[Standards-JIG] UPDATED: JEP-0059 (Result Set Management)

[Ian Paterson]

Peter wrote:
> Well, there's no real reason for <start/> then, is there? In 
> the first request don't specify <after/>, in all other 
> requests specify the unique ID (UID) of the last item.

Yes I agree. :-)

(I sometimes hold back from suggesting the full extent of a 'radical'
proposal in the first post.)

> BTW, the problem of omitted items is not necessarily solved 
> by this. Consider what happens if the list is ordered:

Yes, I *tried* to explain that disadvantage in one sentence in my last
post:

> > - *even if the above to errors do not occur, the complete 
> > result set returned would not necessarily reflect the 
> > result set as it existed at any one point in time

> Dynamic search seems to be ugly no matter how you slice it.

Yes, but the problem above is smaller than the ones presented by the
existing system. <after/> is the best solution I've been able to think
of.

- Ian