Opinion: Why Users Blame the Spatial Nautilius

The recently announced GNOME 2.6 has finally brought many features long awaited by the Linux desktop fans. GNOME 2.6 is all about ease of use, performance and unification and while it’s unfortunately hard to say that the GNOME desktop feels fast, it certainly began to be really easy to use and it has consistent look and feel — and that consistency is what makes up for most of the quality of a graphical user environment. UPDATE: Scroll down the article to read some added commentary.

Among lots of reviews of the new GNOME that have appeared on the Web,
there is no single one that does not mention the spatial Nautilius
file browser. While most reviewers find it really faster than the
old one (which is true), they seem to hate the spatial mode and blame the GNOME
developers for not providing an option to switch it off. They say: “you would
not like the web browser to open each link in a new window, wouldn’t you?
so why do you make me open each folder in a new windows and do not provide
an easily accessible option to switch this mode off?”.

Well, that point of view is one-sided. The whole thing about spatiality
is to provide the user with a real-life-alike interface
that keeps objects’ state and does not alter the contents of any
physical object if not ordered to. Browser mode folder windows
violate these rules by replacing physical object (folder,
represented on screen by a window) contents
with new set of icons every time the user opens a new folder,
and not retaining folders’ state (view mode, sort order,
icon placement).

Think of your hard drive contents as of a desk full of drawers.
Every time you put something into a drawer, you may be sure
that the next time you open the same drawer it will be in
the same place (and the drawer itself will remain in the same
place). So, when you open a folder and try to locate a
particular icon, it should be where you put it before.

Now I can hear all that “what about the web browser” croud again.
Please imagine, what’s the closest real-life representation of
a web page? Well, it might be a book. While reading a book,
you can see only two pages at once, and every time you turn
the page, the new set of two pages replaces the two seen
before. And that’s exactly how web browsers work: clicking
a link replaces what you are seeing with the new content,
unless the link points to another web site (in which case
it may open a new browser window for your convenience).
Reading the book, you may even put some bookmarks on different
pages and that’s exactly how tabbed web browsing works:
you may keep several sub-pages of the same web site temporarily
bookmarked, switch between them with one mouse click
and get rid of them (remove the bookmark) when they are
no longer needed.

So, people in fact love when the machine works in a way
resembling behaviour of real-life objects, but it seems that only when
the “spatial” application is a web browser: they accept
the book metaphor with web pages, but reject the drawer
metaphor with folders and files. Sometimes they even
abuse the physical metaphor of tabbed browsing by opening
multiple pages – not subpages of the same web site! – in
multiple tabs of a browser window. I even know few people
who never open more than one browser window, viewing all
pages in tabs; I hope they do not try to glue a daily
set of newspapers together before reading them…

What is the real cause of all these attacks on the spatial Nautilius?
In my opinion, it is just bad file organisation coupled with
a bunch of old bad habits. It’s really hard to use a spatial
file browser if someone keeps his or her files in a ten-folder-deep
structure. Browser-mode file browsers hide the lack of thought
and organisation in the filesystem structure; spatial ones do not.
Folder structure should be simple and as shallow as possible,
and the “master” folders (something like My Images or
My Music folders known from Windows) should have their own
shortcuts on a GNOME panel, so that playing your favourite
song would only require opening My Music from the panel,
opening the appriopriate album folder and double-clicking
a file icon, instead of browsing straight from the home
directory (or, worse, the root one) through several levels
of subfolders.

Keeping the filesystem structure clear will also reduce symptoms of the next
problem mentioned in many reviews: screen clutter. By the way, I cannot imagine
how spatial browsing must lead to screen clutter: opening folders
with double-middle-click or Shift-double-click closes the parent folder
window at once. And even if it is not enough, one can click one field
in the gconf configuration editor and turn Nautilitus into
“classical” non-spatial file browser. Don’t know how to
use gconf? Then you shouldn’t change the way Nautilitus works,
I presume.

