Mozilla Prism

» see prism.mozillalabs.com as it lived, on the wayback machine

~ the last words ~

Mozilla Labs shipped a great many experiments between 2007 and 2011, and Prism was one of the few whose point I understood before reading the second paragraph of the announcement. Firefox was my browser through that entire window. The problem it addressed was one every heavy Firefox user of the period had: your mail lived in a tab, that tab lived in a window with thirty other tabs, and the thing you needed constantly was the hardest thing on the screen to find.

Prism's answer was to take one site out of the browser entirely. You pointed it at a URL and it produced a shortcut that opened that site in its own window, with no tab strip and no address bar and nothing else. Gmail became an application in the taskbar. The category name for this was site-specific browser, which is an accurate term that nobody outside Mozilla ever said out loud. Under the surface it was still Firefox, running on XULRunner, so the rendering was identical to the browser you already had.

It started as a prototype called WebRunner that Benjamin Smedberg demonstrated at the XTech conference in 2006. Mark Finkle picked it up, kept working on it, and Mozilla Labs relaunched it under the Prism name in october 2007. That was a good moment for the idea. Google had shipped Docs and Spreadsheets the previous year, Gmail was becoming the default mail client for a generation, and the argument that the browser was the new operating system was being made loudly by people who had not yet worked out what that would do to window management.

~ the record ~

~ how it landed ~

It worked, which was never really in question, and it was useful, which was slightly more surprising. Reviewers liked that a Prism window used less memory than a full browser and that a crash in one web app could not take your other forty tabs with it. On the Mac there was a competing tool called Fluid doing the same job with a nicer installer, and the two of them briefly made site-specific browsers look like a category rather than a curiosity.

Adoption never got past the enthusiast layer. Part of that was distribution: Prism was a separate download, and a separate download that makes a shortcut is a hard sell to someone who can already bookmark a page. The larger part was that the desktop metaphor it was reaching for arrived from somewhere else. Chrome added "Create application shortcuts" in 2008, buried in a menu, requiring nothing to be installed, and that quietly settled the question of who would own this feature.

~ the actual end ~

The last release was version 1.0b4 on 5 may 2010, and by that november the project was formally listed as inactive. On 1 february 2011 Mozilla Labs announced it was done maintaining Prism, saying the ideas had been absorbed into a newer project called Chromeless. The Labs mailing list then had to clarify that Chromeless was not in fact a replacement for Prism, because Chromeless was a toolkit for developers building browser interfaces rather than a thing a normal person could point at Gmail. The successor did not succeed anything.

There was a small afterlife. The code went to a company called Salsita, got its original WebRunner name back in 2011, and was cancelled there too around the time Firefox 5 shipped. What is strange about the whole arc is that Prism lost the argument and the idea won it completely. Every Progressive Web App installed straight from a browser, and every Electron client that is really a website in a frame, is the thing Prism was demonstrating in 2007. It just needed the browser vendors to build it in rather than hand it out as a download.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »