Jeremy’s Picks

If you don't already know these, you need to.
ls searches for commands. ge restricts the search to commands on the Golden Eggs list.
man tells you what they do.
create makes a new command.

The first thing everybody wants to do is search the web. Fortunately, making searches easy is Yubnub's forte!
g and gfl google the web, using standard search and "I'm Feeling Lucky", respectively.
You don't even need to use g if you don't want to. Any search that can't be interpreted as a yubnub command is treated as a google search.
y, a9, and bgo search using Yahoo, A9, and Blingo, respectively.
Many language-specific Google searches also exist, e.g. gde or ges.
gok displays a google search side-by-side with common variants of your search.
scro scrapes the ads off the Google results page.

The second thing everyone wants to search for is images. Personally, I use gim, or sometimes xim, but some may prefer yim.
fl and flk search tags and titles/descriptions/tags on Flickr, respectively. flint searches by tags, but sorts by interestingness.
gimflint does gim and flint at the same time, much as gimyim does gim and yim at the same time.
>& finds you all the images on a given site.
gima is the advanced google image search, with lots of options.

Some Yubnub commands are used to create other commands:
url returns the URL a Yubnub command would have redirected you to.
% extracts tokens from a string, allowing you to parse your own argument string.
filterstring deletes tokens from a string, the opposite of %.
ifthen is the type of command that might lead to Yubnub becoming Turing complete.
random generates random numbers. randwd generates a random word.
scrape extracts the bit of text you want from a web page.
split, splith, and splitv display several webpages at the same time.
date prints a date (today by default) in whatever format you need.
delw deletes the specified text wherever it appears.
extractdomainname pulls the domain out of a URL.
ifempty checks for empty strings.
strlength returns the length of a string. (Duh, but what else can I say?)
strreplace replaces part of a string with something else.
ucase and lcase make text UPPER CASE or lower case, respectively.

Wikipedia rules! wp searches article titles, wiki does a full-text search. wikigo sends you directly to the first result (but you should check out >). wpde searches the German version.
a (Answers.com) searches several sources, including Wikipedia. amo is the version optimized for mobile browsers.
brain parses naturally worded questions and searches webpages for the answer.

! and the related commands !ls, !set, !bm, !unset, and !seta allow you to use aliases for Yubnub commands.
mash runs several Yubnub commands on one argument, opening them in different frames.
mo optimizes the output of other yubnub commands for mobile browsers.
> makes a Yubnub command work better than it did before. No, seriously.
multi executes a Yubnub command on multiple arguments.
\\ checks the spelling of a Yubnub command's arguments before executing it.
newwin opens the result of a Yubnub command in a new window.
sidebarload loads the result of a Yubnub command into your browser's sidebar.

imdb (the Internet Movie DataBase) is the place to go for information about movies, actors and actresses, directors, producers, and more.
m uses Yahoo to search for movies playing near a US zip code or Canadian postal code.

gma is a bookmark for GMail. sgma is the secure variant.
ym is a bookmark for Yahoo mail.
send sends an email!
gmails searches your GMail account (after you log in).

am (Amazon) is famous for books, but you can find lots of other stuff there too.
You can buy anything on ebay.
amr searches Amazon for customer reviews.
Google does it again with froogle, a price comparison search engine.

Many Yubnub commands are both powerful and unique. Many can be embedded in other commands. Here are some of the most popular:
simpsons fetches random quotes from the Simpsons.
spllw uses Google to check your spelling.
cmd opens a Yubnub command line in a frame along the bottom of your browser.
script runs some javascript in your browser. Use with caution!
fspell spells what you type with letter images retrieved from Flickr.
qq: returns the status of a QQ user. (QQ is an instant messenger.)
yubnublet can help you create bookmarklets in Yubnub.
speak speaks a single english word, courtesy of Merriam-Webster.
login attempts to retrieve a username and password for a website.
skype displays the status of a Skype user.
ascii renders the phrase you provide with a very large collection of ASCII fonts.
lipsum generates Lorem Ipsum text - meaningless latin-like filler.
gmt retrieves the local time at a location.
convert converts from one unit of measurement to another.

