Facebook's Bunny1

http://svn.facebook.com/svnroot/bunny1/src/README

by khangtoh • 17 years ago

21 7 17 years ago

svn.facebook.com

EAMiller• 17 years ago

I've been using a little PHP hack for this for about a year and a half : http://ethanmiller.name/notes/php_quicklinks/

Easier setup if you use shared hosting...

hedgehog• 17 years ago

To try it: http://www.bunny1.org/

ivankirigin• 17 years ago

I love using it. For a group of people, it is a really simple efficiency booster.

rarestblog• 17 years ago

Facebook reinvents YubNub.org (circa 2005)

pyre2• 17 years ago

Two things:

1) Not everyone knows about every idea that is (or was ever) implemented on the web. I had never heard of yubnub.org before your post, and had I come up with the idea on my own I would have implemented it on my own like they did.

2) They have released the source for it; meaning that I can run my own local version.

pyre2• 17 years ago

It's a little late, but apparently they did know about yubnub. From bunny1.py:

  # these are three good choices for default fallbacks
  YUBNUB_URL = "http://yubnub.org/parser/parse?command="
  GOOGLE_SEARCH_URL = "http://www.google.com/search?q="
  GOOGLE_LUCKY_SEARCH_URL = "http://www.google.com/search?btnI&q="

ccheever• 17 years ago

I wrote a blog post a while back talking about the motivations behind writing bunny1 and the benefits of using it. http://www.ccheever.com/blog/?p=74

Some quick points:

@EAMiller: Cool. One note is that bunny1 doesn't need to run as a standalone server and works fine as a CGI. The instance on bunny1.org is actually running as a CGI. You do need to install some Python modules though, so your version is probably easier to setup for people who have a setup with PHP hosting already in place and not much else.

@rarestblog: bunny1 is inspired on YubNub and it actually falls back to it. Ex. If you do "php array_merge" in bunny1, it figures out it doesn't know what to do with "php" and redirects you to http://yubnub.org/parser/parse?command=php%20array_merge (or something close to that.)

The main reasons it made sense to write a new version and not just use YubNub for what I wanted to do are:

1. Bookmarks/commands written in a real programming language. A few of the most useful ones that are in use at Facebook are pretty complicated and do some things on the server side like talking to databases that would be a mess to try to do with YubNub's macros.

2. Bookmarks/commands version controlled using SCM.

3. Private bookmarks. I didn't want to pollute the YubNub namespace with tens or hundreds of Facebook internal URLs, and since the tool is to boost efficiency, its important that the most common commands be short. Also, even though everything internal at Facebook is behind a VPN, it would make a lot of people uncomfortable if sensitive internal pagess even had their URLs findable by looking through the list of YubNub commands.

Also, if you use Firefox > 90% of the time, check out Ubiquity. It's kind of a similar tool but way fancier. I prefer the portability and lighter weight of bunny1 myself but Ubiquity is more powerful and more user friendly in a lot of ways.

Kind of a funny coincidence... Someone posted this just a few hours after we set up an instance of bunny1 at Quora (where we've found it to be useful even though we're only a group of 4 and don't have a ton of pages to navigate to yet.)

Facebook's Bunny1

Related Stories