Re: vacuum_truncate configuration parameter and isset_offset

On 2025-Mar-26, Nikolay Shaplov wrote:

> git blame reloptions.c | grep "Robert Haas" | wc -l
> 112
>
> git blame reloptions.c | grep Álvaro | wc -l
> 792
>
> For me Alvary is the keeper of the concept of reloption. And he sees
> difference. If he would said isset_offset is good, I will follow. Until then I
> will follow my sense of harmony and code beauty as much as I can.

Well, as I said upthread, I wrote this code far too long ago and I don't
consider myself as its "owner" in the sense that I can block others from
moving it forward. I really wish I had had time to review and get your
new options patch to completion earlier -- we probably wouldn't be
having this discussion if I had. But that regretfully didn't happen.

I think the opinions have been sufficiently voiced, and I wouldn't want
to block Nathan from being able to make progress on the other projects
he has for this release. Given that continued discussion has not
convinced several committers that this is a change worth making, it
seems reasonable to give up on requiring that your proposed change be done.

I'd recommend you to take a break from it for a while. Maybe a design
solution will occur to you that will not break everything (for example:
some options allow a default value to be expressed, others allow the
"isset" flag to be used).

Thanks for your persistence

--
Álvaro Herrera 48°01'N 7°57'E — https://www.EnterpriseDB.com/
"[PostgreSQL] is a great group; in my opinion it is THE best open source
development communities in existence anywhere." (Lamar Owen)