Thread

Earlier this month those clever folks at Happy Cog launched their new (and beautiful) blog, Cognition, in which they wholeheartedly embrace Twitter as their primary comment stream. As Jeffrey Zeldman says in his inaugural post:

Speaking of experiments, there’s our comments section. Everybody knows inline blog comments are going the way of the BBS and Gopher sites of yore. We’re not ready to say “comments are dead” (we’ll leave that for Wired Magazine’s next cover story) but we have noticed the smell, and we’re doing something about it.

Kids today are more likely to respond to a blog post on Twitter than in the article’s comments section; so we’ve collocated our comments on Twitter. Share a tweet-length response here, and, with your permission, it will go there. If you are moved to respond with more than 140 characters, post the response on your website, and it will show up here. Clever, these Americans.

Integrating Twitter comments isn’t new but they’ve taken it to a whole other level than I’ve seen done with EE before:

1. They are hooking into OAuth and passing the user back
2. They must be storing some Tweet data locally

I’m just a lowly front-end savvy chap for whom the mechanics of this are fairly opaque. But it seems clear to me that they gotta be pulling some data into EE from Twitter because Greg Storey’s inaugural tweet comment is now 11 days old and Matt Harris of Twitter tells us that:

At the moment the [search] index is ~5 days of relevant Tweets

Intrigued by this, I’ve asked some questions over on EE Insider but other things that happened last week are clearly occupying people’s thoughts more.

Can anyone shed any light on how those clever Coggers (Mark Huot in particular) are doing this?

Presumably, all they need to store is the Tweet ‘status’ URL, though they could be storing more so as not to be at the mercy of Twitter up-time.

But how?!

Ah, hang on, they must be storing more because the comment text itself differs:

* The tweet includes their custom short URL (wonder what they are using for that)
* The comment on their site strips it out (as it ought)

See attached.

Hey Jonathan…

I found this thread as I was about to embark on re-producing an extension which does what Cognition are doing. It’s a great idea and it doesn’t seem too hard.

It seems it was created by Mark Huot as a custom solution for the project. I wonder if he’s planning to release it?

Have you found anything that already does what Cognition do?

Hi 1px solid (great name!). Nope, I’ve not heard anything. Any insights you can offer I’d be most grateful for!

Jonathan -Any update on this? Did you figure this out? Would love to do the same thing.

1px Solid - You mentioned you were thinking of building an extension for this - update?

thanks

No, nothing, sorry. I might try asking on the EE Podcast some time.

Datagrab can import feeds etc… into EE channels, on my own site I use it to have a copy of my Delicious bookmarks in my EE database.
Have a look, it may help:
http://devot-ee.com/add-ons/datagrab/