Subject | Re: [Firebird-Architect] Re: Design of new built-in functions |
---|---|
Author | Jim Starkey |
Post date | 2006-05-07T02:18:40Z |
Leyne, Sean wrote:
I hope you wouldn't), there is the problem that new system functions
will mask (and break) existing user defined functions.
where the code lives?
introduces a new function that overlaps a site specific user function,
it makes all the sense in the world to change the configuration so the
new function either has a different name or disappears from the
namespace altogether.
--
Jim Starkey
Netfrastructure, Inc.
978 526-1376
>Roman,Unless you agree that additional system function can never be added (and
>
>
>
>>Clear, I prefer the second solution, where
>>"system-defined functions" cannot be overriden by the user.
>>
>>
>
>I also agree. System functions are never to be overridden by user
>functions.
>
>
I hope you wouldn't), there is the problem that new system functions
will mask (and break) existing user defined functions.
>That's a meaningless distinction. What earthly difference does it make
>
>
>>P.S. Just came to my mind: can it be that you suggest to make the list
>>of available system-defined functions (e.g. functions that are
>>available without declaration) configurable on per-database level and
>>provide a possibility to implement them in different languages?
>>
>>
>
>I disagree with this notion.
>
>An SDF is implemented within the engine, they are not external to the
>engine. As such, they can be implemented in any language which the
>engine build can support.
>
>
where the code lives?
>They are 'standard' as defined by the version of the Firebird engine,Pish-tosh. The set changes all the time without notice. If Firebird
>they do not vary by database or by platform.
>
>
>
introduces a new function that overlaps a site specific user function,
it makes all the sense in the world to change the configuration so the
new function either has a different name or disappears from the
namespace altogether.
--
Jim Starkey
Netfrastructure, Inc.
978 526-1376