Re: First Impressions on Using Alphora's Dataphor
Date: 30 Aug 2004 02:59:49 -0700
Christopher Browne <cbbrowne_at_acm.org> wrote in message news:<2p7l7sFhrhjuU1_at_uni-berlin.de>...
> I'm not sure either is the ultimate answer, and it would _definitely_
> be an inconvenience to make a little domain change "over here" that
> leads to 200GB worth of tuples having to be revised "over there."
Alternatively, multiple possible representations. I must admit to not being a fan of these for the usually pushed purpose (polar vs. cartesian co-ordinates, say), but I could see a use for them in version control, such that as and when an update is required on a given tuple, the representation of the UDT in that tuple can be updated to the "newer" version (much like in some DBMSs "alter table ... add column..." adds the column as and when a row is updated or inserted).
In the ML or Haskell system, where you may opt to describe your UDT with an algebraic datatype, you might be able to add in extra cases, and add corresponding cases to the patterns for your operators. Type variables probably wouldn't be of much assistance to you here as you are intimately interested in the internal workings of your new datatype.