Posts tagged: MACOS

Locomotive: Rails for OS X

This is nifty – Locomotive, from Ryan Raaum, a complete Ruby on Rails environment in a self-contained 30MB bundle. And when I say complete, I mean complete: Locomotive contains not only Rails itself, but the Ruby interpreter, RubyGems, the LightTPD webserver with FastCGI, the SQLite database engine, bindings for MySQL and PostgreSQL (though not the server binaries, wisely), and all the other bits and pieces needed for turnkey Rails. There’s also an expanded version of the package with even more goodies. If you have an existing Rails installation, Locomotive will run politely alongside it without messing anything up.

Google wants Mac developers

It’s been a sore point among Macintosh users that almost all of Google’s desktop software is currently Windows only. A couple days ago I learned that Google had started a search for Macintosh developers. As of today, listings for Senior Macintosh Developer (8+ years experience) and Macintosh Developer (3+ years experience) are on Google Jobs.

Only Google can say what they’re up to, but note that Google Earth, Google Desktop, Picasa, and Google Talk are all mentioned in the ads.

AntiRSI

All the keyboarding I do, on top of motorcycle commuting, means that my wrists work very hard. For the past couple weeks I’ve been experimenting with a program called AntiRSI which tries to keep you out of trouble by recommending short “micro pauses” and longer “work breaks” – based on how much continuous keyboard/mouse work you have actually been doing, not just on the clock.

If your rhythm is such that you are naturally taking breaks from the keyboard, AntiRSI stays out of your way. But when the pace increases, it will pop up with a reminder. It’s the ten-second “micro pauses” that I think are saving me. One nice design detail is that these notifications aren’t modal and don’t get in the way of your typing. So if you want to spend a minute or two deferring that “micro pause” while you finish a thought, go right ahead – AntiRSI waits patiently, but doesn’t remove the notification until you actually take that ten-second pause. When I get there, I take the opportunity to do some Bob Anderson style wrist stretches.

Ctrl-T considered harmful

Here’s an odd bug in OS X’s Mail.app: if the cursor is either at the beginning or the end of a message you’re composing, and you press the keyboard shortcut for “Transpose characters” (ctrl-T) the application spontaneously quits.

Bummer!

Caution – don’t idly test this out right now if you have an unsaved message open. I know it’s tempting.

MacFixit has a suggested workaround – use ~/Library/DefaultKeyBindings.dict to disable the key – but it’s not ideal since ctrl-T has valid uses in other applications, like invoking spellcheck in Pico or Nano. Unless you tend to hit ctrl-T accidentally, I’d just leave it alone and wait for the inevitable patch from Apple.

DarwinPorts 1.1

There are almost as many unix software packaging systems as there are flavors of unix – Debian’s APT, FreeBSD’s ports, Red Hat’s RPM, Gentoo’s portage, et al. Under OS X or Darwin, the two main contenders are DarwinPorts and Fink. I used Fink for a long time, but switched to DarwinPorts last year in one of my periodic retoolings, and found I liked it better. My reasons are intangible; my gut tells me that only one of these systems is going to be “the one,” and that it’s going to be DP. Mostly that just means I like the feel of it. It has fewer packages than Fink (2800 vs. about 5000), but seems to be gaining.

Gruber on AppleScript

I’ve always wanted to like AppleScript more. Even though I’ve occasionally been paid to write AppleScript code, and always have a few little snippets gluing parts of my work environment together, I’ve never really gained facility with it. I’ve always summarized its flaws this way: It’s very easy to read for general sense, but damn hard to write. Sort of the inverse of Perl – which, probably not coincidentally, is the world’s most popular glue language. So while this characteristic of AppleScript may have been a conscious decision on the part of its designers, I’d argue it hasn’t served us all that well. (I’d also argue that Automator is a partial acknowledgement of those failings.) John Gruber of Daring Fireball sums it up nicely in his recent posting: