DISQUS

One criticism I haven't seen yet:

Google doesn't permit crawling of its hosted news content, even though Google News relies on crawling news content from other sites. (Recall the same was true for the comments on news stories Google announced a few weeks back.)

E.g., this robots.txt file http://ap.google.com/robots.txt states that nobody may crawl this hosted news article http://ap.google.com/article/ALeqM5gURr_MR-Wtuq...

One consequence: you can use Google Search to find Yahoo News-hosted content, since Yahoo News doesn't block (http://news.yahoo.com/robots.txt is 404), but you won't be able to use Yahoo Search to find Google News-hosted content.

It's nice to be the market leader in search.

Robert, I personally study your link blog. It's in my Google Reader and I learn a lot about tech news trends from it. Almost certainly, I apply that information to what I do.

The idea of attribution is very tricky. Do you know that the vast majority of the time, when a blogger learns something on Techmeme and blogs about it, Techmeme won't get credit for the discovery? That's very similar to the situation you alleged. Is that wrong? I'm not sure. BTW, you're probably among the best bloggers as far as crediting Techmeme goes, but I bet you still don't do it half the time.

Anyhow, for the record, I should state that lately, Techmeme has NOT used your link blog.

Hey Robert. Yeah, Techmeme doesn't aggregate all the forms of video as deftly as it could if it were omniscient. But there are some strategies for getting your videos linked to on Techmeme:

Basically, create a BLOG POST that's the clearinghouse for all your video stuff related to that news event. Create the post when the news happens. Give it an appropriate title. Embed any videos in that post. Link to your steaming video in that post. And add as much text as you can. No text is OK as long as there are videos, but the more text the better, for semantic analytical purposes. Just a few sentences will help.

Among other reasons, doing this helps because many/most links from other blogs will go to that one post.

Note this might be a good idea even if Techmeme didn't exist...

You sure about that? I was taking a look today, and I think that figure is just people who installed the application. (Which is still impressive, but doesn't indicate people actually sharing.) Notice the users of the app who haven't specified their shared feed get Mario's items as their default shared feed. That creates the impression that more people are sharing than in actuality...

iPhone sales begin 6PM Friday, right? So lining up Thursday evening means an all-night and all-day wait. 18+ hours. Are you really up for that?

Robert: no, most Techmeme readers weren't former Google/Yahoo News readers. (Think about it...even if you happen to be an exception.)

Techmeme readers are people who previously scanned lots or blogs and/or Digg/Delicious/Slashdot for interesting Tech biz/product/issue/trend stories. Maybe Techmeme "competes" with those sites, but only a small minority of their readers were at risk of defecting, and their hold on those readers was tenuous to begin with.

Robert, this is essentially your post from 6/6 but with "Plaxo" in place of "Sun blade servers". So I'd like to refer you to my earlier response. :)

BTW, I'm quite pleased with the direction Techmeme is headed. In fact, the nature of this post and the ensuing comments very much reinforces that.

Techmeme actually wants to be a selection of interesting Tech news stories. A particular take on "what matters", suitable for a particular audience.

In many ways, it wants to be the anti-Google News, though it other ways it needs to be more like it.

I'm agnostic on professional vs. amateur. Ideally, whoever has the best take on the story for Techmeme's readers should be the one appearing.

Techmeme often comes out ahead when people compare it to Google News (recently on Read/Write Web, earlier on eWeek). Though to be fair, these kinds of comparisons are usually done by people who BROWSE tech news. People who do mainly news searches will clearly find Google News superior because Google offers an excellent news search while Techmeme offers none. (And Techmeme indexes fewer sites in the first place.)

BTW, I personally give the advantage to Techmeme on the blade server story, which sounds rather boring, something best passed over. Is the fact that you covered it for Podtech affecting your judgement in this case?

Hey Robert,

Oh, I always love to read reverse engineering reports...

Techmeme's algorithm is so complicated at this point that I could write a long article on various ways linking can help you and/or hurt you, ranking-wise. Instead, I'll just make the following vague claim: linking to new news in a way that doesn't appear to be an act of quid pro quo should never hurt you.

Remember, so many factors enter Techmeme's rankings lately, that it's easy to attribute to link patterns something that's actually caused by other factors.

BTW, is it just me or is the amount of news today freaking you out?

Your link blog is a marvel, and gives me ideas on how to improve Techmeme. I just promoted it in Techmeme's twitter feed:
http://twitter.com/Techmeme/statuses/78199982

So perhaps threatening to not link works as well as not linking! :)

Well, surely we're not all adults, in a literal sense anyway. Though it's not really about "the children" for me.

