Archive for February, 2006

Upgrade Day

Tuesday, February 28th, 2006

Today I migrated from a PowerBook G4 to a brand spanking new MacBook Pro. I’ll post more later describing my experiences. First up, this

Thursday, February 23rd, 2006

David J Anderson, on Software Development vs Construction:

In other words, if you ask a software guy to design a means of transport from San Francisco to Hawaii, he won’t come up with a boat or a plane, he will put a bridge in a While loop

Coolest.Sidewalk.Art.Ever

Thursday, February 23rd, 2006

This is cool.

Thanks to Katherine Derbyshire for the pointer

Motivation

Wednesday, February 22nd, 2006

Driving to work today, I set my car computer to display the estimated instantaneous fuel economy. Then I modified my driving to keep the number as high as possible. I take it as a personal victory I discovered the display maxes out at 99 mpg (not bad for an ’95 Dodge Grand Caravan).

Sometimes I tried to drive to get to my destination as fast as possible. Sometimes I tried to take the shortest route possible. Sometimes I tried to make the fewest lane changes.

I like turning my daily activities into competitions. I wonder if the same approach can work for software development . . .

  • I’m trying to be the best team player today.
  • I’m trying to be the one with the least amount of knowledge about the system that is not shared with at least 2 others on the team.
  • I removed the most number of lines of code today.
  • Fewer people who work on my code ask me questions about it.

(My personal challenge is how to turn weight loss into a competition without triggering my gaming-the-system filter)

Testers of interest

Sunday, February 19th, 2006

I’m glad to see that James Bach is posting to his blog. If you are interested in learning how to be a better tester, I suggest you read James’ writings.

Someone working on hooking up Watir and Firefox

Thursday, February 16th, 2006

Alex Fritze, author of jssh, pointed Angrez Singh to me. Angrez is working on extending Watir to drive Firefox, and he’s using jssh for the control channel.

At first, Angrez just had general questions about building the extension version of jssh, and why his Firefox builds were called “DeerPark”.

As we exchanged more mail, I was able to help get him answers to more complex questions, one of which involved the “browse” button for the input type=”file” picker. He also sent me the code he’s written so far, and I’m anxious to try it out.

I see that he is still using AutoIt to send key events, which is a Windows-only tool. Maybe I can find someone who knows how to do that using jssh, so this solution will be totally cross-platform (win, mac, linux).

Right now, Firefox does not ship with a remote-control interface such as the one Watir uses to contol IE. Using jssh seems like the least amount of effort to get something good-enough up and running. Eventually, I hope to see, and maybe work on, adding such a thing to our production code.

In the meantime, I look forward to playing with Watir and Firefox.

Developer Tests and the Mozilla Codebase

Wednesday, February 15th, 2006

Are you a developer writing code-facing tests in the Mozilla Codebase? If so, please let me know. I’m looking for ways to make it easier and/or quicker to write such tests, and would love to learn about how others have addressed this issue.

I’m keeping a list of tests I find in the codebase, and I’m working on a jssh-based remote control library for use in other test harnesses. Please feel free to add to it.

Quotes from the peanut gallery

Wednesday, February 15th, 2006

Last night, while watching House, in a disappointed tone: “Those aren’t maggots, they’re mealworms!”

Slogging through the nominations list

Thursday, February 9th, 2006

We’ve had 3 triage meetings for Firefox 1.5.0.2 so far, and we’re looked at a bit more than half of the bugs that were nominated for this release.

We granted blocking status to some bugs because they had a reviewed patch that has been “baking” on the trunk long enough to expose regressions, and there have not been regressions.

We granted blocking status to some bugs that are really bad bugs, and we hope someone writes a patch soon.

We denied blocking status to some bugs that do not have patches, but if someone writes a patch we will reconsider taking it for the release.

We denied blocking status for some bugs where there are lots of dependencies and complex issues under discussion. If someone more familiar with such a bug can present a simplified case for including a fix in this release, we might include it.

We denied blocking status for some bugs where there has not been any discussion or evidence of any investigative or patching activity.

Second-Most Important Meeting Phrase

Monday, February 6th, 2006

“Do you need me for anything else in this meeting?”, followed (hopefully) by “Great. Contact me if anything comes up.”