Twitch

Update: a few corrections due to feedback and a minimal update on my tinkering with QSB.

I have been putting off doing this for a while given that application launchers are sensitive affairs for a number of reasons:

  • They are intensely personal choices
  • They (literally) attach themselves to our cerebellum and beyond through their repeated use, tricking our motor control centers to deal with some actions reflexively
  • They are the preferred option of many users who regard the keyboard as their primary means of input, a habit that future generations may (or may not) find quaint and charming but which is currently considered (regardless of scientific reasoning) as pivotal to their productivity.

As such, any discussion on the Internet regarding them is thus automatically devoid of purely rational reasoning, so allow me to state outright that this is about what I think about these tools, and that you are perfectly free to go off and ponder, reason through and post your own views on the matter (preferably in that order, although I’m positive some people will skip the reasoning bit).

Furthermore, considering that has been around for eight years and had marked influence in a number of similar utilities, drafting a post comparing Alfred, and in this day and age would ordinarily be asking for trouble if it weren’t for the fact that I’ve recently been trying to use all three and still think there’s room for improvement.

First of all, allow me to remove two “contenders” from this analysis:

  • own works fine as a simple application launcher on its own, but does not easily let me do things I need to do every day, like fire up IM and talk to someone with three taps on the keys.
  • has a number of devout followers (tip o’the hat to Justin Blanton), but I could never really get to grips with it for some reason.

I also happen to be familiar with and on Linux, but the point here is that I’m not going to go on about anything but the aforementioned three apps.

Secondly, allow me to explain that I have rummaged about in the source of both and . I couldn’t get to grips with source and found a lot more interesting given that it was more modern code, so I started keeping around a personal, experimental fork of where I actually tried to at fixing some of the niggles I have with it.

Then, of course, real life happened, and I decided to with Alfred, which, frustratingly is free on the App Store but which has most of its useful features available as an (off-store) “powerpack” – which I’ve since confirmed is due to the Mac App Store currently not supporting in-app payments/upgrades.

But still, I find it annoying to have perfectly good free software that almost does what I need, so in the table below, and despite implementation details, I’ve tried to compare them where it concerns me, which includes abstracting whether or not a feature is built-in or an add-on. For instance, there are plugins that are built-in and activated by default, and there are plugins that are not built-in at all.

So, here are the features that I really care about:

Feature Detail Alfred
Contacts search With instant photo display No Yes Depends on UI
With drill-down into any field No Yes (including web search of contact URLs) Yes, but somewhat messy
IM can IM instantly on iChat No Built-in Plug-in
can start iChat voice chat or Telephone instantly No Built-in Requires fiddling with services
Applications Switching Yes Yes Yes
Process management (quit, hide, force quit, etc.) Yes, but counter-intuitive Yes, as an action upon a selected object. Same as QSB
Invoke menus No Built-in, automatically drills down to the right menu Plug-in, requires fiddling
File system Finding files or folders Yes (very basic) Yes (drill down) Yes (drill down)
Doing more than just open files (open with, show in finder, attach to e-mail, etc.) No Yes Yes
Extensibility Plugins Limited to “powerpack” buyers, and not there yet Yes (minor niggles) Yes
Modern source code ? Needs some dusting can’t really tell yet

The winner for me is pretty clear, I think. But, again, that’s my opinion.

To stave off some of the canned replies, I also care about support, a community, regular maintenance and a degree of future-proofing. And I understand the marketing angle pretty well (ten years’ worth of that make you appreciative of the backstories behind apps).

But that doesn’t mean that the best fit for my needs has to be the most popular, which is why I’m more than a bit concerned, especially given that has some pretty serious bugs where it regards running on Lion at the moment that seem to have no feedback and that I can’t even begin to wonder how to fix on my own.

And the prospects of finding myself stuck with either Alfred (which does pretty much nothing special on its own) or (which tries to do entirely too much and carries too much baggage) are definitely not appealing.

My original plan for was to basically gut the whole thing until it had absolutely nil “fluffy” network features (which a colleague of mine euphemistically referred as “taking the Google out” of it) and make sure that worked, but I’m guessing that won’t be happening – removing dependencies and cleaning up the build system so that it can be built from a minimal source tree under 4 might take three to six months on its own at my pace (which hasn’t kept me from making a start on Github and having portions of it already building), and I don’t have the luxury of having “20% projects” like the guys (plus, like my own boss pointed out this week, department heads are the pointy-haired guys in comics now, and that takes a lot more time than you’d think).

Still, I hope this post at least rekindles some interest in , which I’d love to see make some kind of headway.