Re: cdt glossary 0.1.1 (repost)

"Gene Wirchenko" <genew_at_ocis.net> wrote in message news:cobb231c8d4ooh12tjra9shgjhd48b80oq_at_4ax.com...
> "Brian Selzer" <brian_at_selzer-software.com> wrote:
>
>>"paul c" <toledobythesea_at_oohay.ac> wrote in message
>>news:%qgVh.89055$6m4.56691_at_pd7urf1no...
>
> [snip]
>
>>> Anyway, at the risk of being too long-winded, I must say that I object
>>> to
>>> definitions that imply how a phenomenon such as transactions must be
>>> implemented, even though I realize that many people these days don't
>>> consider a "definition" to be a definition unless it tells them what to
>>> do.
>
>>I'm not sure I understand your objection. I was just trying to point out
>>that a set of operations evaluated in one order may produce a different
>>result than the same set of operations evaluated in another. If the order
>>is specified, then there's no question which result was desired.
>
> I think the objection is the forcing: there should be a defined
> result of a transaction, but *how* it occurs is irrelevant. It might
> be: "The result of a transaction shall be as if the transaction
> statements were executed in the specified order." Note that this does
> not require that the statements be done in order, or even that they be
> done. The implementation might be able to optimise.
>

Good point. I'm not sure I agree, though, that *how* is irrelevant. The implementation is irrelevant (or should be), but the user's stated intent as to how the results are arrived at may not be. Even though the result is the same, a transaction consisting of a delete and an insert has different semantics than a transaction consisting of an update--the latter also implies that there's some relationship between the old and new values.

> [snip]
>
> Sincerely,
>
> Gene Wirchenko
>
> Computerese Irregular Verb Conjugation:
> I have preferences.
> You have biases.
> He/She has prejudices.
Received on Wed Apr 18 2007 - 14:21:00 CEST