Re: [emergency-msg] Backward compatibility

["R. Allen Wyke" <emtc@nc.rr.com>]

Hold on guys - we are talking about 2 things here. My original email had
to do with the XML Schema defining CAP - not the WSDL defining any
services that support CAP. You both are talking about the later. Just to
restate, here is what I was saying.

There was a statement that some implementors (one way broadcast for
example) are having trouble with having URIs for extra attachments, etc.
- they have no ability to access and use them. So, I asked the question
about validating the actual alert instance against the schema, which
would be hosted at the other end of.....yes, a URI. If they don't
support accessing URIs, then it would not be possible for them to
validate the schema unless the schema was "built directly into" the
application. Following this a step further, it means that supporting
future versions of CAP would definitely require code changes, validation
against the current schema (v1, v2, whatever), does not necessary mean
that. 

>From a system/application design standpoint, the approach of hardcoding
the support for a schema in an app is not wise and is very short
sighted. It locks you into a single version of a standard, or at least
makes it hard to support newer versions, because it means your system
must be updated to support the newer versions.

Allen

On Thu, 2003-10-02 at 13:54, Rex Brooks wrote:
> Ah! Thanks for the clarification, Art,
> 
> I was a bit confused but the connection has now been made and I agree 
> with you. Partly my confusion was due to the fact that I am about to 
> venture into VoiceXML waters, where, as it turns out, v2.0 which is 
> what most everyone is using or preparing to switch to imminently, is 
> not backwardly compatible with v1.0 and we just had a rather lengthy 
> discussion in another venue about whether to consider compatibility 
> with 1.0 and decided against it. Generally, this is something I am 
> rather vehemently against, but I was convinced due in no small part 
> to my own ignorance of said subject as I try to get up to speed on 
> it, AND not having to learn v1.0 was the path of least resistance. 
> That's not a good argument for such a practice, and I don't recommend 
> it. I just admit to it. I was also overruled on a number of issues. 
> However, I mention all this because I'm getting my feet wet these 
> waters largely because of my participation in Web Services for Remote 
> Portlets TC, where I somehow volunteered to chair the markup 
> subcommittee. It strikes me as a very likely that VoiceXML for ICS 
> Forms via telephonic devices on the fly will rear its head all too 
> soon. Eventually, I'm sure we will also see voice activated 
> pda/notebook type mobile and/or handheld devices that we will want to 
> use effectively in EM, and while we are looking at infrastructure and 
> versioning issues, we might as well start thinking about these 
> things, too.
> 
> Ciao,
> Rex
> 
> At 9:15 AM -0700 10/2/03, Art Botterell wrote:
> >At 7:24 AM -0400 10/2/03, R. Allen Wyke wrote:
> >>Web Services? I'm sorry, I am not sure how this applies to this
> >>discussion (validation of schemas)?
> >
> >What I meant specifically was that the ability to dynamically 
> >discover services and interfaces (via UDDI and WSDL) isn't available 
> >in many communication systems... so we can't assume that a revised 
> >schema would propagate immediately and automatically throughout the 
> >user base.
> >
> >And more generally, that we need constantly to be on guard against 
> >letting assumptions about the transport layer color our document 
> >designs.
> >
> >Sorry if I confused anyone.
> >
> >- Art
> >
> >
> >To unsubscribe from this mailing list (and be removed from the 
> >roster of the OASIS TC), go to 
> >http://www.oasis-open.org/apps/org/workgroup/emergency-msg/members/leave_workgroup.php.
-- 
R. Allen Wyke
Chair, Emergency Management TC
emtc@nc.rr.com
http://www.oasis-open.org/committees/emergency