After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 332951 - hex wep key entry allows non-hex characters
hex wep key entry allows non-hex characters
Status: RESOLVED FIXED
Product: NetworkManager
Classification: Platform
Component: nm-applet
unspecified
Other Linux
: Low enhancement
: 0.7
Assigned To: Dan Williams
Dan Williams
: 341011 (view as bug list)
Depends on:
Blocks:
 
 
Reported: 2006-03-01 05:30 UTC by Christopher Aillon
Modified: 2008-01-24 04:47 UTC
See Also:
GNOME target: ---
GNOME version: ---



Description Christopher Aillon 2006-03-01 05:30:43 UTC
Upstreaming from: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=181755

We can probably do this with a signal handler.


------------
Description of problem:
When attempting to a wep protected wireless network, the NM authentication dialog is presented for entry of the hex wep keys / ascii wep keys / wep passphrase.  When you select hex wep keys, the entry field for the key will accept non-hex values.

I can see this being a reasonable behavior if switching between hex/ascii/passphrase saved the value in the field, but this is not the case.



Version-Release number of selected component (if applicable):
NetworkManager-0.5.1-12.cvs20060213

How reproducible:
Always

Steps to Reproduce:
1. Attempt to connect to wep network using NM
2. Wait for authentication dialog, choose hex wep key
3. Enter non-hex values
  

Actual Results:  Non-hex values were accepted as valid input for a hex only field.

Expected Results:  Field should have been filtered to 0-F.
Comment 1 Christopher Aillon 2006-03-01 16:56:55 UTC
But then an issue of course becomes, what happens when someone fills in a passphrase, wants to select ASCII key, but accidentally selects hex key... Do we clear the textbox?  We'd almost have to...  Would be nice for the already typed info to stick around if they go back to passphrase, though.
Comment 2 bjohnson 2006-03-01 20:50:56 UTC
I am no GUI designer.  But here are my thoughts.

As you fill in a key and the select a different key type, the box is cleared.  This is the current behavior.  As Christopher Aillon points out in comment #1, this is sub-optimal from a usability point of view.

However, I think if you ended up keeping a value per key type, so that if you re-selected that type, the value would re-appear, would be confusing for end users.

Perhaps the correct behavior is that one value should be maintained and always displayed no matter what key type is selected.  Then, if the key was invalid for that key type (ie. non-hex characters entered for hex key type), the confirmation button could be disabled and a message displayed that the key contains invalid characters.

The confirmation button is already disabled when the key length is wrong in this situation, so it should be pretty easy to tie into that logic (easy to say without looking at the code).
Comment 3 Christopher Aillon 2006-03-01 22:37:32 UTC
Might that further confuse users? -- "Why is my confirmation button disabled?  I typed in all the info!"

Is that a good thing?  Probably...
Comment 4 bjohnson 2006-03-01 23:07:54 UTC
I think it would only make sense with a companion message something like "your key contains illegal characters" or "your key contains characters other than 0-9, A-F".

Otherwise, you are correct, it would probably be very confusing.
Comment 5 Teppo Turtiainen 2006-09-17 13:13:39 UTC
*** Bug 341011 has been marked as a duplicate of this bug. ***
Comment 6 Dan Williams 2008-01-24 04:46:54 UTC
input is clamped now in trunk and stable; if/when we fix the password-clears-on-WEP-type-change we shouldn't blow away the entry but still filter it.