Re: Error conditions

On Fri, Mar 09, 2007 at 01:35:31AM +0100, Lars Marius Garshol wrote:
>  (1) Should TMQL specify unambiguously which queries are erroneous?
>  (2) Should TMQL name all defined error situations?
> 
> By our requirements we have to do (1), so I assume that's not what  
> you mean.

No, of course not. It is (2) whereby I would probably drop the word
'all' above.

> >   From a language perspective this is not necessary, but OTOH, it
> >   impedes compatibility between TMQL processors as one application
> >   has to expect potentially different sets of exceptions.
> 
> Well. What good is a TMQL API that only standardizes the exception  
> classes and nothing else? That's effectively what this is.
> 
> >   Structured Discussion
> >
> >      ? Should the TMQL standard name all error situations
> >        + higher compatibility between TMQL implementation
> >        - less freedom for implementors
> 
> I think that doing this
> 
>  - adds a lot of effort to specifying the language,

A bit, yes. At least we would have a list (maybe 10?, just guessing)
errors named.

>  - causes problems for implementors, because error situations tend to
>    be highly dependent on implementation strategy, and

OK, that should actually not happen. If the query is valid, then a
processor MUST perform, regardless how it is implemented. 

> In short, I'm strongly in disfavour of this.

Certainly a reasonable position.

\rho