[Gllug] On Linux desktops...

David Damerell damerell at chiark.greenend.org.uk
Thu Oct 18 11:16:58 UTC 2001


On Thursday, 18 Oct 2001, Alex Hudson wrote:
>On Thursday 18 October 2001 11:14 am, you wrote:
>>I do, yes, know what the generally used meaning of 'GUI' is. If you're
>>using some private meaning, perhaps you should include a
>>Hudson-English dictionary in your mail messages.
>Go look up the old Boing systems, and anything that used a vector display 
>(particularly CAD systems).

Which are probably GUIs by my definition (and, hint, that used by the
rest of the world) too. What's your point?

>>>Okay, let's make it a more obvious example for you. We're running lynx (in
>>>an X terminal or not; doesn't matter). I can move the mouse and obtain a
>>>pointer. Does it matter that the pointer is a text cursor or mouse
>>>cursor?
>>If we mean 'is it a GUI', yes. Obviously the distinction between GUI
>>and not-GUI is pretty marginal and not very interesting here, but
>>that doesn't mean it isn't there.
>So, if I have a cursor for a mouse pointer, it's not a GUI, whereas if I do 
>it is...

Yes. As I say, it's a marginal and not very interesting distinction
here; and I would hesitate to describe lynx under X as providing a GUI
simply because it's getting the pointer from X; but the pointer,
itself, is part of the X GUI.

>You define for me the difference between GUI and "text-mode",

_Graphical_ user interfaces have graphics, that being what the word
means. Text interfaces don't. There are some curious edge cases
involving the line-drawing characters and whatnot, but there you go.

>>Which is sufficient. Text, not graphics, hence not graphical.
>So, when I run the command line on a text-mode display, it's text mode. When 
>I run it under a framebuffer, it's a GUI. Right.

No; it's a text mode interface either way.

>>Yes, I really enjoy guessing what all those little pictures do. 
>If you're talking about effective GUI design principles, I'm inclined to 
>agree with you. However, if you're trying to suggest that icons are a poor 
>interaction paradigm, I can point you to a number of studies which say you're 
>wrong.

You can point me to any number of studies that say something patently
false, but who cares? You're not going to convince anyone that - for
at least some purposes - little boxes with words in don't convey more
information than little boxes with pictures in; for instance, the
Emacs menu here has an item marked 'Buffers', and you just can't
suggest a little picture that is of more use than the word 'Buffers'.

>>IYO, anyway. Me, I would say it varies depending on the information
>>being represented; which is why I have mostly text windows with
>>graphical boxes around them...
>though, it displays data. Basic information is available (the colouring of 
>stocks on a listing, for example), but more information can be achieved 
>graphically (bar chart comparing performance of stocks over a year period, 
>for example) than in a text data equivilent. 

As I say, it depends on what's being represented.

-- 
David Damerell <damerell at chiark.greenend.org.uk> flcl?

-- 
Gllug mailing list  -  Gllug at linux.co.uk
http://list.ftech.net/mailman/listinfo/gllug




More information about the GLLUG mailing list