|
Boost : |
From: Sérgio Vale e Pace (svpace.forum_at_[hidden])
Date: 2005-02-14 09:59:14
sorry for the short answer, but I was kind of sleepy. A couple a
months ago I was working in a project something like:
a group have several salesman that has several sales
in the interface one selects a group and a list od salesman shows up,
then one could select a salesman and it will be shown some statistical
analysis profile on this guy, or it could search for a specific
profile and the best match salesman get selected.
In my first draft design if a salesman get selected then I query for
his personal data, and all his sales and perform the analysis, on the
other hand if a profile is send I load all sales and perform analisys,
query for the most probables salesmans on the group and perform new
analysis. The problem is that the DB server was remote and centralized
so queries generate network trafic, disconnections, waits, angry users
and so on.
Then I redesigned the system to query for all salesman and sales in
the group (inside a group there are a resonable low number of then)
and perform the mix and matching in memory which was boring,
repetitive and error prone, and mostly easy to acomplish with good
relational algebra operations.
The real thing was a litle more complicated than that, but in a
nutshell that´s it Hope my history is of any help in rtl development,
basicaly I would love to be able to perform relational operations
togheter with other C++ operations. Also, I hope/wish to see
Boost.RTL/Rel ;) around soon;
On Sat, 12 Feb 2005 05:13:43 -0200, Sérgio Vale e Pace
<svpace.forum_at_[hidden]> wrote:
> right now it's just a general interest, but being able to solve
> relational algebra equations using in memory sets of data already came
> up as an issue in past projects. primarily when trying to avoid round
> trip queries to the database
>
> On Fri, 11 Feb 2005 10:22:08 -0500, Arkadiy Vertleyb
> <vertleyb_at_[hidden]> wrote:
> > "Sérgio Vale e Pace" <svpace.forum_at_[hidden]> wrote
> >
> > > well, if it counts for anything I would be very intersted in such library
> >
> > Of course it counts, and it is very much appreciated.
> >
> > Do you have any particular problem you want to apply it to, or is it just a
> > general interest?
> >
> > Regards,
> >
> > Arkadiy
> >
> > _______________________________________________
> > Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
> >
>
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk