Saturday, December 29, 2007

Getting to know RSS

About 6 months ago, I started using RSS (rss/atom/feed - I'll use these terms interchangeably). As my RSS using habits become more advanced, I've been meaning to write down my experience of how I started, because I don't think my experience is particularly unique, and it might just push someone over the hump of getting to know RSS.

For a frame of reference, here's where I started: I knew about RSS for a long time before I used it at all. My thoughts were something like this: "I like going to web sites to see what's new. If I use an RSS reader, I may be able to efficiently see the content I want, but I'd miss out on seeing what else the site might have, like exploring outside links and such."

The problem with RSS is it's just so vague. RSS defined, is basically "It's an XML file, of site content." That means about nothing, to the people involved - internet readers, or site creators. On the publishing (site creator) end, a feed's content can take many forms: headlines only, full articles, intro sentences only, pictures included, pictures excluded. RSS doesn't say "translate your site into a feed this way". Site creators can publish their feed however they want. On the user end (the feed reader end), the experience is just as undefined (and therefore variable): list of healines - click to read; feed description shown with link to original post; the body of the original post shown instead of the feed description; a client program on a computer; an online client. These options make for wildly different experiences. Users, by choosing different feed readers, can "do" RSS however they want.

So one day I tried a couple of clients and subscribed to some feeds to see what would happen. There are many readers that technically "read rss files" and "display them to the user", but not as many take the whole experience or the real world into account. Let me skip the boring specifics, and say that Google Reader does things right. Don't even bother with the rest. (Bloglines.com works similarly, and fits the mold too, I should point out.) Most of the others aren't much more practical than using bookmarks/favorites in a web browser.

When I started subscribing to a few things in Google Reader, everything clicked for me. G.R. will tell me when a site has an update. As soon as I subscribe to site X's feed, I never again have to go to site X to find that there's nothing new there. Furthermore, G.R. will tell me /how many/ new (unread) things there are at each site. Furthermore, G.R. will do all this for me for 30 different sites simultaneously.

Where it shines best is sites that have infrequent updates (my magic threshold is one update a day or less). Have a friend who blogs once every 2 months? Subscribe to the feed, and you'll never have to remember to go look if there's something new. Plus, you'll know within an hour, when they do. A weekly column that posts on /about/ the same day every week? Don't bother reloading for a few days straight; subscribe to the feed and you'll just get notified when it's there. You get the picture. The more I used it, the more I found uses for it that I hadn't considered possible before:

- Someone made a retro-fitted feed for homestar runner (a flash comic that I would always have considered outside the scope of RSS).
- I can subscribe to my brother's Picasa Web Albums. Whenever he uploads new pictures, I know it.)
- Specific ebay searches have feeds. (I know it when a new Transformers [toy] prototype hits ebay.)

I did discover that you can go overboard. Particularly for sites that have a lot of updates. When I come back to Google Reader and see a site has 30 updates since I last looked (e.g. BBC news), I feel like I'm behind. I would either avoid subscribing to sites that have a lot of updates, or just keep them separate from the sites you actually want to keep up on. (Google Reader has organizing features that let you do that.)

That's about my whole experience. I've since introduced a few people to Google Reader (using a verbal version of the above), and they all use RSS habitually now. I'd like to hear if this post sways anyone else. In the mean time, at least it's written down.

Sunday, December 02, 2007

Let's listen to music together

Friends,
I'm back on my I Should Use My Skills To Make Music Fun To Listen To kick again. It's a bit of a misnomer, because this "kick" is nearly perpetual.

First, I daydreamed, I should make a site where I can recommend artists/albums/songs to others, in a time-ordered fashion, as in "here's what I'm listening to now, it's great". And everyone else could do the same, and in doing, I would discover new music through friends, socially, conversationally. That was "Phase I". Phase II, I thought, music will be fun again! My friends will say "Listen to this. It's great." And I will, and I will say back, "you know what? You're right. That's great!" Everyone will love it, everyone will use it, it will take over the world with all it's awesome features, and it will only take way more effort than I'm willing to spend on this project; Which brings me to Phase III, but first I have to explain this hypothetical site's biggest un-feature. There's no music sharing going on, just music recommendations. The presumption is, users can easily listen to whatever recommended music on their own, and finding a given artist/song is not a burden. It doesn't matter how. (I happen to through Yahoo, which I have to recommend honestly; it's $9/mo or $72/yr. But this is of course, not the only way.)

So, Phase III. Really, I don't even know if this idea would even be neat. And I think there are already tools online we could use to try this out, and see what happens. Namely, del.icio.us (or similar), and Google reader (or similar). As I hear new artists/albums I like, I will add them into delicious, tagged as something like "music_recommend". My friends will subscribe to my music_recommend tag in Google reader, and they will see that I have recommended a few artists since last they looked. Of course, my friends will have feeds that I will subscribe to, and we will be merry together.

