Incorrect case for a folder path on a fat-formated volume creates a duplicate folder

Bug #235878 reported by Michael Roy
2
Affects Status Importance Assigned to Milestone
nautilus (Ubuntu)
Invalid
Low
Ubuntu Desktop Bugs

Bug Description

Binary package hint: nautilus

See video: http://www.youtube.com/watch?v=x1fsngMKQyU
See also forum thread: http://ubuntuforums.org/showthread.php?t=805971

Bug is in: Ubuntu 8.04
(Assumed) Package: nautilus 1:2.22.3-0ubuntu1

Rationale:

Fat filesystems are not case-sensitive, and are not compliant with Unix naming conventions. Fat filesystems are still common on dual-boot systems and are widely used on USB thumb drives and digital camera memory cards, and some attention should be given to make sure fat's limitations are handled gracefully, especially given its continued use.

Expected to happen:

A user has a folder with subfolders and data, called 'Oh-no' on a fat filesystem. The user is working with tree view open on the side panel in Nautilus. The path of the folder is /fat/Oh-no. The user types the path incorrectly, mistaking the case - '/fat/oh-no.'
Nautilus should:

A. Ignore the mistake and point to '/fat/Oh-no,' possibly correcting the path's incorrect case for the user (this is what MS explorer on Windows XP does. In Dolphin on KDE 4.1 beta 1, the path is not corrected for the user, but no unexpected behavior occurs. [see "what happened instead"]).

-Or-

B. Return an error, saying that '/fat/oh-no' does not exist (as it does does not, using strict case-sensitivity).

What happened instead:

In tree view, on the left hand side pane, a second folder appears. Both '/fat/Oh-no' and 'fat/oh-no' are visible. Both appear to contain identical subfolders and files, though only one copy of the data exists.

Why this bug should be fixed:
The creation of a duplicate folder invites the user to press "delete" to remove the redundant files. Yet there is only one copy of the data. Nautilus misrepresents the data on the drive, compounding user mistakes.

See also the following old bug, which may be related: https://bugs.launchpad.net/ubuntu/+source/nautilus/+bug/9758

slink3r (brian-slinker)
Changed in nautilus:
assignee: nobody → brian-slinker
status: New → In Progress
Revision history for this message
Pedro Villavicencio (pedro) wrote :

Hello Slink3r, are you going to work in this report?

Changed in nautilus:
status: In Progress → Incomplete
Changed in nautilus:
importance: Undecided → Low
Revision history for this message
slink3r (brian-slinker) wrote :

I tried contacting the nautilus list, never got a response.

Revision history for this message
Michael Roy (mikemlp) wrote :

That is unfortunate. Now that I am aware of this behavior, it is easy enough to work around, I just wish that I hadn't learned by losing data. The only thing I can suggest is to try again with upstream. Perhaps "bump" the message on the 16th, 1 week from original posting. It may have gotten lost among pre-existing discussions of ongoing debate on the list.

On the mailing list, you said, "I've noticed this bug has been reported in the past, and that on the bug-page it was mentioned that if this bug were to show up again to e-mail the list." Do you have a link to the past bug report in Gnome's bugzilla? Linking to the bugzilla report that you saw might also be helpful to anyone considering a response on the mailing list, if you choose to "bump."

Out of curiosity, I notice that this bug is marked "incomplete." I am unsure of the criteria for each of the various "status" options, so I'm imagining, without having all of the information, what is required to change a bug's status from "incomplete" to "confirmed." Does it require several users to notice the same behavior? If so, then another user has reported identical behavior in the forum post that I have linked to in the original report. Does it require that upstream acknowledge the bug? If so, then it seems we are back to waiting for the nautilus mailing list.

I think this scenario is pretty easy to duplicate, but please let me know if any additional information is required, and I will be happy to assist.

Finally, one additional note: If one follows the youtube video exactly, deleting /fat/oh-no, and causing /fat/Oh-no to be deleted, if one then decides to create a new folder, /fat/Oh-no following the deletion, this creation of /fat/Oh-no will also "respawn" /fat/oh-no automatically. The only way that I could find to end this behavior was to log out of gnome, return to GDM and log back in. In summary, it appears as though nautilus "remembers" the duplicate state of a previously deleted folder affected by this issue while in the same gnome session. As you might imagine, this made it take longer to get a "clean" video of this behavior in action. I share this anecdote in hopes that it might help track down the source of the issue.

Thank you both for your efforts, I will check back periodically.

Revision history for this message
Pedro Villavicencio (pedro) wrote :

Any news on this?

Revision history for this message
Sebastien Bacher (seb128) wrote :

could somebody try if that's still an issue on intrepid and describe easy steps to trigger the bug?

Changed in nautilus:
assignee: brian-slinker → desktop-bugs
Revision history for this message
Sebastien Bacher (seb128) wrote :

We are closing this bug report as it lacks the information, described in the previous comments, we need to investigate the problem further. However, please reopen it if you can give us the missing information and don't hesitate to submit bug reports in the future.

Changed in nautilus:
status: Incomplete → Invalid
Revision history for this message
Michael Roy (mikemlp) wrote :

This bug is still present in Intrepid final. Please re-open.

Easy steps to re-create:

1. mount a fat-formatted volume

2. create a folder with a capitalized name, (put a subfolder or file within the folder -optional)

3. change the folder name to lower-case in the nautilus path, and hit enter

4. watch 2 folders appear in tree view, one with the capitalized name, one with the lower-case name.

(See the video, the formal steps at the top of the page. The same steps apply to Intrepid).

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.