- From: Jim Whitehead <ejw@cse.ucsc.edu>
- Date: Wed, 11 Apr 2001 18:03:09 -0700
- To: "Greg Stein" <gstein@lyra.org>, <ietf-dav-versioning@w3.org>
> I believe my position matches Tim's precisely. > > I see no reason to believe that a label is anything other than a > sequence of characters (in some specified encoding). I do not see that > language has any relevance to the topic. A label is a human-readable string. To correctly display an arbitrary sequence of Unicode characters, a computer must know the language of the string. This allows the computer to, for example, draw the characters left to right, or right to left, as appropriate. Thus the language tag is necessary. However, this is all predicated on the label being just a single string. What if label is changed into the pair: (label_id, label_displayname) Comparisons could be performed against the label_id (which is a computer-generated ID), and the label displayname could be shown to the user. This would keep the human readability, while avoiding the i18n string comparison problems. Header marshalling would be ugly, but feasible. - Jim
Received on Wednesday, 11 April 2001 21:04:36 UTC