Josh Haberman

Looks like there’s another Gazelle in the herd. Microsoft Research has released a paper about a web browser called Gazelle:

In this paper, we introduce Gazelle, a secure web browser constructed as a multi-principal OS. Gazelle’s Browser Kernel is an operating system that exclusively manages resource protection and sharing across web site principals.

It’s easy to have my first reaction be annoyance. After all, I’ve been using the “Gazelle” name for over a year. For the same reason I was a bit annoyed when a BitTorrent tracker named Gazelle was released last year.

Then I remember that my “Gazelle” isn’t all that discoverable, even for someone looking to do due diligence about this. For reasons I don’t fully understand, my Gazelle is not returned very high in Google search listings (proof that working at Google isn’t a guarantee of good search listings). Here are my search rankings for various search phrases:

• gazelle software: result 83 (blog)
• gazelle project: result 86 (blog)
• gazelle program: result 89 (blog)
• gazelle system: result 4 (project homepage)
• gazelle download: result 34 (project homepage)
• gazelle parse: result 1 (github), result 2 (project homepage)
• gazelle parser: result 1 (project homepage)

Conclusion: unless you already know Gazelle is for parsing, or you think to call it a “system,” it’s pretty impossible to find it by its name. The reason I tried “gazelle system” is because the HTML title and headline on the homepage is: Gazelle: a system for building fast, reusable parsers. And that seemed to pay off — it has a much higher ranking for that phrase.

Guess I need to learn some search engine optimization. I should either use the words “project”, “software”, etc more on the homepage, or I should put them in a meta tag or something.

Not that Microsoft would have decided to change the name of their browser thing, but being discoverable is good in general.

In related news, I got the domain gazelle-parser.org, and replicated the homepage there, but I haven’t updated any links to point to it yet. Part of my vision for this site is that it will host a repository of known-good grammars, and let you parse snippets of text for these grammars through a web interface.