Coredump

I just noticed that my Twitter timeline RSS is not being displayed in my tumblog, so I went back to Yahoo! Pipes to sniff around. It appears that the Twitter feed is not coming up valid (or Pipes could not parse it).

(Update: The Twitter feed is valid, according to the W3C, but the Feedburner one is not. Oookay, so Pipes accepts an invalid feed, but not a valid one?)

As a workaround, I plugged in the feed through Feedburner and fed the Feedburner RSS back to my Pipe, et voila! my tumblog is back to normal.

In related tumblog news, popular social bookmarking site, del.icio.us, has been rebranded as, plainly, delicious.com. The old URI still works — I like it better than the boring new one (love the new layout though, much cleaner and faster), which lacks the, er, whimsy-ness of the old domain — but just to be sure, I upgraded the link to the new feed.

I’ve made several improvements to my tumblog.

Previously, I merely plugged in the RSS output from Yahoo! Pipes to Feedburner, and served up the content using Buzzboost. Using techniques from Cristiano Betta’s lifestream and elsewhere, I simplified and modularized the Yahoo! Pipes mashup.

I then used the pipe’s JSON output and styled it with some simple CSS. I didn’t want too much dependence on other JavaScript libraries so I used native code as much as possible, except for dynamically loading the JSON object from Pipes (for which I used the JSONscriptRequest class). I’m still having trouble getting the logic of reloading the content asynchronously: for now, I simply destroy the child element containing the tumblog items and rewrite everything — expensive, I know.

I’ve also been thinking of creating a Google Gadget out of this, but the iGoogle API’s FetchFeedAsJSON method is somewhat limited. I may have to use my current approach of dynamically creating a JSON script request into the gadget.

Well, without too much fanfare, the code is over here, and the tumblog is right here.

These recommendations were “generated by comparing (my) interests with the feeds of users similar to (me).”

I know, it’s supposed to be a tumblelog, but what the heck. I was able to squeeze out one for another domain I maintain: iandexter.com.

It was supposed to be for side projects, but since I’m neck-deep in work, I decided to make it a launch pad for my “presence” in the internet. But why stop there? So I decided to aggregate all that content, using Yahoo! Pipes:

Plunk the resulting RSS to Feedburner, and use BuzzBoost to put out HTML:

Then, voila! My very own tumblog.

Never browse Google Reader feeds at work when you’re subscribed to Fleshbot.