What’s worst, attacks on the spatial browser try to stop the innovation.
While it is hard to call the GNOME’s spatial Nautilius “innovative”, as spatial
browsers have a long history, to mention only the famous Macintosh Finder,
it is certainly innovative to bring this idea back to life, after all these
years of browser-like file managers domination. Even Apple and Microsoft,
after years of commitment to spatial interfaces (though much half-hearted
in case of Microsoft), turned back and tried to ride the ‘web browser
alike interface’ hype, with not so brilliant results. And now, when the time
to ressurect the spatial ideas has finally come, people accustomed to the
bad interface design try to defend it only because for the past years
they have been using it! They seem to be againt re-learning the new interface,
even if it promises to be so much more straightforward and natural,
and keep using something that reminds Windows 1.0 MS-DOS Executive
or Windows 3.0 File Manager and not a modern file browser.

While spatial Nautilius is not perfect (why oh why does it need 2
to list 3000 files stored in one folder while Windows NT 4.0
Explorer lists 10000 files in 15 seconds on the same machine…), it is able
to recreate the desktop metaphor that started the graphical desktop revolution
with Xerox Alto and Star so many years ago. Please, don’t stop all these
good ideas coming back again. And remember that the spatial applications do
not organise your work by themselves: you have to help them and keep your data
organised yourself and then you will see how much spatial interface may make
work easier and more effecient.

Radoslaw Sokol is a network administrator in Poland.

OSNews’ EIC’s opinion:
Personally, I am all for the spatial interface, but not on top of the current hierarchical file systems. When filesystems become fully DB-based and MIME-based with no folders (they would be sorted based on extended attributes criteria, not based on folders) *then* a spatial interface would make absolute sense because each view/window would represent a different “search” result based on attribute search: For example, “show me the files that belongs to XX package, were created before 2004, are JPEG or PNG and their EXIF information includes the word ‘Greece Holidays'”, a query created by an easy to use dialog, like in this example. Each query can be saved down if the user wishes to, and because each one is unique, it would need to be represented in a different result window, that’s where the spatiality should come in.

But as things are today, the spatial Nautilus is going live before its time. Explorer-like file managers are a better bet for the kind of filesystems we utilize today. Yes, spatiality is the future. Not just yet though because the filesystem part needs a grassroot “upgrade” too.

The above example comes from BeOS, where its file manager, Tracker, is also a spatial file manager (unfortunately also on top of a hierarchical FS with folders, so spatiality was also not well suited despite the fact BFS was an advanced fs for its time). When Tracker became open source in 2000, the FIRST thing users added to the codebase was the option to not open each folder on its own window. It was the most common problem users had with the fm, just like today with Nautilus. This says a lot and unfortunately the Gnome Project didn’t learn from history and they basically repeated the same mistake. — Eugenia


  1. 2004-06-14 8:35 am
  2. 2004-06-14 9:00 am
  3. 2004-06-14 10:21 am
  4. 2004-06-14 10:50 am
  5. 2004-06-14 11:49 am
  6. 2004-06-14 11:54 am
  7. 2004-06-14 12:01 pm
  8. 2004-06-14 12:04 pm
  9. 2004-06-14 12:20 pm
  10. 2004-06-14 12:24 pm
  11. 2004-06-14 12:25 pm
  12. 2004-06-14 12:30 pm
  13. 2004-06-14 12:34 pm
  14. 2004-06-14 1:03 pm
  15. 2004-06-14 1:28 pm
  16. 2004-06-14 1:29 pm
  17. 2004-06-14 1:52 pm
  18. 2004-06-14 2:48 pm
  19. 2004-06-14 3:41 pm
  20. 2004-06-14 3:56 pm
  21. 2004-06-14 4:07 pm
  22. 2004-06-14 4:24 pm
  23. 2004-06-14 5:23 pm
  24. 2004-06-14 5:32 pm
  25. 2004-06-14 6:09 pm
  26. 2004-06-14 7:17 pm
  27. 2004-06-14 7:42 pm
  28. 2004-06-14 8:03 pm
  29. 2004-06-14 8:05 pm
  30. 2004-06-14 9:42 pm
  31. 2004-06-14 9:45 pm
  32. 2004-06-14 11:15 pm
  33. 2004-06-14 11:50 pm
  34. 2004-06-15 12:00 am
  35. 2004-06-15 7:43 am
  36. 2004-06-15 12:50 pm
  37. 2004-06-16 10:49 am
  38. 2004-06-16 8:16 pm
  39. 2004-06-17 9:49 pm