Re: ISSUE: sequences need more explanation

Hi Shawn,

I like your suggestion, and will add text along these lines to the next 
Working Draft.

Jonathan

At 10:38 AM 1/14/2002 -0500, Shawn Vincent wrote:

>Summary: Please add verbiage to the spec clarifying important points
>          about sequences.
>
>It is clear that a lot of thought and effort went into creating a
>model for sequences that is usable and yet supports its XML legacy.
>
>In particular:
>
>    1. An item is IDENTICAL to a sequence of length 1 containing that
>       item.
>
>       That is, 1 is the same as (1).
>
>    2. There are no nested sequences.
>
>       That is, (1, (2, 3), ()) is the same as (1, 2, 3)
>
>    3. The empty sequence is often used as a nil/null value.
>
>These points (and especially the first two) are vital to understanding
>and using XQuery.
>
>However, it took me a long time to understand these points, as they
>are not clearly and explicitly stated, and this sequence model is
>different from most other languages (it has similarities to APL,
>perhaps).
>
>I'd propose adding some verbiage to the spec calling out these points
>explicitly.  This would ideally be done in the section that defines
>sequences, singleton sequences, etc.
>
>Thank you for your time.
>
>       -Shawn.
>
>
>--
>Shawn Vincent, svincent@exalt.com
>Exalt Inc.

Received on Wednesday, 23 January 2002 13:11:36 UTC