The magic of rss has a pleasant side effect. If I get lazy (which I will), and don't recommend anything in 6 months straight, my friends won't get angry at me for consuming their attention for nothing. They just won't get notified that I've recommended anything, because I haven't. The magic of delicous is two-faced. There's a spot for a description, which we can explain "I like this because..." (if we want), and I could see that being nice. But then, there's the problem of what we put in the url field. I don't know, but I also don't care. I'm just going to read the title (which should have the artist or album name) and I'll just look it up on my own. I don't need to click on any links. (But maybe to the band's official site, or something. I think it will have to be something unique, so delicious doesn't think its the same entry as some previous one.)

Anyway, if anyone thinks this would be neat, I'll need a volunteer, or 2, or more. It's kind of a tall order, with rss knowledge and nearly unlimited music access required. There aren't many people I know who even qualify, but you know who you are. If I get a volunteer, I promise to start such a feed, and update it. I also promise to subscribe to your feed, and listen to at least some of your recommendations. (Probably all of them, but this is an experiment; and I don't like making promises.) If this turns into something neat over time, perhaps I'll build an actual system around it.

Comments?

Disclaimer 1: as usual I haven't researched if my idea has been implemented by someone else. But perhaps even if it does, my offer still stands.

Update: this sub-optimal description of a good idea is being implemented right now.

Thursday, September 13, 2007

RIP Firefox extension

There are not many Firefox extensions I can't live without. (Until recently, there weren't any. There's something about extensions that make me feel like I'm messing with "how things are supposed to be".) But now there's one, and I've been using it for at least a few months. So I know it's here to stay. Let me recommend an extension called "Remove It Permanently". It rids your tired eyes of animated distractions while you're trying to read, in an elegant, non-obtrusive way. Something I've never seen another ad-blocker do well.

http://rip.mozdev.org/

Tuesday, September 11, 2007

Another Best Buy experience

Every time I go to Best Buy, it seems, I have another story to tell. Here it is, as best as I can remember.

I put down my hard drive, and my impulse-buy on-sale Nintendo DS game at the checkout counter. The cashier offers me some special rewards thing that I decline, then lets me know about the 1-year guarantee on the game. Just to make sure, I ask if it's for the hard drive. No, it's just for the DS game in case it goes bad, or "starts skipping". Then came the bump in the road.

Cashier: "What's your phone number?"
Me: "Sorry, I can't give that out."
Cashier: "Can you just make something up so I can type it in here?"
Me: pausing, giving funny look, "...All 5's?"

The cashier types in all 5's. At this point I'm kind of glad that everything's pleasant. I'm actually sort of fond of these ritual conversations. There's just a certain line I won't cross about giving out my personal information. (Related word to the wise: never say, "I'm not going to give you that"; say, "Sorry, I can't give you that." But that's a different story.) The cashier continues filling the on-screen form. "What's your name?" I give her my name, which she misspells. I partially correct her.

C: "What's your address"
Me: "I can't give that out... You need my address?"

She starts filling in a made up address. She keeps prompting me to make things up so she can type them, but I just keep looking stunned. I got my bearings enough to help her make up a zip code. The computer tries to match her nonsense to real addresses, and gives her a list of addresses it thinks are close.

C: "I guess I need a real address."
Me: still looking stunned.

She just selects one, and finishes. Then it says "invalid phone number". She makes up a better number, which the computer is finally happy with. Then, on the credit card swiper, it prompts me that the info is correct. It doesn't just ask if it's correct, though. It says, "Do you agree that the following information is correct:" above a bunch of made up crap. The button I need to click reads "I agree".

Me: "What will happen if I agree to this?"
C: "Nothing... nothing bad."

Then, I clicked it. I swiped my card, which I imagine has all my info embedded in it anyway, and went on my way. But I'll be back.

Thursday, April 19, 2007

Email spelling

Moved in part by the argument on the bottom of this page,

http://www-cs-faculty.stanford.edu/~knuth/email.html

I am dropping the hyphen when I write the word email. Goodbye, brave hyphen that was once in "email". You helped bring a culture into its modern day. I will remember you fondly.

Thursday, April 05, 2007

More Fun with Google Trends

I heart Google Trends. There's so much in there to find, if you have the right queries to tease out real trends. I was playing on Google Trends again today, and I thought I would share my results...

If the last three years has anything to say about it, gardening is on a slow and steady decline. (Sad; let's hope it's just misguided statistics!)


While we're on the topic of seasonal searches, check out this pattern for "calories". Holy new year's resolution!


This one I thought was odd. Searches for "database" and various competing database products all seem to ebb and flow together. It's as if people don't get excited about (or bored with) SQL Server, or MySQL. They all get excited and bored together. I can't figure this one out.


And finally, I've been following the popularity of OpenID (that is, this graph) for a while now. It's about time OpenID gets more searches than the DC Comics superhero, Green Lantern.