Summary of NWI -- the page

Maciej Pilichowski bluedzins at wp.pl
Fri Jul 10 18:13:42 CEST 2009


On Tuesday 07 July 2009 22:20:05 Matthew Woehlke wrote:

> > _This_ remark -- only here (AFAIK). _Such_ remark -- in wiki too.
>
> I'm sorry, I must be blind. I'd like to revisit this, but I am not
> seeing what part of the wiki it is I need to revisit. 

Now, I don't see it either :-))) I'll find it and post a quote :-).

> > I disagree. Application launch costs, and user should not control
> > all the time if this app quits, or not.
>
> Um... but the point is, closing the window does *not* close the app
> in these cases. (Well maybe if mail editor is separate app, but I
> think not; 

Mail editor can be launched as standalone app, so I see one window and 
I expect it will stay after closing document.

> And if it is 
> expensive to relaunch such satellite window, IMO app is doing
> something wrong.

It depends, it could be part of the workflow, but external app after 
all (like Kate as an editor).

But again -- we should end up with consistent UI, if user has to think 
what will happen -- "is this part of the app or not" -- this would be 
wrong.

> > It should stay put until I am finished. And I am finished when I
> > close the app, not the doc.
>
> I am certainly finished chatting with someone without closing
> kopete!

? If you are finished, why do you not close the kopete?

> I would like to
> > say to Composer:
> >
> > -- I launched you, because I have to write 5 emails. So stay put
> > and don't go anywhere.
>
> Hmm... is the problem that you need a "send and start writing new"
> button, or that composer takes too long to re-launch from "write
> new" button in kmail? If it was near-instant to launch, would you
> still have a problem with close-document closing composer (but only
> if launched from e.g. kmail and not launched stand-alone)?

Yes. Because I prefer if computer system is shaped like real-life. And 
I get typewriter (well, not anymore) and I fill several pages. I 
don't hide&get typewriter with each page.

App for me is like a tool, doc is like a material I work with. And in 
real-life it is usually 1:N relation (one hammer, 100 nails). 

> >> Closing when there are siblings is a little different, but I
> >> suspect there would be bug reports when using TAI-used-to-be-TDI
> >> applications otherwise.
> >
> > I don't see it, but ok.
>
> Let's say I have three konq tabs. Right now they are TDI and ctrl-W
> closes tabs if >1 tab left. Now with TAI (each tab == separate app
> instance, own process, etc.) would you be confused if ctrl-W just
> made tab empty? This is what I mean.

Ok, I would prefer that the behaviour would be unchanged.

> > State: last window app in the container.
> >
> > Action: user closes the application window.
> >
> > Possible effects (set by user globally):
> > a) entire container closes
> > b) nothing happens (action is ignored)
> > c) application goes to "blank state"
> >
> > So do you agree user should have such choice to select which
> > a/b/c she/he would like the best?
>
> Yes, I think so.

GREAT! :-) As soon as I get access to changing the summary I add this 
back.

Cheers,


More information about the Kde-usability-devel mailing list