both Artur and Nick’s articles).
5. By David Berlind at ZDNet. (His article doesn’t link out to other bloggers who covered this story).

Now, how many blogs are linking to each?

Artur has three links, according to Google’s Blog Search at the time I wrote this article. (No outbound links).
Nick has five links. (Nick links to me as his only outbound blog link).
Martin’s has zero links. (No outbound links to other bloggers).
My blog has eight links. (The most links!!! and I link out to two other bloggers, so most inbound and most outbound links).
David Berlind has no links. (No outbound links to other bloggers).

So, lesson learned. If you wanna be the top dog on TechMeme, don’t link to anyone else but get them to link to you.

Or, is something else going on? I’m sure Gabe will say that most of the eight people who linked to me don’t count to his algorithm because he only looks at what the seed bloggers (folks who’ve been hand picked to be counted) are linking to. That might be true, but I’ve looked at enough result sets now to start a theory that there’s something else other than a straight counting of links going on.

Looking at Geni, new kind of family tree software

One of the guys, David Sacks who was CTO, who started PayPal has spun out and started another company. This one aimed, Geni, at helping you to document and share your family tree.

I have an interview with David and a demo too (I’m embedding the demo here).

[podtech content=]

Inside the Gears of Google

So, what did Google just do with its introduction of “Gears?”

Setup its suite of office applications to go offline. Oh, and everyone else’s too.

Zojo? ConceptShare? Zimbra?

All will be able to go offline soon if their developers adopt Google’s Gears.

UPDATE: David Berlind at ZDNet has the best insight I’ve seen posted so far about Gears and has a podcast/interview with Linus Upson, director of engineering at Google

Right now I’m using my computer while not connected to the Internet. I’m typing in Windows Live Writer’s window and I’m surfing Google Reader while unconnected.

Some things that work better? Google Reader doesn’t hesitate or “stutter” every 20 posts like it does when it’s online. Some things that don’t work as well? It doesn’t download images so posts have grey boxes in them.

I just reconnected and it instantly shared all the items I had maked as shared.

Nice implementation and works simply and easily. I’ll definitely be using this on my next plane ride.

I wish it ALWAYS worked in offline mode, though. Why do I need to click a button to resubscribe? The Gears team told me this is a choice the developer will need to make.

Oh, while offline, if I reloaded my browser, or accidentally closed it, it goes back to my Reader automatically. That’s nice.

Look for more offline applications to come from Google soon. They don’t have any to announce yet, but said they are working on their Docs and Spreadsheets.

What really just happened? The Web got a little further away from Microsoft’s platforms and Google revealed a little more about its platform dreams.

What do you think? Are you looking forward to developing an app using Gears?

Google brings developers offline with “Gears”; new offline Reader

Right now in Sydney, Australia, the first of 10 Google Developer days are starting up and the audience there is hearing about several new initiatives. The most important of which is “Google Gears,” an open source project that will bring offline capabilities to Web Applications — aimed at developers. More on that in a second.

Also being announced today (and tomorrow here in San Jose) at Google’s Developer Day:

1. A new version of Google Reader (shipping this afternoon) that is built with Gears that will enable that reader to work offline. To make use of the new offline capabilities you’ll need to load the Google Gears plugin first (also available now).
2. A new mashup editor that is aimed at developers familiar with HTML and JavaScript. The Google Mashup Editor offers a simpler way to deploy AJAX user interface components atop existing feeds and Google Web Services. This competes pretty headon with Microsoft’s PopFly that was shipped about a week ago. Unfortunately this is going to be a limited beta, so you’ll have to sign up for it and wait.
3. Google Web Toolkit has passed a million users and they just shipped a new version of that.

Regarding Gears. It works on Macs, Windows, Linux on IE, Firefox, Opera. Enables versioned offline storage. Extension to HTML/JavaScript. UPDATE: Opera support isn’t finished yet and Safari support is coming “later.”

They are showing me a demo of the new Google Reader using the new Gears plugin. After you load the Gears plugin you get a new icon at the top of your Reader window which enables offline capabilities of Google Reader. They showed how Google Reader then downloaded 2,000 feed items. They took the browser offline and it continued to work great.

Codename of Gears was “Scour.” The team made fun of their code name saying that the world had “Ajax” and now it needed to “Scour.”

Gears is a 700kb install. Gears consists of three modules that developers can talk to via an API (details will online at the site). The three modules consist of:

1. A Local server. An object that your app can talk to and get stuff (images, JS files, HTML, etc). Atomic updates.
2. A data store. SQL light. You can talk to that via SQL. It’s like a cookie on steroids that you can talk to via SQL so developers can store data offline and talk to that store.
3. A worker pool that’s designed for background tasks. Keeps track of offline and online and helps sync and keep your apps “snappy” no matter what state your connection is in.
What won’t Gears be good for? Something like Google Search. Why? Because that assumes you want to search all the items on the Web. It’s better for Web developers who are building apps that have a specific amount of data to interact with.

Gears supports using Adobe’s Apollo and Flash and should support other technologies including Microsoft’s Silverlight.

Gears will be submitted to a standards organization eventually, they said, but want to make sure the technology is rock solid first.

More on this from TechCrunch’s Nick Gonzalez, and from Artur Bergman on O’Reilly’s Radar site who are sitting a few feet from me.

The detail in this post came from Google’s Vice President of Engineering, Jeff Huber, among other people from Google in the room — I’ll try to get more names shortly.