* Aryeh Gregor wrote:
>The difference is that if you have "must" requirements that are
>specific to a single conformance class, you can write a test suite and
>expect every implementation in that class to pass it. For "should"
>requirements, you're saying it's okay to violate it, so test suites
>d
On Sun, 10 Jul 2011 22:34:59 +0200, Aryeh Gregor
wrote:
On Sun, Jul 10, 2011 at 3:59 PM, Charles McCathieNevile
wrote:
Privacy and security restrictions leap to mind. There are things that
really are "should" requirements because there are valid use cases for
not applying them, and no rea
On Sun, Jul 10, 2011 at 3:59 PM, Charles McCathieNevile
wrote:
> Privacy and security restrictions leap to mind. There are things that really
> are "should" requirements because there are valid use cases for not applying
> them, and no reason to break those cases by making the requirement a "must"
On Wed, 06 Jul 2011 00:32:42 +0200, Aryeh Gregor
wrote:
Generally, if something is important enough for interop that we want
to test it, we don't want to make it a "should" requirement. It
should be a "must". What examples do you have of "should"
requirements that you want to test?
Privac
On Sun, 10 Jul 2011 20:01:17 +0200, timeless wrote:
On Sat, Jul 9, 2011 at 7:23 AM, Arthur Barstow
wrote:
Although there are ongoing discussions regarding exceptions, there were
no
objections to this CfC. As such, I will request publication of a LC
specification to encourage broader review
On Sat, Jul 9, 2011 at 7:23 AM, Arthur Barstow wrote:
> Although there are ongoing discussions regarding exceptions, there were no
> objections to this CfC. As such, I will request publication of a LC
> specification to encourage broader review and comments.
Sorry, I'm in the midst of sending com
