Note by Toltec

[Bo Thorsen]

On Tuesday 13 July 2004 21:30, David Faure wrote:
> On Tuesday 13 July 2004 21:20, Martin Konold wrote:
> > Am Dienstag, 13. Juli 2004 21:09 schrieb David Faure:
> > > On Tuesday 13 July 2004 20:58, Joon Radley wrote:
> > > > Changed. We need to define the legal format of a UID to avoid
> > > > encoding issues.
> > >
> > > Yes. But I think it's simple to define: us-ascii only.
> >
> > What about an even more narrow class  [A-z,0-9] ?
>
> We have  <uid>libkcal-1004052451.119</uid>, where - and . don't fit in
> your range.
>
> I don't see what's wrong with defining it as "us-ascii", since we know
> it's the set that won't give problems.

So I didn't remember correctly :)

The problem with us-ascii is that you have things like \0, \n etc. in 
there also. Don't know if it's a problem in reality though.

Bo.

-- 

     Bo Thorsen                 |   Praestevejen 4
     Senior Software Engineer   |   5290 Marslev
     Klarälvdalens Datakonsult  |   Denmark