Re: [Usability] GNOME3: Handling/Opening Documents (Website for Contributing Ideas?)



> Obviously system administrators have to know what programs are (and that
> includes home users who want to extend the capabilities of their
> computer), but that has nothing to do with how those capabilities are
> represented in the user interface.

well, I don't agree... from one side or the other the problem remains:
user should know the existence of applications when they install the
desktop or have to search the internet for "something" (a "program")
that can solve his problem/task.


> [...] . I'm strictly speaking about the user
> interface here, under the hood it could still use the old concept of
> applications and simply reload the document in the editing application.

Finally there would be two (or more) document "states": viewing mode and
edit mode. this can be an idea but I'm not so sure that this would be
better than simply think about "applications".

> Well, thinking about Gimp, document windows actually are very
> simplistic. The GUI of Gimp mostly consists of the utility windows.
> Attaching a document to those utility windows could simply turn it into
> a Gimp document. 

Yes, it's true for Gimp...

> Of course it would be more complicated than this, but
> "complex GUI infrastructure" doesn't need the application concept at
> all.

Well, i don't know... I'm not so sure... finally (think about cinelerra,
main-actor, cd-roaster, blender, evolution, etc...) the user will
actually see those groups of toolbars/sub-windows/menus as something
really different from the "document", mocking the "document-centric"
view of the desktop.

> Tools like Blender (or Cinelerra) are an obvious exception. They are not
> remotely part of the desktop, so these considerations simply don't apply
> to them. 

Why a text editor is part of the desktop and a video editor not? only
because it doesn't agree with gnome usability rules or is not ufficially
a "gtk/gnome" application? For example I don't feel gedit as part of the
desktop more than nedit or emacs. ... 

> > 6 - Instant save deserves several drawbacks... the discussion would be
> > too long but basically in my opinion it's really better to leave to the
> > user the decision about when saving changes or discard them totally:
> > full history or undo-lists are too high requirements for every
> > application-development.
> 
> This is nonsense. I already use instant save in my application
> (Scratchpad) since many months and it works like a charm. It has no
> persistent undo yet and only a rather simplistic tagging system (which I
> rarely use). [...]
> There are some rough technical challenges to overcome for more complex
> document types than plain text, but from a usability point of view, I'm
> very sure that instant save is the way to go.

Sorry, I respect your opinion but I don't think mine as a "non-sense",
so I still don't agree: basically the versioning system is a good and
amazing idea but I agree with some other mails which stress the
excessive I/O load and the difficulty of having and maintaining a
system-wide versioning system. 
Just think about the disk usage or how to work on removable media or the
CPU load if the "document" is complex to be save (and not just a text
document), or what would happen if before saving "optimisation options"
can be applied (for example compression options when saving changes in
images but there're plenty of them in other applications) or simply the
logic to apply for versioning: I'm editing a document and how would you
save the various "versions": every time I make a change? every 5 minutes
if changed? And how many versions would you store? and what if an
intermediate version is really "wrong" or inconsistent (so saving it is
useless) ... and if I changed the document without being aware of it
("save before exit" dialog no longer apply to warn me) and I really
didn't mind change the document? ... and if I changed too much the
document and I decide that I want to save it with another name: are you
sure there're no situations where "save as" would be usefull?

Too many questions to say that "automatic saving" really helps the user
instead of making things apparently simplier but really more difficult
to manage.

In my work I've found that sometimes it's possible to really semplify a
process but, in this way, managing exceptions or problems in the
"automatic system" is more difficult for users because it has really
hidden too much the real mechanism... sometimes you have to match a
trade-off and let the user be aware of certain processes.

Perhaps "saving" is this situation... perhaps not and can be really
semplified... I don't know now... but I have a feeling: it's better to
explicitly save and I would like to have control over it, decide
when/where/how many times and how to do it.

> 
> > Finally, but this is my opinion, an environment where applications feels
> > like applications is better because is more "task oriented", more
> > natural. A "mixed" scenario would only clutter the desktop semplicity.
> 
> We already have this mixed scenario: Spatial Nautilus is purely document
> based (and I love it). 

What does it mean that nautilus is "document based": you still have
icons and mime-types associated with applications. The only difference I
can see between "Spatial" and "not spacial" is the way the underliying
filesystem is treated (folder like filesystems and devices are
"graphical entities" that can be activated/deactivated on demand).

I find it interesting for a not-expert users (not for developers like me
that really like filesystem structure) but I still really use "browsing"
mode at work to avoid cluttering the desktop with several "forder
windows" (or having to double click the center button each-time) and to
use the really comfortable "folder tree". But it has helped so much my
parents using the desktop so I think it's fine for not-expert users, but
this is another story already discussed in deep...

> 
> The GNOME Storage feature page still says: "Storage was not designed in
> a vacuum, but is a chunk of a greater design for a revised desktop
> system."
> My greatest hope is, that such a document oriented "revised desktop
> system" will ever see the light of the day. Whether as GNOME 3 or
> something totally different. So far, it's just a distant dream. :)
> 

Yes, I've only expressed my doubts, but it would be really interesting
to see it in action... but I hope not in gnome :-) since a want to use
it forever :-))))




[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]