Re: What is a better word for "de-duplication"?

Abel Online wrote:

Jim Melton wrote:

I think that my serious suggestion a couple of days ago got lost because of my joke. The joke was "uniquify". The serious suggestion was "duplicate elimination". That is very clear, not ambiguous, easy to say, and doesn't depend on any vocabulary that requires further explanation.

I'd go for "duplicates elimination". At first I didn't really got it: did you mean eliminating things twice (you only live twice, so you can die twice?) but then I saw it: you meant eliminating duplicates! ;) Since duplicates never arive one at a time (they wouldn't be duplicate (n) if they were), I think the plural is in place here.

I doubt that as much as 5% of the users of XQuery, SQL, and even XSLT will be able to figure out what "node factorisation" means -- heck, even I'm uncertain why you think that the term is sufficiently meaningful to seriously suggest it.

I admit, I confess! I belong to the other 95 %. I thought factorisation had something to do with the exclamation mark in mathematics: 5! = 1*2*3*4*5 = 120. Or the other way around: finding unique factors gives you all prime numbers that multiplied together give you your query number (isn't that called factoring out a number?).

deduping is sexy librarin term for this...

perhaps there is some term from set theory with converting multisets(bags) to sets using lesser known symmetric difference operation with a 2 copies of itself (and another operation)...cant remember/think too late to google

in SQL/RDBMS world I think making something 'distinct' is clear enough

-- Jim Fuller