Re: sequences and "addval('myseq', value)"

> On 6/8/2004 11:46 AM, pgsql(at)mohawksoft(dot)com wrote:
>
>>>
>>> This strikes me as a complete nonstarter.
>>
>> Tom, I have to chuckle here. You HATE every suggestion I ever make. I
>> can't think of one thing I've suggested over the years that was ever met
>> with enthusiasm. Never change. :-)
>
> I happen to agree with Tom on this entire thread. I do not think that
> sequences should be abused as a replacement for global shared variables.
>
> I do think that PostgreSQL should instead have some sort of shared
> memory segment for user variables. The accessor functions would place
> locks and the like, and would have mechanisms like UNDO on rollback.
> With all the limitations that has, inflexible size, garbage collection
> and so on, it would yet be far superior to anything you've come up with
> here.

My original suggestion was to have some sort of global variable system. I
thought using an existing construct would have been more palletable.

I was wrong.
>
>
> Jan
>
> --
> #======================================================================#
> # It's easier to get forgiveness for being wrong than for being right. #
> # Let's break this rule - forgive me. #
> #================================================== JanWieck(at)Yahoo(dot)com #
>