[sylpheed:37201] Re: Button order in modal dialogs

[k0 at trixtar.org]

Thu, 27 Jul 2023 14:02:34 +0300
Anton Shepelev <anton.txt at gmail.com> :

> Hello, all. The following Usenet article may be of interest to Sylpheed
> users and developers:
> 
> -------------------------------------------------------------------------
> Date: Tue, 7 Mar 2023 21:27:14 -0500
> From: bad sector <forgetski at INVALID.net>
> Newsgroups: alt.os.linux
> Subject: about button order
> 
> 
> I started asking myself why do I keep hitting the NO instead of the YES 
> button when trying to empty the trash in Sylpheed seeing that emptying 
> it was the very reason why I had right clicked the Trash folder in the 
> first place. It threw me for a loop (other dialogs might use CANCEL and 
> OK buttons) and I wanted to know why. In this case the buttons are on 
> the right side of the dialog with the NO button next to the right edge 
> and the YES button inboard from it i.e. the decisional progression of 
> buttons is perpendicularly adjacent to the right edge and this attribute 
> may have more to do with response than the direction that a user 
> normally reads in.
> 
> Seems to me that my attention first goes to the dialog body but then 
> attacks it from the side where the progression of buttons is 
> perpendicular to it. Try it, see what you think. I haven't tried with 
> buttons that completely fill the dialog in one direction or another, 
> there might be more to learn...
> 
> Methinks that in a line of buttons with a decisional progresion 
> perpendicularly adjacent to the dialog edge where the buttons are, the 
> button most likely to be clicked should be the one that is closest to 
> that edge. Also, if a person calls for the dialog then the least-likely 
> button to be clicked is the CANCEL or NO button, and it should be farthest.

Yes, we hail from a hunting past and them hormones are still twitching in our bones, more competent behaviorists will point out the details but for my money it's the triangulating instict that is at work  here. Simply put, we are still running down the prey! And the way that instinct works remains for ever indominable, here come two examples. Say the word 'mother' in you NATIVE tongue, what do you feel? Next say it in any other one of the many languages that you know, again note what you feel. Examples two: Look at the torque 'pointer' on a bolt tightening tool as it sneaks up on the preset 60 degree postion, or a compass needle approching what you want to see during hard cornering (less so a speedometer needle), or in fact the hands of a real clock... next look at a digital counter displaying the same information. With the first 'needles' that 'point' we instantly 'see' us 'pulling angle' on what we want to run down, but the digital horror we invented just years ago confronts us with as sterile a process as hearing mother in another language. I wrote this up in Time Magazine in another life, nobody paid attention then and nobody pays attention now.

With a dialog we'd normally steam right to it but on a monitor that can really be from any direction so it cannot be presumed. Next we mentally go around it looking for a crack in the defenses or a line of buttons resembling what we want, a bit like a boxer would. Finally we attack on the side of the optimal risk/reward ratio, where we see the buttons that we want to see and of course expect the button that we want to be at hand on first strike (I am aware of the war between Gnome and everybody else with the buttons issue).           

Another cardinal point that should never be forgotten is 'why did he user call the dialog'? There should be back doors but we don't call them 'back' doors for nothing, they should not be on the front porch. I'm minded of the way sound is handled in my computer. If I plug my guitar into the sound card it's OBVIOUSLY because I want to hear it and not because I want to go into a pulse dialog to set soundcard X:0:mic for input. 

'nuff said, I      l   o   v   e        Sylpheed :-)   

-- 
https://imgur.com/R9cXaXa.png