Re: 4. ID assignment and the empty string

/ Norman Walsh <Norman.Walsh@Sun.COM> was heard to say:
| / Ian Hickson <ian@hixie.ch> was heard to say:
| | |   2. ID assignment is performed.
| |
| | It should be stated explicitly that the value used for ID assignment
| | is the normalised value obtained in the previous step.
|
| Agreed. Done.
|
| | Also, should the following:
| |
| |    xml:id=""
| |
| | ...be special-cased in the spec? At the moment a strict reading of the
| | spec leads me to believe that such an attribute would make the element
| | have an ID consisting of the empty string, which would thus be found
| | by using, e.g.:
| |
| |    document.getElementById('')
| |
| | I would assume that much software has been written with the assumption
| | that the empty string implies the lack of an ID. I would recommend
| | changing step 2 of the processing section to read:
| |
| |    2. If this normalised value is not the empty string, ID assignment
| |       is performed using the normalised attribute as the ID.
| |
| | This would take care of both these comments.
|
| I'm inclined to agree.

The Core WG discussed this issue and concluded that the empty string is
no different than any other invalid value. Please let us know if you find
this answer satisfactory.

                                        Be seeing you,
                                          norm

-- 
Norman.Walsh@Sun.COM / XML Standards Architect / Sun Microsystems, Inc.
NOTICE: This email message is for the sole use of the intended
recipient(s) and may contain confidential and privileged information.
Any unauthorized review, use, disclosure or distribution is prohibited.
If you are not the intended recipient, please contact the sender by
reply email and destroy all copies of the original message.