The Day I Nearly Dumped Firefox

I remember well the moment when the beta version of Netscape Navigator 0.9 was released in October 1994. It was so clearly superior to the main Mosaic browser I was running at the time, that there was no question about using anything else...


I remember well the moment when the beta version of Netscape Navigator 0.9 was released in October 1994. It was so clearly superior to the main Mosaic browser I was running at the time, that there was no question about using anything else thereafter.

At first, Netscape Navigator steadily improved, and the Netscape home page become the centre of the fledgling Web – the first thing you checked every day before doing that wacky "surfing the Internet" thing. But then, two things happened.

First, Microsoft's Internet Explorer started to improve. This was based on the original Mosaic, and had been hastily licensed from Spyglass after Microsoft failed to appreciate the importance of the Internet and thus did not include a browser with Windows 95 as standard (to begin with, you had to buy an add-on called Windows Plus.)

At the same time, Netscape Navigator became increasingly bloated and unusable, and the company lost any sense of strategic direction. When it announced that it was open sourcing the Netscape Navigator code it was hard to see it as anything but a last, desperate move.

I used the new Mozilla for a while, alongside various other browsers, but things really changed when Firefox came along (as something called Phoenix). Even in its earliest form, it was evident that this code was full of promise, and I've been using it ever since.

But this weekend, after more than a decade with the Mozilla family, I was close to dumping Firefox.

It began, innocuously enough, with a minor upgrade to Firefox 4. Firefox 4.01 apparently dealt with a host of minor bugs, and so I allowed the system to install it without a thought. After all, had there been anything big, I reasoned, it would at least have been version 4.1; 4.01 was clear signal that the changes were minor and would not break anyting.


When, a little later, I cam to use a fairly obscure but (for me) indispensable add-on, it wasn't there. Checking the Add-ons Manager, I was informed that it had been disabled. To console me for the loss, Firefox kindly offered to delete the whole thing so that I wouldn't be troubled again.

My first thought was to try to undo the upgrade, but Synaptic would only let me go back to version 3.6.17 or 3.6.10. Since this would have been a pretty massive loss of bug-fixes and features, I was unwilling to go this route, not least because I was informed: "If you force a different version from the default one, errors in the dependency handling can occur", which hardly warmed me to the idea.

There was no other add-on for Firefox that would replace the missing functionality, but I knew that Google's Chromium had something similar. So I duly installed Chromium and then the relevant extension, and was rather good. In fact, it was actually better than the extension that Firefox had just disabled for me.

So, as the result of an extremely small, unforeseen glitch, I find myself a satisfied user of the Chromium browser. Despite my initial frustration, I still have Firefox on my machine, and it's still my main browser, but the experience has made me wonder how many other people hit these apparently small obstacles, and are driven to download Chrome or Chromium, say – and like it so much that they do switch? Could that explain the current rise of Chrome, and the gentle decline of Firefox's market share?

Obviously, Mozilla can't test every add-on when it upgrades Firefox. But perhaps there is something that can be done to the architecture so that this kind of thing simply doesn't happen for such minor upgrades. Alternatively, maybe there should be a roll-back feature so that you can always undo such upgrades when you find they have problematic consequences.

This issue leads me to wonder whether Google has solved this problem with Chrome. Since it has adopted a far more aggressive automatic upgrade approach, I would expect more problems with add-ons not working, but it's not something I've really heard about or researched.

So, what's your experience with Firefox and Chrome upgrades? Is one better than the other in this regard? And would you ever be tempted to switch when a key add-in no longer functions? Or have you already, perhaps?

Update: Thanks to the kind efforts of Justin Scott, who leads the add-ons team at Mozilla, my particular problem has been resolved. Since general issues were raised, I thought I'd post some details from Scott of what was going on:

the version you linked to uses a max compatibility of 4.0 instead of 4.0.*, so when you updated to 4.0.1 it was disabled. Additionally, it does not have automatic updates configured, so you were not automatically upgraded to the newer release. Both of these issues aren't present with add-ons obtained from our official gallery, so fortunately the vast majority of users shouldn't see issues like this and those that do are likely power users installing add-ons from untrusted sources.

This might be worth bearing in mind if you too have upgrade problems.

Follow me @glynmoody on Twitter or

Find your next job with computerworld UK jobs