As many people know, I stopped working at Mozilla in December of last year. The story deserves its own post, though I doubt it’s one I’ll ever write. This post, however, is more of a post-post-mortem, to write about what I’m doing now and where I’m headed.

In mid-February, I sold or disposed of 99% of everything I owned. What’s left of my physical possessions fits into three, fairly heavy carry-on sized suitcases and an over-packed messenger bag. Suffice to say, a year ago, I never saw myself making this type of move.

The beginning

I guess most of this started back in 2008.

That year, I woke up one morning and decided to visit Thailand for my yearly vacation. Most people I know take a couple two-week vacations every year. I prefer to take one long vacation, usually three to four weeks. Being able to truly decompress and relax is something wonderful. But, back to Thailand.

I knew people who had gone to Thailand and, for some reason, I felt the urge to head West. Five months later, I was there. I spent a month in Thailand, enjoying the beaches in the south and the mountains of the north, making friends with locals and expats alike. Coming back to work was the hardest thing I had ever done. While my job was more or less my life, something about the freedom of traveling clinched inside of me.

After returning, my mind was on leaving again. I talked a lot about walking around the world, stopping for a month or two at a time, whenever it felt “right”. Of course, doing so wasn’t something I was yet prepared for. But the seed was planted. I wanted to travel. I wanted to see the world.

In 2009, I got my chance to see more of the world. Traveling over 100,000 miles by air, and many more by car, train, and even boat, I hit four continents, and almost ten countries. More importantly, this lifestyle started to feel right.

The real beginning.

Personally, I’m not really one for superstition, astrology, or any of the related ideas. I find them to be patently false. On the rare occasions when fortune telling works, the resulting fortune seems far too vague for my taste.

Yet in 2004, I attended a party with a coworker that resulted in a startling glimpse of what my future would be. An hour or two into the party, a woman at leaned over to me and asked if I was a Libra. “Indeed, I am.” Unfortunately, she didn’t know the precise information about my particular birthday. A quick trip downstairs and back she was with a large book in hand. I’d later purchase this book just because of how accurate it was.

She read, with increasing accuracy, the details of who I was – who I am. The book described how I was born to travel, how I wasn’t ever satisfied in one place for too long. Sure, it’d be possible to settle down, as long as I could continue to travel while settled.

The next morning, I left for a planned two-week road trip across the US, over seven thousand miles of travel.

Today.

I’m writing this blog post at the Mohammed V airport in Casablanca, Morocco. This year, I’ve completed fifteen travel segments, thirteen by air, two by train. In a few minutes, I’ll be on my fourteenth flight.

What I realized was that I’m happiest on the road and in the air. I relate to movies like Up in the Air, not because I’m destined for a hollow life of travel, but because the freedom that travel brings is truly wonderful. While it wasn’t actually my intention to visit Morocco on this trip, because I did, it likely means I’ll visit six continents this year (curse you, Antarctica!), including two new-to-me ones.

So with that in mind, this blog is hopefully going to change a bit, to include interesting stories from my travels and hopefully some pictures as well.

After our kickoff meeting with Happy Cog, the design firm we selected, we’ve talked a couple more times and they’ve now started the design process. One of the things we all agreed on was that getting the first pass at a design out before SXSW would be perfect. Doing so gives us a long week of feedback from both members of the Mozilla community as well as members of the design community.

We’re now intending on presenting the new designs in about a week at the weekly design lunch (which you can watch this week and next on Air Mozilla). David will have more information as we get closer to next Thursday.

The first question we need help with is: What other sites out there do you feel does a great job from an experience point of view? Why?

There are lots of websites with great design out there. Hundreds and thousands. But which sites do you think have a great user experience? What sites should we draw on for inspiration with the redesign?

The second question: What words (category labels) do you feel would be meaningful to your audience?

The Mozilla community is huge. There are so many facets and so many different things we do. What words describe Mozilla in general? This can be a long list, but I think three to five words would be ideal. What’s the most meaningful set of words that describe Mozilla?

Besides those two questions, there’s a bunch of other questions that have already been answered and we’re looking for feedback on. Feel free to browse through the questionnaire and provide feedback on any of the questions and answers. Preferably, add your comments to the “Feedback” section of the questionnaire, but feel free to add them to this blog post, send them to me (or David) in email, or ping me (or David) on IRC.

I plan to post about my month long frolic in Thailand soon enough – a month’s worth of photos take a while to sort through – but this seemed so incredibly strange that I had to blog it.

A couple days ago, AOL released AIM for Mac Beta 1 (henceforth called “AIM”), a completely rewritten version of AIM for Mac after letting the the original AIM for Mac languish for years.

I download AIM and was greeted with a package installer. Now, for an app like AIM, this is completely unnecessary. AOL: Just give me a disk image to open and copy the app out of. I’ve seen the negative press before from .pkg files; specifically, some vendors use them to install random files throughout your system even when they don’t have to. As a precaution, I investigate them now. Much to my surprise I saw the word “Talkback” in one of the files.

