Re: Remove hacks for old bad qsort() implementations? - Mailing list pgsql-hackers

From Zeugswetter Andreas OSB SD
Subject Re: Remove hacks for old bad qsort() implementations?
Date
Msg-id E1539E0ED7043848906A8FF995BDA57902DD360B@m0143.s-mxs.net
Whole thread Raw
In response to Remove hacks for old bad qsort() implementations?  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: Remove hacks for old bad qsort() implementations?
List pgsql-hackers
> 2. If you've got lots of equal keys, it's conceivable that having the
> index entries sorted by TID offers some advantage in indexscan speed.
> I'm dubious that that's useful, mainly because the planner should
prefer
> a bitmap scan in such a case; and anyway the ordering is unlikely to
> be preserved for long.  But it's something to think about.

How about always adding the TID as last key when using qsort for create
index ?

Andreas


pgsql-hackers by date:

Previous
From: Mike Aubury
Date:
Subject: Request for feature - ECPGget_PGconn
Next
From: Gregory Stark
Date:
Subject: Re: New style of hash join proposal