A scathing review of Firefox 11 finally pushed me to an experiment I should have done a long time ago — using Google Chrome as my default browser.

I’m not an early adopter. I may hear about something the early adopters will take up, but I let them vet it out before I take a stab at it. In the case of Google Chrome, I did actually try it out when it was first released, and … it failed. I had Japanese text on my website that would not render at all.

And it was a pretty specific situation too. The Japanese text would disappear when I list Helvetica as the first font for my style sheets to render, and it would happen only on Windows XP machines. As it turned out, the version of Helvetica installed on my computer was corrupt, and it didn’t affect just Chrome but anything using Webkit, including Safari and iTunes.

But the impression was made — I can’t use Chrome if it can’t render Japanese on my own websites. Then there was the matter of add-ons. When I tried out Chrome, AdBlock Plus and Rikaichan hadn’t yet been ported, and I depend on those add-ons.

I would eventually fire up Chrome for cross-browser testing and to use Google Apps at my last job — Firefox was having issues with GMail at some point — but, like Internet Explorer 6 users who couldn’t see the value of upgrading, I didn’t see the value of switching.

Those first impressions from when I gave Chrome a shot persisted. Yes, now that I’ve upgraded my operating system (and got rid of that errant Helvetica font), Japanese fonts appear on my web pages, and most of the add-ons I use from day to day have been ported to Chrome.¹ But I gave Chrome a number of chances, and those failures — or the appearance of failures — didn’t endear me to it.

There’s nothing standing in my way now of using Chrome as my default browser, which I’m doing now as an experiment.

I didn’t realize it myself, but I have been grumbling about Firefox for a while. When it loads my last session, it always freezes up trying to render my Twitter stream. And when I visit my own jQuery-heavy sites, I can see Firefox drawing the page before displaying it properly. Even Internet Explorer 9 (on 64-bit, that is) doesn’t draw that slowly.

While I am impressed with Chrome’s speed so far, that hesitation remains. Will Chrome get in the way of my user experience again? And how bad will it be the next time? This experiment attempts to addresses these questions.

¹I would like to see a Foxytunes port, but YAHOO! has enough trouble keeping the thing updated for Firefox.

I want to buy a MacBook Pro laptop.

Anyone who knows me will probably be taken aback by that sentence. I’m a Windows user and have been since 1995. But don’t mistake my adherence as advocacy.

When it comes to tools, I use what’s available. Hopefully, it’s the right tool designed for the job. You wouldn’t want to use a Philips screwdriver for something that needs a Thomas wrench, but if scissors aren’t available, a fingernail might do the trick.

One morning, I tweeted a complaint about iTunes on Windows, to which a friend of mine replied on Facebook with a question: “Why are you still using a PC?” I have standard answer to that question: I’m not in the income bracket for Apple products.

Yes, I’ve saddle myself with the Windows platform out of economics, not out of brand loyalty or platform advocacy. No, I don’t love Windows, and right now, I’m going through installation hell, having moved from Windows 7 32-bit to 64-bit. I didn’t work extensively with Apple computers till I got a web software engineering job at a Mac shop in 2010.

Till then, the last OS with which I worked on a Mac was System 7. Uh-huh. The late ’90s right around the time Steve Jobs came back into the fold. And boy, did those machines really suck. Back then, I was baffled by the Apple fan boys who insisted on the supremacy of the Mac operating system when it seemed neither Windows nor Mac worked very well. I remember Macs running System 7 crashing as much, if not more, than Windows 98.

A decade later, my previous perceptions were wiped away by the stability and ease of MacOS X, and I began to understand the accolades heaped on Apple. I have to admit — I was impressed. Impressed enough to decide that yes, I would welcome a Mac in home.

But I’m not so enamored that I would completely remake my technology stack to be exclusive to Apple. Not at this point. (That might change if/when I actually do get a Mac through the door. Price has always been a barrier of entry for me with Apple.)

In the 18 months I was working on a Mac in the office and Windows at home, I didn’t feel too much of a cognitive dissonance. Actually, working on a Mac convinced me to upgrade to Windows 7, and I found the transition from Windows XP to Windows 7 a lot easier because of it.

The truth is, I’m a mercenary when it comes to my tools. Could I make better music on a Mac than a PC? It might be easier, but I think the songs would sound the same given my level of skill with audio engineering. What about my work on the web? Most of the software I use to build websites are cross-platform, and I found no huge differences between user experiences.

In other words, it’s not the tool, it’s the person who’s using it. Yes, something could be said about tools getting in the way of the person using them — as Windows tends to do when it breaks down — but when they work, I’m usually too concentrated on the task at hand to notice the system running them. Which is how it should be.

Except when it comes to the COMMAND and CONTROL keys. Switching between those keystrokes is just messed up.

