Interclue - Clever Link Previews

Interclue is a Firefox extension that adds more information about a link when you hover over it. It's not that annoying as the famous Snap tooltip and it certainly provides much more information about the page. As you can see in the screenshot above, you need to hover over a link, and then over the site's favicon that appears next to the link. Interclue shows a preview of the page that includes most of the actual content, a thumbnail, information about the last update, the number of words, links, images, a tag cloud with the most popular del.icio.us tags. But the most impressive thing is that all this data loads almost instantly and it's actually useful, especially when you want to judge the quality of a search result (including if it returns a server error). The preview window can be resized and dragged around, but it disappears when you move your mouse unless you lock it.



When the link points to a file (for example, a PDF), Interclue shows its file size and the last modified date, as reported by the web server. This is useful if you don't have a fast Internet connection and you don't want to wait for an hour to download Adam Bosworth's speech about health. Fortunately, the file is quite small.



The previews are pretty clever: for example, the preview for a link to a YouTube video shows the actual video. The extension "tries to avoid doing any look-ahead operations on [links that produce side-effects] by avoiding links with certain keywords in their text and link (such as 'logout', 'delete', 'remove' and so on), and also by turning off Linkclues for internal links on secure sites, which are typically used for banking, ordering goods and services".

You'll find a lot of things you can change in the settings and also the options to disable the previews for certain domains. Overall, the extension takes a great concept from Cooliris and transforms into a brilliant tool.

Google File Search

Web pages are useful, but if you've ever wanted to find a specific file on the web, you noticed it's not very easy. Fortunately, search engines like Google could be used for this tricky task.

Sometimes people create a web site, put some files in a directory, but forget to add an index file. So they end up with an unprotected directory that lists all of its files and subdirectories, when directly accessed from a browser. If someone links to the directory or submits it to Google, it becomes available to anyone who performs a search.

Because these directory listings are built using similar templates (depending on the web server), you can add to your query the most distinctive traits:

* The title starts with "index of" -> add to the Google query: intitle:"index of"

* They typically contain these words: "parent directory", name, "last modified", size, description -> you can add to your query "parent directory", for example

* Since most sites use Apache servers, you could also add Apache, that appears in the footer of a listing for Apache web servers


To find the page from the screenshot, you could use a query like:
intitle:"index of" firefox 2.0 rc1 source

Of course, you could use this idea to find any kind of file from a PDF e-book to an MP3 podcast or song. Some of the files are shared by breaking a copyright law, so you must you use your judgment before downloading them.

But finding files using this technique is too complicated, you'll say. First you have to enter a very complicated query, then visit all these strange-looking web pages and perform a new search in the current page to actually find the file. Then there are so many dead links and disingenuous webmasters that try to trick you with fake pages.

Some people with too much time on their hands built web apps that make it easy to search for files using Google. Briefli builds the query internally, loads the first results from Google and displays the links to the files on the same page. Moreover, the files that actually match your query are highlighted. To play the MP3s inline, you could add the del.icio.us bookmarklet to your browser and for Office files and PDFs, use Docufarm.



A site optimized for finding and playing MP3 files is mp3Salad. It lets you play all the MP3 files from a directory using a simple Flash player and even export the entire listing as a playlist.

The avalanche of file hosting sites brought a new to search for files: restrict the search results to one or more of these sites. Some examples of popular file hosting sites: esnips.com or megaupload.com. This custom search engine lets you restrict the search to 127 file hosting sites.

And then there are BitTorrent sites. Because they're so many, this custom search engine is useful to search across the most popular ones.

Google actually indexes some of these files, mostly Office documents, PDF files, text files. You can restrict a Google search to a file type by using the filetype: operator in your query (examples: bash linux filetype:pdf restrict the search for [bash linux] to PDF files). This way you can search inside these files and not only in a listing of filenames.

For files residing on your hard disk, a desktop search engine like Google Desktop (Windows/Mac/Linux), Windows Vista's search, Mac's Spotlight are great and should be used before searching on the web.

Maybe one day Google will come up with a nice file search engine that indexes unprotected directories, FTP servers, file hosting sites, torrent sites. But probably the legal challenges outweigh the advantages of a such a search engine (Yahoo has a music search engine, but only for China).