Twitter friends again!

Jul 15th 2008 at 5:30 PM

Update 12/2009: Mentions and retweets are excluded from the friends_timeline.rss. Nowadays, it's better to join the mentions.rss feed with the home_timeline.rss.

Remember how angry I was at twitter? They turned off open friends feeds subscription since the WWDC?

It turns out the solution I needed was a one-liner cron job.

curl -u email:password http://twitter.com/statuses/friends_timeline.rss > twitter_friends.rss

::slaps head::

Now I can use Google Reader again. Thanks, Twitter API. We can be friends again.

Tags: code, web

Comments

davidd on Jul 16th 2008 at 1:17 AM

I really wish I had some clue as to what all this stuff means. Okay, I know it's "code" but, like, what do you do with the line of code? Where do you even go to paste this in to make things work? Rhetorical question, really. I know it's not something you can explain in one or two Plurk-length sentences. What's the best way to start learning this coding stuff? I mean, there's so much, so many kinds... none of which are BASIC or COBOL. Heck, I can't even get a WordPress site to look right, let alone do groovy stuff. No, I'm not looking to learn coding, not at this time in my life. My point, rather, in case you haven't caught on, is that I'm really impressed by the "smart people" like you and Nari and Zannah and Sjon who know how to do this stuff. It's a skill, more of a talent, really, which to me appears so arcane as to border on magic. I think it's really cool what you guys can do! Enough fanboying. ________ I think I google the terms "oneechanbara" "dvd" almost every day now. Nothin' yet. ________ Multiple edits to fix redundancies and typos. Sorry!

zannah on Jul 31st 2008 at 6:44 PM

How often does that have to run? Once a day?

dblume on Jul 31st 2008 at 7:00 PM

I run it hourly. Remember that I'm dealing with two delays. The rate at which my cron job runs, and the rate at which Google Reader checks the public feed I created. (You may not have to deal with worrying about the second one.) Here are my best case and worst case scenarios: Best Case A friend tweets at 12:59 My cron job picks it up at 13:00 and writes it out. Google check the public feed at 13:01, and I see it! A delay of a minute or two. Yay! Worst Case A friend tweets at 13:01, just after my job checked. My cron job checks at 14:00 and writes it out. Google miss by minute or two, and gets it next time, at 15:01. A delay of a couple of hours. Drat. For an enterprise application, you'd want to check more frequently with conditional GETs and honor any HTTP 304 replies, probably.

dblume on Aug 5th 2008 at 10:29 PM

You can accomplish the same with without needing a cron job by using Yahoo Pipes, too. Here's how.

(no name) on Jan 21st 2009 at 9:45 PM

Doh. That *is* simple. Thanks.

(no name) on Jun 14th 2009 at 11:31 AM

It seems a good idea when you are on your own Linux box but what if you are away from your pc???

(no name) on Aug 24th 2009 at 12:42 PM

Quote: "Worst Case A friend tweets at 13:01, just after my job checked. My cron job checks at 14:00 and writes it out. Google miss by minute or two, and gets it next time, at 15:01." Sorry if there's some reason why this wouldn't work- I'm a student and hobbyist coder but I'm not familiar with the twitter API. Couldn't you just run your cron job at every XX:50, so that google reader can't miss it? Then your worst case is only a 1 hour delay.

dblume on Aug 24th 2009 at 8:27 PM

I have no control over when my friend tweets, nor when Google Reader gets around to reading my feed. If I change my cron job to work at XX:50, the odds are that Google Reader will start reading my feed just before I update it, at XX:49, ensuring the worst case is still the worst case.

← The Unofficial Plurk API in Python

So Tweet It Is →

Backup of David's Livejournal

Recent Entries

Navigation Tips

Twitter friends again!

Comments

Tags