[Cosmo-dev] Cosmo 0.6 as first dogfood release

[Mikeal Rogers]

> On 1/2/07, Mikeal Rogers <mikeal at osafoundation.org> wrote:
>
>> We're at the point in our development where caldav should be pretty
>> much done and we're moving on to new features. If caldav is broken in
>> any way we should fix it, like we would with any other regression.
>
> not true. we've only implemented through draft 12, and a few select
> bits from later drafts. we are definitely not done.

The only important changes for us is the addition of the collation  
attribute and removal of caseless attribute, which isn't that big.  
The change to support only utf-8 and ascii actually closed a bunch of  
tasks we had.

I'm really doubtful that our lack of implementation in the changes in  
the last few drafts would cause an interop problem in the existing  
clients.

 From caldav-draft-15

Changes in -15

    a.  Switched to using collations for text-match element in calendar-
        query report.
    b.  Removed caseless attribute from text-match element.
    c.  Removed UNICODE4 reference.
    d.  Removed mailing list comment.
    e.  Made calendar-home-set property a SHOULD as it was in previous
        drafts.
    f.  Now require https download of external attachments.
    g.  Changed some improper uses of 2119 terms to lowercase.
    h.  Updated some references to latest specs.

Changes in -14

    a.  Reverted to normative reference to 2518 and added informative
        reference to 2518bis.
    b.  Reinserted section describing preconditions/postconditions.
    c.  Removed redundant compliance statement in last paragraph of
        Section 3.
    d.  Clarify that only text/calendar is allowed if supported- 
calendar-
        data property is not present.
    e.  Removed redundant compliance statement in Conformance paragraph
        of Section 6.2.1.
    f.  Removed redundant compliance statement in first paragraph of
        Section 7.5.
    g.  Fixed incorrect whitespace in elements in example in Section
        7.7.6.
    h.  Fixed incorrect CDATA descriptions in various places.

Changes in -13

    a.  Changed mailing list draft description.
    b.  Added security review suggested text to Security Considerations.
    c.  Changed external attachment support to require http URI
        downloads, and optionally others.
    d.  Added reference to text-match element in Internationalization
        Considerations section.
    e.  Changed 'undefined' to 'not specified here' in ETag behavior
        section.
    f.  Added reference to RFC4346 with note that it obsoletes RFC2246