Firefox 1.5 was released along with many changes to back-end services many people outside the developer community aren’t aware of. So instead of blogging about changes in this release I thought I’d take some time to stop and take a look at what went on with web services that support it.
Safe to say, without the great work of the Mozilla sysadmin team the release wouldn’t have gone very well. polvi and justdave worked very hard throughout the release to make sure everything stayed afloat — and I think we all owe them a debt of gratitude for their excellent work.
Seeing the overwhelming traffic to addons.mozilla.org (AMO) was pretty cool — until it started to bring down the application. 🙂 But not to worry, Dave and Alex were able to add more LVS nodes to the mix and the site is running well at this point.
On the marketing side of things, Rebron and Beltzner managed to work out the new look for AMO. They did a great job organizing content geared towards the correct audience, and I think the new look and content really hits the spot as far as addressing our main userbase — non-technical users who want to get things done!
We did manage, however, to hit MozDev pretty hard (just overall, not just from AMO). Thanks to shaver for helping us fix the search-engine page and relieve some of the pressure on MozDev.
On the AUS side of things, Chase has continued his dominance over the impossible by pushing out the new builds using the now cumbersome Bouncer 1.0 interface (one-at-a-time — more on this later) while managing to fix AUS problems caused by the disk issue on stage in the last week of November. I think we all owe Chase a bottle of port, or maybe at least a Guinness and a pat on the back.
Bouncer handled things pretty well during the release. So well, in fact, that Netcraft wrote an article on its performance. I thought that was pretty cool.
That said, Bouncer still needs a lot of love. Lars and I finally hashed out some of the blockers for Bouncer 2.0 and we should be able to help Chase out by providing the upload utility for adding multiple builds via a sum file. It’ll save him a lot of time before releases.
We also plan on releasing Bouncer (finally) to the public and opening it up for improvements. A community site is also in the works, once we have time to do it (weekend project!).
One of the cooler things I saw was Lars writing a conversion script to sync up the Bouncer 2.0 dev database with the Bouncer 1.0 production database in about 20 minutes. Lars, you’re a madman.
It was an eventful week. The release was a success, and I was reminded of a few things that makes it all worth it:
- A focused community can accomplish so much in a short period of time
- Firefox is huge
- There are so many more opportunities to make things even better
So thanks to everyone who played a part in this release. It was once again a tremendous experience.
The whole is greater than the sum of its lesser parts.