Re: [VCARDDAV] vCard format and lack of interoperability

[Alessandro Rossini <[e-mail]>]

On 20 Nov 2012, at 18:16, Doug Royer <[e-mail]> wrote:

> And is sounded like you agree it can break vendors.

No, actually I think that the standard way to introduce extensions does not guarantee that these extensions do not crash other vendors. A poorly implemented vCard parser may crash with unknown properties, no matter if their names start with "X-" or not; a correctly implemented vCard parser would ignore them.

> So if your argument is to remove a standard way of adding them, because they will add them anyway.

No, my argument is that the standard way to introduce extensions may encourage vendors even more to ignore the interoperability issue.

> If that is your argument, and given that there is no protocol police force, why would anyone spend money to change their product so that is was not compatible with their legacy products? It would be an invitation to their customers to spend time and money looking a competitive products. What company would consider that?

By way of an analogy, Microsoft did exactly this when they committed to comply with the W3C web standards. They cooperated with the W3C to improve some of the web standards, and they rebuilt Internet Explorer almost from scratch to support these standards, breaking the compatibility with older versions of the browser.

> If you have done all of this work and found that there are common objects done differently by different vendors, write an Internet draft and propose a standard way to do those common objects. That's how to improve the specification.

Yes, I found out that common contact information such as email addresses, telephone numbers, postal addresses, web addresses, and instant messaging addresses can be represented in two ways: by means of standard properties, or by means of standard properties grouped together with non-standard properties. The second way is currently used by Apple (and other vendors targeting Apple); it is unnecessary and prevents interoperability. Therefore, I suggested to improve the specification by removing grouped properties and non-standard properties.

Looking at the archives of this mailing list, it seems that the vCard working group was partially aware of this issue, but preferred backward compatibility to interoperability:

http://www.ietf.org/mail-archive/web/vcarddav/current/msg01850.html

Personally, I would have chosen the opposite.

Cheers,
--
Alessandro Rossini
http://alessandrorossini.org
http://twitter.com/alerossini