For those who don’t know, Talkback is the name of the crash reporting system Mozilla uses for applications based on the Gecko 1.8 branch (and before) including Camino 1.6.x, Firefox 2.0.0.x, and Thunderbird 2.0.0.x. It’s an old, fragile system – both the client and the server – acquired from the Netscape days. Talkback also happens to be a group of four systems I manage at Mozilla. When I said it’s fragile, I mean it. The server often falls over and we scramble to fix it. The client looks fairly ugly and doesn’t provide the features we need (such as compatibility on Intel Macs).

AOL couldn’t be using Talkback. Not that Talkback.

Of course, seeing Talkback anywhere forced me to install AIM and survey the damage.

AIM uses quite a bit of open source software. The quick list includes:

• MOKit
• Pantomime
• PSMTabBarControl
• NSPR
• NSS

The last two are likely familiar to Mozillians.

Overall, it’s a pretty happy list, though MOKit and Pantomime seem to be older, less maintained projects (side note: anyone know of a replacement for Pantomime?). But I’ll be darned if AIM doesn’t include Talkback, right alongside the open source software.

Talkback in AIM is different than Talkback in Mozilla apps. For one, AIM uses version 2.0b4, Mozilla uses version 2.0b1. For two, the size of the Talkback application in AIM is 888 KB to Mozilla’s 480 KB. That, of course, is likely a consequence of AIM’s Talkback being a Universal application. Someone has gone through the trouble of updating Talkback specifically for AIM. Why in the world they would do this, I have no idea. Especially when a better alternative exists.

AOL did an “okay” thing by updating AIM in spite of iChat being the de-facto AIM client on Mac. They just completely ruined it in the end by incorporating a crash reporting system that is clearly ready to break down any day now.

See you on the other side!

The Talkback database is now around 500 GB, which is way too large to be manageable. As a result, the scripts that removed old data are unable to complete and, thus, unable to lower the size of the database. There are several ways we can “fix” the issue:

1. Let the size of the database grow infinitely until Talkback is replaced.
2. Manually run clean up commands to (hopefully) keep the database in good shape.
3. Remove a bunch of data to all the scripts to run and move to keeping 60 days worth of data instead of 90.

There are advantages and disadvantages of all of these, but after some discussion, we’ve decided to do #3. Removing data will allow us to have a manageable database and will allow us to keep the database in shape for the long term by lowering the amount of data we keep.

Sadly, as a result of this, we’re going to lose some (critical) data. Namely, remove this data will affect full stacks, topcrash reports, and smart analysis reports.

Tomorrow we plan to take Talkback down for a few hours and do a full cold backup of the database. After the backup, next week we plan to remove full stacks from existing crash reports. If you have a stack you want to save, please put it in the relevant bug.

New reports will, of course, generate full stacks, but all stacks from old crash reports (about 90 days worth, give or take) will be gone. Because full stacks will be gone, the main topcrash reports will not necessarily be correct and the smart analysis reports will be completely broken (they rely on full stacks to generate properly). This will last for a period of 10 days, then both reports will return to normal.

Note: The removal of full stacks does not affect the stack signature. The signature will remain.

We know this situation isn’t ideal, but it’s just yet another reason to convince your friends to upgrade to Firefox 3.

If you have comments or questions, please email me.

Scott and David: Your work on Thunderbird over the past years is most appreciated by your millions of users, myself included. I wish you both good luck in your new endeavors.

JT: Oh, JT… you who brought me so much work in my first couple of months… As hard of a time as I give you, you will most definitely be missed by all of us here (except me, of course). Good luck at your new job.

Paul: While you’re not leaving ’til next week, I just wanted to say that it’s been great working with you. Coffee breaks, housewarming parties, nights in the city, buying iPhones, etc etc. Since I started, you’ve been a big part of my experience here at Mozilla and I look forward to seeing you outside of the office more frequently.

To the four of you, I say farewell and wish you the best. So long, and thanks for all the memories. I’m sure I’ll be seeing you around.

I’ve thought about this a lot. There are numerous unconfirmed bugs that are legitimate bugs or enhancement requests that go unnoticed. Additionally, most bugs that are still unconfirmed are likely duplicates of other bugs, invalid, or, after the amount of time that has passed, “works for me”. So, is it really a problem to have a few thousand bugs that no one’s looked at? Absolutely.

If legitimate bugs exist, we need to triage them and get them in the hands of developers. Bugs that aren’t legitimate should be closed. I’m much opposed to the “expired” idea because there are so many bugs that could get lost in the shuffle.

So, after talking it over with Tomcat, we’ve started setting minor and major personal goals regarding lowering the number of unconfirmed bugs. One of our first goals was getting the number of unconfirmed Firefox bugs under 7000. Now that that’s done (happened two days ago; check out the graph comparing unconfirmed bugs with assigned bug to get a glimpse of how out of control it was), we’ve set our targets on the total unconfirmed bugs.

On May 28, 2007 (just four days ago), we hit a post-EXPIRED peak of 16,693 total unconfirmed bugs. After talking it over with Tomcat, our goal for the month of June is to get that number under 16,000. Pretty easy goal, really, given that we’re already down to 16,571.

The reason I’m blogging about this is because we need your help. If we’re ever going to lower the number of unconfirmed bugs to something reasonable, we need participates from all over the spectrum. Lately, a few developers have spent time cleaning out their Firefox components