Finally, I get feedburner

So I have off for Martin Luther King day today. I found myself with a chunk of time, and decided once and for all, to make feedburner work for me. Happily, I can now say that I did that, and can move on.

You will notice the bright orange RSS logo on the side there… that now squishes my blog, del.icio.us links, and flickr images together for a nice big mess. Just the way I like it.

Feedburner is not hard, but I had no patience for it. I was under the assumption that you added your feeds, and there would be a grand final step where you could parse them all together. I kept looking around for this final step, but it turns out this is not the case. Instead, you choose one of your feeds, and then add little “enhancements” to pull in specific alternate feeds, such as del.icio.us, and flickr. Lame, compared to what I was thinking, but much faster than writing my own parser.

Of course, I’m sure everyone else out there understood this, so I’m being redundant. Oh well. That’s what you get when you try to learn everything from scratch.

Twitter inspires similar thoughts (when broken!)

First, I just watched what can only be described as a “herd” of squirrels, run across a neighbors front lawn. I’ve seen one chase another before, but I have never seen 5 all move together with a direction change. Very odd. Anyway…

A few days ago, I questioned in passing, when Twitter would start operating as a decentralized service. I asked because it was down at the time (outages are annoying, and I was bored). Apparently it was due to the MacWorld keynote address… so basically, Apple did it. Who cares, I was still bored. My friends were probably doing something amazing…

My two-second solution to the outages (and my current boredom) was decentralization of Twitter. And apparently, I was not the only one. Some questioned twitter’s usefulness as an emergency alert system. (My gut tells me that system wouldn’t work. In an emergency, most will grab their phones, not run to post a tweet. Also, there’s no way I’m paying $20 for “industrial twitter.” Anyway, I digress…)

There were also a few implementation ideas tossed around. I would implement based on a bittorrent model. You want your friends updates (call him friend B). You both have friend A in common, and friend A has friend B’s updates. There are now two places to obtain friend B’s updates. Voila. What about storage… what happens when you don’t want to hold all of my 1700 updates? Well, that’s where a few redundant twitter history servers pick up the slack, and algorithms figuring out when they start picking that slack.

[there are those squirrels again!]

Protected permissions would have to be dealt with, so give the client instructions to handle them. And I do mean client. Extend twitterific or a similar program, they would be fine for such a task [though I may leave phone-based clients out]. Why use a client instead of a dedicated server? One of twitter’s biggest draws is ease of use. Lets keep it that way… installing a server app is annoying to some, impossible to others.

I may have overlooked some flaws, but the basic model seems alright to me. A social network like this seems tailor-made for a distributed model.

So those are my two cents. Which I’m only spending so that I’m not bored when twitter is down.

what I've been up to

Though I doubt it will mean anything to anyone… I threw this diagram together Monday night. It’s my application design for a default NCBI application. I’m pretty excited about it. Not too fancy, but a nice start.

I’ve also been thinking about writing a Firefox extension. It’s about time that I play around with it, and I was thinking that a pretty simple app (like twitbin) would be a good place to start. No plans yet, but… ideas.

Speaking of ideas, when is twitter going to become open source and (then) de-centralized? That would be awesome…

Alright, now time to take the design and make it generic… woo hoo overloading!

Website usability annoyance no. 1

When you do a search, and you click on a deep page. The website makes you login, which you do begrudgingly. Where do they take you? To your home page. Do they even offer to send you to the page you were looking for? Almost never.

Come on, guys & girls. That’s just too easy…

Update to some random thoughts

1) I took a look at the book on Processing, and it does look quite nice. That being said, I’m not quite sure what I’d use it for. So it fell a bit on my list.

2) Letterman is still pretty much the same. I haven’t seen Leno. Jon Stewart, however, looked TOTALLY crazy last night. It was really quite scary.

3) Rob Coburn, you had better hang out with Raluca and I pretty darned soon….

4) Despite ordering the Python in a Nutshell book, I have semi-officially decided to put off learning Python for a bit. I think that I am going to re-focus my attentions on Scheme, and work my way through the Little Schemer, The Scheme Programming Language, and SICP. I’ve been through some of them, but haven’t gotten to the end of any of them.

5) The Giants won! Beat the Bucs, and are going to play the Cowboys next week. I hate the Cowboys, so this is going to be a big deal. I’ll probably try to watch it.

6) Steve, congrats on the potential new-old gig!