Hi,

Before I give my vote, I just want to clarify a little what I meant.

Assume you have package A, which can carry some information.

Then you figure out you need a new information element, but all the old
information still remains exactly the same. You create package B.

Now, package B would be fully backward compatible with package A.

But, if there is no versioning, and backward compability guidelines, I would
have to indicate explicit support of both package A and B.

Of course, if package B is NOT backward compatible, it would have to be a
compeltely new package.

Another option is of course to deal with these kind of things inside the
info packet itself.

Regards,

Christer



 

> -----Original Message-----
> From: Eric Burger [mailto:[EMAIL PROTECTED] 
> Sent: 30. lokakuuta 2008 18:53
> To: Christer Holmberg
> Cc: SIP IETF
> Subject: Info Package Versioning
> 
> Channeling Keith here: is there a real use case for package 
> versioning?
> 
> I cannot imagine backwards-compatible packages. Once one 
> starts changing the stuff in an INFO package, it is a 
> different package. That is, rather than
> 
>     Info-Package: foo-package;version=1
>     Info-Package: foo-package;version=2
> 
> You, in reality have:
> 
>     Info-Package: foo-package
>     Info-Package: something-like-foo-but-not-really-foo-anymore
> 
> This is more especially so when one considers there is no 
> possibility of negotiating package versions, whereas there is 
> a possibility of negotiating different packages.
> 
> For the one instance one could come up with in theory for 
> needing versioning, remember one does get to have an 
> arbitrary parameter:
> 
>     Info-Package: foo-package               /* Version 1 */
>     Info-Package: foo-package;bar        /* Version 2 */
> 
> 
> Is this latter mechanism sufficient?
> 
> Yet another poll:
> 
> [ ] We must have abstract versioning as part of the Info 
> Framework document
> 
> [ ] We do not need to have abstract versioning, as it does 
> not do anything for anyone
> 
> [ ] We do not need abstract versioning, but it would be 
> worthwhile including some text,
>      like what is above, to show people how to do it.
> 
> [ ] Something else? _____________________________________
> 
> 
> 
> On Oct 27, 2008, at 4:48 PM, Christer Holmberg wrote:
> 
> > [snip]
> >> The wording could be better. There are no explicit IANA 
> >> considerations
> > about registering an INFO package. Its just implied here. 
> Its a little 
> > confusing here what is intended to be part of the
> >> *definition* of a new package vs. what is intended to be in the 
> >> content
> > of a particular package instance in an INFO message.
> >
> > We do need to have a section on what needs to be part of an info 
> > package definition. The mime type is one thing, for sure.
> >
> > I guess we should also look at what/if we can re-use from 
> SIP events.
> >
> > We also need to decide whether info packages should have a version 
> > number system. From experience I think that would be very useful.
> >
> > Info-Package: foo-package;version=1
> >
> > ...or something like that.
> [snip]
> 
> 

Attachment: smime.p7s
Description: S/MIME cryptographic signature

_______________________________________________
Sip mailing list  https://www.ietf.org/mailman/listinfo/sip
This list is for NEW development of the core SIP Protocol
Use [EMAIL PROTECTED] for questions on current sip
Use [EMAIL PROTECTED] for new developments on the application of sip

Reply via email to