[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [p2-dev] Match/Context naming
- From: Susan Franklin McCourt <susan_franklin@xxxxxxxxxx>
- Date: Wed, 24 Feb 2010 07:43:28 -0800
- Delivered-to: email@example.com
I have the same problem with BooleanQuery that you have with MatchQuery - both kinds of queries are boolean.
IsolationQuery - the answer can be determined in isolation
RelativeToCandidatesQuery - the answer depends on the candidates
I realize that second name is too long, but I would know exactly what it meant.
Thomas Hallgren ---02/23/2010 11:28:55 PM---On 02/24/2010 12:28 AM, Susan Franklin McCourt wrote:
On 02/24/2010 12:28 AM, Susan Franklin McCourt wrote:
I have problems with the term 'match' since I consider all queries to be
a filter that produce a result of instances that 'match'. Susan's
suggestion also shows this by adding the term to both queries. I think
the term itself thus becomes redundant. It doesn't help bringing clarity
into what it is that differentiates the two types.
To me, the term 'relative' suggests that the other query is 'absolute'
or perhaps 'concrete' which in turn seems a bit far fetched.
On 02/24/2010 03:33 AM, Simon Kaegi wrote:
> In SQL we have the various aggregate functions.
An aggregation to me is something that brings several things together
into one thing. Our 'context' _expression_ is applied on a collection and
produces a subset of that collection. Can that be called an aggregation?
Our 'match' query can always be turned into a 'context' query using the
<match each> -> everything.select(each | <match each>)
Perhaps I'm too focused on the _expression_ aspect but I'd like the names
to be explanatory so that the user intuitively knows which one to pick
for the purpose at hand. Here's another suggestion:
p2-dev mailing list