Re: virtual nnvirtual

[Eric Abrahamsen <eric@ericabrahamsen.net>]

Eric S Fraga <e.fraga@ucl.ac.uk> writes:

> Hi Andrew,
>
> thank you for the long response to my queries.  Very helpful.
>
> Regarding /o, now it makes sense (the cycling nature of the retrieval)
> and, yes, it seems to be working just fine.  Because some of the groups
> in the virtual collection have not had any recent activity, I was
> confused by seeing significantly old articles in the list provided.  I
> now know that I need to multiply the number of messages, M, to retrieve
> by N, where N is the number of groups in the virtual collection, to
> ensure I get the M most recent emails I'm looking for.  Simple enough,
> especially if I then sort by date.  Thank you for explaining this!

As Andy mentioned, this is also kind of a philosophical/design question.
We can choose how this algorithm works, it just happens to work this way
right now. Personally I would find it less surprising if you didn't get
any messages from long-inactive groups until you "dug all the way back"
in time to them, but that would require a different code approach (and
probably would be less efficient).