Re: [COMMITTERS] pgsql: Use a bitmask to represent role attributes - Mailing list pgsql-hackers

From Stephen Frost
Subject Re: [COMMITTERS] pgsql: Use a bitmask to represent role attributes
Date
Msg-id 20141223184347.GK3062@tamriel.snowman.net
Whole thread Raw
In response to Re: [COMMITTERS] pgsql: Use a bitmask to represent role attributes  (Bruce Momjian <bruce@momjian.us>)
Responses Re: [COMMITTERS] pgsql: Use a bitmask to represent role attributes  (Bruce Momjian <bruce@momjian.us>)
Re: [COMMITTERS] pgsql: Use a bitmask to represent role attributes  (Tom Lane <tgl@sss.pgh.pa.us>)
Re: [COMMITTERS] pgsql: Use a bitmask to represent role attributes  (David G Johnston <david.g.johnston@gmail.com>)
List pgsql-hackers
* Bruce Momjian (bruce@momjian.us) wrote:
> I am with Tom on this --- there is more wasted space in the 'name'
> column pg_authid.rolname than by shoving 40 boolean values into a
> bitmap.

I agree that we waste a lot of space in 'name' but I don't think there's
an easy solution to that problem.  This, on the other hand, seemed like
a relatively minor change to an internal catalog definition.

> Adding the complexity of a bitmap doesn't make sense here.  I
> also apologize for the late feedback.

The complexity of the bitmap on the SQL side actually makes the code
side cleaner. :/

It would be great to figure out a way to get feedback like this earlier
on in the development.  This patch has been floating around for quite a
while, with intentional breaks for feedback taken prior to incremental
improvements and documentation additions.  Clearly that gets back to the
discussion around the commitfest situation.

> Offtopic, what I would really _love_ to see improved is our display of
> object permissions:

That's a whole different discussion and really belongs on a new thread.
I'm certainly curious what ideas you have regarding how to fix this;
it's not like Unix permission display is particularly elegant. Is there
something you've seen that looks nicer while still conveying the
necessary information in a relatively small amount of space?
Thanks,
    Stephen

pgsql-hackers by date:

Previous
From: Alvaro Herrera
Date:
Subject: Re: replicating DROP commands across servers
Next
From: José Luis Tallón
Date:
Subject: Re: Proposal: two new role attributes and/or capabilities?