The asterisks basically help to prevent making certain people irate. (Yes, I do get emails about dirty words.) Because they're not really obscuring the words, as you pointed out, there's not a huge downside to this, in fact some see it as a fun sort of "accent". (I think the South Park creators commented once on the upside of bleeping.)

I would say Robert's classification for his link blog is somewhat aspirational, but it's a good aspiration, because his link blog is excellent. Exceptional in many ways. Alot of people would be wise to scan it. I'm one of them.

That's it. I broke down and signed up for Twitter: http://twitter.com/gaberivera Warning: entrails.

) = :)

Heh, how DID it get on Techmeme? It almost seems like there was a little engineering on your end... )

Robert: I agree. For many key technologies you've been an early and vocal (and even articulate) adopter.

But c'mon, we're approaching a decade of syndication and you remain the world's most voracious feed reader! So to repeat myself, in "some respects" you're an edge case.

Accept it and embrace it. :)

"I read more feed items in Google Reader than any other human being alive"

So maybe in some respects, you are in fact an "edge case"? http://scobleizer.com/2006/01/26/scoble-im-not-...

P.S. Edge cases are great. And Techmeme loves them.

Robert, in general, to answer your question, no, the "root" article should not always be linked. Because stories evolve over time, the "root" articles tend to lose value for Techmeme's readers. How many days after Google's YouTube acquisition should TechCrunch's original scoop have appeared with Gootube stories? I'm fine with "1 day" as the answer myself.

In this specific case however, I see your point. It's only been a little over one day, and the story hasn't yet evolved much.

TDavid's remark touches on another point: very often non-originating posts tell the story much better. In these cases, it's actually in Techmeme readers' interest when the non-originating post is emphasized, though I can understand why the originator would get frustrated with that outcome.

Anyway, I think it's my job to find the right balance. That job's not done yet.

Scratch that..."Wikipedia" did appear in the title, but as the very last word in the French subtitle portion. Anywayz, a descriptive title probably would have helped yesterday. Too bad Techmeme can't rewrite your title.

Robert,

Philippe's post DID make Techmeme, the day before:

http://www.techmeme.com/070208/p56#a070208p56

It later passed because Techmeme's a news site, and old news goes away.

Whether Duncan's post should have received such placement the next day is debatable, but he did offer some more facts and analysis, and actually used a descriptive title, unlike Philippe, who somehow didn't even mention Wikipedia in his title. :)

Ah, wasn't aware of scoobietron.wordpress.com until looking over your stats. Looks like you're keeping that guy busy. :)

How do you say "Scoble"?

Check out TechNet.org . Not sure how influential they are, but here's how they describe themselves:

"TechNet is the bipartisan, political network of CEOs and Senior Executives that promotes the growth of technology and the innovation economy. TechNet focuses on politics and policy by bringing its members together with our nation's policy makers to sustain and advance America's global leadership in innovation. In addition to its offices in Washington, DC and Silicon Valley, TechNet has a presence in the high-tech centers of Southern California, New England, Texas and the Pacific Northwest. TechNet's members represent more than one million employees in the fields of information technology, biotechnology, ecommerce and finance."

Wow, thanks Thomas.

John, not expecting it to come to that! Anyway, I've always cited gray areas as I have above. And I don't advertise my sites as "fully automated" or anything like that because there's no workable notion of "fully automated". You can't get a tech news vertical without manual configurations to ensure that.

Rex, thanks. Actually, I described it as a "pseudo-vacation" above. I only work half of the time on pseudo-vacation.

[Third installment. Maybe I'm done here. Rather happy to have written this up somewhere.]

You said I claimed not to "mess" with the site in my first comment. Nope, I mess with it.

Here's an example: http://www.techmeme.com/061215/p17#a061215p17

See that "Tribute to Lonelygirl15" post? I removed it almost immediately after spotting it.

Why? Its appearance was a fluke. It was basically a flaw in my system combined with an unusual type of redirection used by the longelgirl15 site that fooled my system into thinking people were discussing that post. They weren't, so I felt best to zap it.

Such examples are uncommon but they arise. Another: a couple times a month my scraper selects the wrong title for a given article url. Whether or not I fix the scraper, I may manually change the title if the erroneous title is awful enough.

Another example, where I need to be deliberately vague (sorry): suppose I see evidence of excessive logrolling that degrades results, and therefore change a particular global system parameter that has an immediate impact on a headline currently live. Is that "messing"? I dunno. But stuff like that has happened once or twice in the past month.

I guess to summarize everything: 1. Direct changes by me personally are rare, and normally to work around bugs. 2. Many, many individuals "mess" with Techmeme. 3. Please recognize my system is always getting smarter and will seem increasingly humanlike over time.

Epilogue: interesting peripheral issue: Would Techmeme be better if editors under my direct control actively messed with the live site? Maybe...I'm undecided myself.