Re: A question about 'next'

[Enrico Colombini <erix@...>]

Mark Meijer wrote:
> I think (but I'm not sure) that, in principle, the order is never
> guaranteed. And in practice, the order is always the same as long as
> the table is not changed (and in certain cases even after the table is
> changed, but I don't think this is usefully predictable).
>
> Changes to tables inside tables (e.g. changes to "v", inside "t" in
> your example) should have no impact on the containing table ("t").

I think that too... but I'd like 'official' confirmation (or rebuttal).

I could save some processing time on my CPU-starved device by avoiding 
extra tables to guarantee traversal order, but over all I want to play 
it safe and avoid nasty surprises.

  Enrico