Wandering Thoughts archives

Why people still like TCL/TK

This is the complete source code for a little X program I call tkmsg. It puts up a window with a text message (given as the command line arguments, ala echo), sized to fit the message, and when you click in the window it goes away.

set myname "tkmsg"
wm title . $myname
wm iconname . $myname

frame .frame
button .text -text $argv -command {exit}
pack .text

(And this also automatically handles standard X things like the -geometry command line argument.)

For all its warts, TCL/TK is pretty much the closest the Unix world has come to a Bourne shell for X programming: something you can use to easily whip up quick little utilities and scripts. And it's pretty good at that job, as you can see.

(And like sh scripts, it can be written by people who are just bashing rocks together; you don't have to choke down a big pile of library documentation just to do something simple. The most time consuming bit of writing this script was finding out how to print out stuff to standard output, so I could see if TK did anything peculiar to $argv or if I could just use is straight.)

A bonus to writing documentation

A quick observation:

There is nothing quite like writing up the documentation for a new feature you've just programmed to poke you into realizing just how lame your quick implementation is, and spur you into doing a better, less limited version. Somehow it is just hard to put the limits of a quick hack down in the black and white of text; making the limitations explicit makes them all the more cringe-worthy and petty.

A thought about interactive development environments

As a followup to yesterday's entry:

One of the reasons that interpreters and other interactive, immersive development environments improve your productivity may be indirect: by giving you rapid feedback, they give you the feeling that you're getting things done, which in turn improves your motivation and morale.

(And of course this is addition to the direct fun you get from having shiny flashing things happening right here, right now.)

Of course, this is just a feeling of productivity; it may or may not be real productivity. Which may explain the eternal appeal of various work avoidance techniques like rearranging your desktop; there too you are getting feedback, so the little circuit in the back of your mind that equates feedback with productivity can be happy.

(I find it interesting that serious work avoidance behaviors in people I know often involve rearranging things, as opposed to more passive activities like reading LiveJournal.)

You can then extend this to how people who aren't getting anywhere and are getting demotivated and unenthused as a result, seem especially prone to such avoidance techniques. They're not just avoiding yet another rewardless slog, they're also getting some sort of 'look! things happening! I must be productive!' reward from the back of their mind.

We do know that people like doing things that have visible effects. For example, people often report that they're faster navigating through text with the cursor keys than using the mouse, despite the actual objective measurements generally saying otherwise.

(Disclaimer: I am half thinking out loud here, and may be talking through my hat (or my bicycle helmet).)