bl is a bookmark for the Bloglines RSS reader.
tec (technorati) seems to be the most popular way to search blogs.
gbs does a simple google blog search.
glog displays google search results side-by-side with google blog search results.
blub subscribes you to a blog via Bloglines.
bps finds blogs that link to a URL.
rssread uses Google's RSS reader to open an RSS feed.

gm (Google Maps). If you still think MapQuest is cool, you should really check this out.
gearth starts Google Earth (if you have it installed) and zooms you in to the address you provide.
gloc searches the yellow pages and puts the results on a map, although you can do the same thing with Google Maps.

d is an english dictionary meta-search command.
dict and mw also search english dictionaries.
rae searches a spanish dictionary, I think.
thes will find another way to say something.
sss finds italian synonyms and antonyms.

xe converts currencies at the current exchange rate.
stock will retrieve current information about a particular stock ticker symbol.

leo is a German <-> English word lookup.
gproxy translating from english to english can be used as a proxy :-)
guess takes a sample of text, and guesses what language it is.
autotr is Google's attempt to translate anything into English.
tr can translate some language pairs.

Want to know what other people are bookmarking? Del.icio.us is the place to go.
deli finds the links with the tags you specify.
dpop looks for recently popular links with the tags you specify.
dgdt is a bookmark for diggdot.us, a meta-meta-links site.

Sometimes you don't want the hassle of downloading the graphics.
mo optimizes the output of other yubnub commands for mobile browsers.
gmobile strips images, CSS, and javascript from a page.
lynx displays HTML source.

If you want to know the weather in a US zipcode, these commands will give it to you:
w (The Weather Channel)
wu (Weather Underground)

Finding bittorrents is a little tricky because there's no one definitive search.
ts (TorrentSpy) is currently the most popular torrent search command.
tor is a meta-searcher, but it also requires you to verify your humanity before you can use Yubnub.

gn grabs news from many sources, courtesy of Google. As usual, you can do language specific searches too, e.g. gnewse.
cnn (Cable News Network) searches the eponymous news site. (You must provide a search term.)

One step beyond bookmarks, these commands require that you know exactly what to type in, and are intollerant of misformated parameters.
php will retrieve the manual page for the PHP function you name. But you have to know the correct name.
paview is the canonical example of how Yubnub makes parameterized bookmarks obsolete. With a little work you can end up with paview3 instead.
The most common type of parameterized bookmark takes you to a particular user's page, if you know the username. E.g. lj.

cd2 searches Djangos and Spun for used CDs. Or you can do them separately with spuncd and djangos.
allmusic searches for music by artist name.
gmp3 searches for mp3s, courtesy of Google.
hit find the date when an artist or song made the UK top 40.

sudoku generates a sudoku puzzle for you to solve.
wih is a bookmark for the warishellgaming forums.
thott searches a World of Warcraft database.
omerta is a bookmark for the text RPG of the same name.

Just a few of the other interesting things you can search with Yubnub:
gig searches Google homepages, I think.
g-ism finds out what the web thinks of your search term.
moz searches for FireFox extensions.
dpts (Debian Package Tracking System) retrieves the complete status of any Debian package.
groups searches Google Groups, aka Usenet.
gl uses Google's Linux-specific search.
gs (Google Scholar) searches academic papers you can get online.

Find out who owns a particular domain name or IP address.
dd (Domain Dossier) provides the most complete information, but runs the slowest.
whois and who provide basic results, quickly.

Programmers are always looking up the details of a particular module or language feature, and Yubnub tends to attract programmers.
flex2c and flex2p find class and package information for Flex 2 from the Macromedia LiveDocs.
flex1c and flex1p find class and package information for Flex 1.5 from the Macromedia LiveDocs.
jd finds the man page for a Java class, courtesy of javadocs.org.

Here are some popular bookmark commands that don't fit into the other categories:
/. is a bookmark for slashdot.
%s is a bookmark for Yahoo. (WTF?)
bark is a bookmark for an mp3 of a small dog barking.
porn is a popular bookmark for a popular site for, you guessed it, porn.
6 is a bookmark for myspace.
gread is a bookmark for Google Reader.

revs does a reverse lookup on a US telephone number.
phone looks up a US or Canadian phone number by name or address.