Re: Proposal: Commit timestamp

Jan Wieck wrote:
> On 2/7/2007 10:35 PM, Bruce Momjian wrote:
> > I find the term "logical proof of it's correctness" too restrictive. It
> > sounds like some formal academic process that really doesn't work well
> > for us.
>
> Thank you.
>
> > Also, I saw the trigger patch with no explaination of why it was
> > important or who would use it --- that also isn't going to fly well.
>
> You didn't respond to my explanation how the current Slony
> implementation could improve and evolve using it. Are you missing
> something? I am discussing this very issue with our own QA department,
> and thus far, I think I have a majority of "would use a pg_trigger
> backpatched PostgreSQL" vs. "No, I prefer a system that knows exactly
> how it corrupted my system catalog".

No, I _now_ understand the use case, but when the patch was posted, the
use case was missing. I would like to see a repost with the patch, and
a description of its use so we can all move forward on that.

> > As far as TOAST, there is no question in my mind that TOAST development
> > would happen the same way today as it did when we did it in 2001 --- we
> > have a problem, how can we fix it.
>
> Looking at what did happen back then and what happens in this case, I do
> see a difference. There were concerns about the compression algorithm
> used ... it still is today what was the first incarnation and nobody
> ever bothered to even investigate if there could possibly be any better
> thing. Do you think lzcompress is the best we can come up with? I don't!
> So why is it still the thing used? Maybe it is good enough?

It is simple/stupid enough, I would say, and the compression space is a
mine-field of patents.

--
Bruce Momjian <bruce(at)momjian(dot)us> http://momjian.us
EnterpriseDB http://www.enterprisedb.com

+ If your life is a hard drive, Christ can be your backup. +