 

I’ve been saddled with Movable Type since the start of the 2000s mainly because I coupled an external database with Movable Type a bit too tightly. It was a design decision that snowballed into a monstrosity from which I could not easily extricate myself.

Recently, I started a blog on WordPress.com and discovered I liked the responsiveness of the WordPress UI more than Movable Type. So I started to experiment with WordPress on my own server. The Network feature in 3.x appealed to me because I like being able to use one interface for multiple blogs.

(Side note: I had actually tried out WordPress a few years back, but the lack of support for multiple sites kept me in the Movable Type fold.)

I’ve been impressed so far, and I feel like I want to learn more about WordPress development. So I wanted to mirror the WordPress network I created for a development environment.

And that’s when the wheels came off.

Round and round: Redirection loop

WordPress does some funky things with redirection that I’ve not had time yet to reverse engineer. I just discovered that simply copying the file system and loading a database dump into a new database is not enough. The WordPress Codex site has documentation on creating a development mirror, but it only really works for single-blog installations. When you try to enable the WordPress Network configuration values in wp-conifig.php, you get a redirection loop.

On an unrelated task, I wanted to find out how to change the domain of the primary blog in a WordPress Network. A Google search led me to these instructions, which actually are a bit more heavy-handed than they need to be but contain all the right information to get it done. I noticed as I tried to get my domain changed that I was running into what seemed to be the same redirection loop.

Perhaps my trouble with creating a development mirror is related to changing the domain of the primary blog in the network? As a matter of fact, it was.

Change your primary blog’s domain in three queries

Despite what the instructions say, you do not need to replace every instance of your domain in your database dump. In fact, you can change the domain of the primary blog with three MySQL statements:

• update wp_site set domain = 'newdomain.com';
• update wp_options set option_value = 'http://newdomain.com' where option_name in ('home', 'siteurl');
• update wp_blogs set domain = 'http://newdomain.com' where id = 'blog_id';

Of course, substitute newdomain.com in those examples with your own domain name. blog_id is the numeric id of your primary blog in the database, which is most likely “1”. NOTE: The table names in the example may not reflect the actual table names in your database. But some form of wp_ and site, options and blogs should be in the names.

You will need to change the DOMAIN_CURRENT_SITE constant in wp-config.php to your new domain name as well.

define('DOMAIN_CURRENT_SITE', 'newdomain.com');

These steps assume you’ve already set up a domain separate from your production site, and you’re pointing that domain to a development copy of your WordPress file system. They also assume you have access to a MySQL command-line interface.

In short, creating a development mirror of a WordPress Network involves changing the domain of your primary blog to prevent a redirection loop.

 

In his book Listen to This, Alex Ross profiles Marlboro Music, a summer retreat for classical musicians, with a particular focus on one of the festival’s directors, the pianist Mitsuko Uchida. In one anecdote, Uchida attempts to convince Romanian pianist Radu Lupu to visit Marlboro:

“I got very excited, describing how people do nothing but play music all day long. But he said no. His explanation was very funny. ‘Mitsuko,’ he said. ‘I don’t like music as much as you.'”

That sentiment sums up my relationship with software development.

Of course, Lupu has some level of dedication to his art, if winning the Van Cliburn competition in 1966 is any indication. At the same time, it’s refreshing to hear a performer of Lupu’s caliber admit to his limits, if only in jest.

I earn a living working with web sites and databases. I like it. The work day passes quickly when I’m head down in a project, turning an idea into a tangible product. Part of it is the wonder of getting something to work. The rest is ego in not letting that stupid bug or that idiotic configuration defeat me.

But there comes a point where the proverbial wall smashes into my face. HTML5? Node.js? Ruby on Rails? Hey, great! Wonderful! I’ll get right on that, but first, there’s this other project I’m working on that suits my state of mind right now. So I’ll get back to you.

There’s a lot of punditry out in the land of developer blogs about what makes a “good” developer/engineer/programmer/whatever. A lot of the talking points are the same — gotta be curious, gotta code outside of work, gotta have passion, gotta not treat it like some 9-to-5 job. If that’s the case, I was a damn good developer in the early 2000s, but I’m nowhere as good now.

Sorry, software engineering is not how I choose to center my life. It’s certainly how I choose to center my professional career, and I take those skills to other things I do. But I don’t like programming as much as you, and I don’t think that makes me any worse off.

Sure, I’ll code outside of work in moderation. I’m curious to a point, and my passion has its limits. But my 9-to-5 job is exactly that, and my “life” portion of my work/life balance contains more things than just code.

If anything has sustained my career for as long as it has, it’s a principle that drives any work I’ve done, be it music or journalism or development. Rands puts it succinctly and in italics: “We must not ship crap.”

Live by that and everything else falls into place.