[cap-talk] Last Call for ref_send API 1.0

[Kevin Reid]

On Apr 10, 2007, at 12:34, Tyler Close wrote:
> On 4/10/07, Kevin Reid <kpreid at mac.com> wrote:
>> I think you're agreeing with me, but this phrasing doesn't really  
>> make it clear. I infer that by "that resolved state" you mean a  
>> state of being resolved /to a particular reference/ (that is, that  
>> being resolved to some other reference would be "a different  
>> resolved state"), but I wouldn't think that except for the context  
>> of my question above. I request that this be stated more clearly.
>
> Is the following javadoc clear on this point:

Yes.

>      * will forever be in the rejected state, with that reason.  
> After this
>      * initial state transition, all subsequent invocations of either
>      * {@link Resolver#resolve resolve} or {@link Resolver#reject  
> reject} are
>      * silently ignored. Any {@link Do observer} {@link #when  
> registered} on

Is there a reason this is silent? It seems to me that this would hide  
mistakes.

E resolvers offer "resolve", which throws if already resolved, and  
"resolveRace", which returns a boolean indicating whether it wasn't  
already resolved.

-- 
Kevin Reid                            <http://homepage.mac.com/kpreid/>