Re: [cgmo-webcgm] ISSUE: prohibit picBehavior on fragment of linkuri's URI parameter?

[Lofton Henderson <lofton@rockynet.com>]

I want to be clear about my motivations, entirely pragmatic, for my 
recommendation ...

Although I appreciate the aesthetic appeal of cleaning up these legacy 
warts on 1.0, on the other hand, probably everyone has implemented this 
(there's a 1.0 test of it, I reckon?), and it's more disruption to 
deprecate / remove it than its removal is worth.

-Lofton.


At 06:20 PM 9/22/2005 -0600, Lofton Henderson wrote:
>Ref:  http://lists.oasis-open.org/archives/cgmo-webcgm/200509/msg00093.html
>Comments:  44
>==========
>
>ISSUE:  prohibit picBehavior on fragment of linkuri's URI parameter?
>
>DISCUSSION:
>Dieter writes, about specifications in the last paragraph of 3.2.2.3,
>"2nd sentence discourages usage within fragment, 3rd sentence should hence 
>be changed to:"For specifying picture behavior, WebCGM 'linkuri' instances 
>shall use the Behavior string only.""
>
>This is about the inclusion of picture behaviors within the fragment, in 
>the URI specified as the first parameter of 'linkuri'.  Some time ago, 
>shortly after we agreed on the definition of "deprecate" and put it into 
>A.2, I raised the question of whether we should really apply "deprecate", 
>because it meant that we would have to prohibit it in 2.0 content.  As I 
>recall, I recommended changing the word "deprecated" to "discouraged", 
>which would not prohibit the use of it.  No one commented or objected.
>
>Making Dieter's suggested change -- "..SHALL use the Behavior [3rd] string 
>only -- is a normative, conformance change from 1.0 to 2.0.
>
>As we saw from the last telecon, deprecating or obsoleting (removing) 1.0 
>things in 2.0 comes with a price -- you have to then consider some gnarly 
>backward-compatibility conformance issues.
>
>RECOMMENDATION:  leave it as is.
>
>Regards,
>-Lofton.
>