Dear FOSS Developers

I’m speaking specifically to FOSS developers who are involved GUI / Human Interface type development and design.
There’s apparently a percentage of ya’ll who seem to think that the proper way to execute a program represented by an icon is to just click on it.
I have one thing to say to that. Or. Several things. With the same message…
Every once in a while someone seems to think that it’s proper to make it so if you click on, say, a Firefox icon, it means ‘Start Firefox RIGHT NOW’. This is ridiculously, painfully WRONG. A desktop contains objects. If you touch something, it doesn’t mean “EXECUTE THIS RIGHT NOW, NO MATTER WHAT IT IS”.
Imagine if this were taken on in the real world. Your car would start whenever you touched your keyring. If you touched the handle on your sink faucet, water would come gushing out immediately.
This pattern has been cropping up in KDE on occasion, though I’m assured it will not be the case in KDE4. Gnome is rife with it. And Puppy Linux, an OS that can run via LiveCD, will happily start up a 60meg program (Firefox) if you happen to touch the Firefox icon on the desktop. Just touch it. A process that can take 4-5 minutes in LiveCD mode.
So, in conclusion. No. For the love of all that is sane, do not do perpetuate this bad design.
Warm regards, me.


A wandering geek. Toys, shiny things, pursuits and distractions.

View all posts by

3 thoughts on “Dear FOSS Developers

  1. The worst part of it is that my web browser has already adopted this behavior! All I have to do is click once on a link or an image, and bang! the thing goes and activates it! Durned thing. 🙂

  2. I have to point out that “touching” is not “clicking”. I’d be pretty horrified if merely waving your mouse over an icon activated it — the virtual equivalent to “touching”. A click is the virtual equivalent to pressing a button. And yeah, I expect things to happen when I press a button.

  3. @thisdamnhouse: no, sorry, I have to really disagree with you. In fact, you even said it in your post:
    I’d be pretty horrified if merely waving your mouse over an icon activated it
    The key words there are ‘over’. You never touched the icon, you metaphorically moved your mouse over it. Surely any activity that activates when the mouse rolls over it is terrible (witness the current trend of ‘mouseover menus’, another horrifying trait). But on a GUI desktop, the pointer is the equivalent of your hand – it’s showing where you will be doing an activity. In all modern GUI designs (heck, going back as far as there HAVE been GUI desktop environments – PARC Xerox proto-Mac days), you always had to click on an object to specify you were working on it.
    A click is the virtual equivealent to pressing a button, and yeah, I expect things to happen when I press a button.
    Note that in the metaphor we’re speaking about, there is no button to press. An icon is not a button. It is a visual representation of the entire object itself. It has no ‘start’ ‘stop’ ‘begin’ ‘end’ button, it just is. How are we to determine that clicking on the object means “Run this immediately”? Greater minds than I have come up with the concept of a double-click. “Select this item, and execute it.”
    Objects should not be started, from a dead stop, with a single click.

Leave a Reply

Your email address will not be published. Required fields are marked *

The reCAPTCHA verification period has expired. Please reload the page.