Re: [Usability] New Menubars

2009-04-20 Thread Philip Ganchev
On Thu, Apr 16, 2009 at 6:43 AM, Anton Kerezov  wrote:
[...]
> Gome-DO is a good thing but IMO they have bloated it too much with all kinds
> of useless plug-ins (of course you can turn them off). What is good about do
> is that when you type it learns what you want the end result to be and
> autocomplets it - that is something that is worth intgerating in the new
> menu system.

This could be a disadvantage because it would interfere with
habituation. The user would always have to check if the command they
want is in the position they expect, in case the computer has learned
since the last time the user used the command.

Anton, sorry but I still do not understand the purpose of your idea.
It might help if you give an example use-case.  Is the purpose the
same as with my idea (text search for menu items and any other
commands)? If so, then I think we should evaluate and compare our
design choices. Do you agree?

[...]
> - you have a box that assits you request a group of commands and lay out
> them in the menubar area (the way the items are shown is topic that should
> be thought on a lot)

How do I request the group: text search? Can it show commands (menu
items) that are in different menus. For example, when I type  "scal",
it will find "Layer -> Scale layer..." and "Image -> Scale image..."?

> - help you execute commands but that would not be the primary goal of this
> interface because you cannot learn so many specific words and if you do you
> will forget them easily when you don't use the system for a while.

What do you mean by "commands"? In my proposal, help with executing
commands is a major motivation; "command" to me includes menu entries.

Alan Day wrote:

"There's a lot of interest in ways to combine command (text) and
graphical interfaces at the moment - an interest that I share. There's
definitely an advantage there for repetitive actions, as well as some
exciting possibilities around natural language input."

And a huge advantage in finding the action you want (compared to
looking through the menus) and streamlining the user input (compared
to mousing through the menus).

Natural language may be good, but it is likely to be difficult to
design and implement well. We should pick the low-lying fruit (text
search) now, thus creating a framework to incorporate NL with more
thought.
___
Usability mailing list
Usability@gnome.org
http://mail.gnome.org/mailman/listinfo/usability


Re: [Usability] New Menubars

2009-04-20 Thread Anton Kerezov
On Tue, Apr 21, 2009 at 2:18 AM, Philip Ganchev wrote:

> This could be a disadvantage because it would interfere with
> habituation. The user would always have to check if the command they
> want is in the position they expect, in case the computer has learned
> since the last time the user used the command.
>
> Anton, sorry but I still do not understand the purpose of your idea.
> It might help if you give an example use-case.  Is the purpose the
> same as with my idea (text search for menu items and any other
> commands)? If so, then I think we should evaluate and compare our
> design choices. Do you agree?


I agree. This was just a thought about possible ways to go but I think that
the best way for a start would be only search for commands (e.g. menuitems +
desctiptions). One of the reasons why this was discussed was the possibility
of the space not being enough to display all items. Do you have any ideas
how this could be done?



> How do I request the group: text search? Can it show commands (menu
> items) that are in different menus. For example, when I type  "scal",
> it will find "Layer -> Scale layer..." and "Image -> Scale image..."?


Yes text search that searches for all the text available in the current
menus + descriptions.



> And a huge advantage in finding the action you want (compared to
> looking through the menus) and streamlining the user input (compared
> to mousing through the menus).
>
> Natural language may be good, but it is likely to be difficult to
> design and implement well. We should pick the low-lying fruit (text
> search) now, thus creating a framework to incorporate NL with more
> thought.


I totally agree with you here. It is good to start simple and evolve the
system over time when people learn how to use it and request more
intelligent ways of showing the items.

I will write use cases but wouldn't it be better to make a wiki page? And if
yes where I can do that (which section so on ...)?

-- 
Anton
___
Usability mailing list
Usability@gnome.org
http://mail.gnome.org/mailman/listinfo/usability