Re: binary protocol, again

From: "P(dot) Christeas" <xrg(at)linux(dot)gr>
To: Daniele Varrazzo <daniele(dot)varrazzo(at)gmail(dot)com>
Cc: psycopg(at)postgresql(dot)org
Subject: Re: binary protocol, again
Date: 2012-07-21 11:34:22
Message-ID: 201207211434.23200.xrg@linux.gr
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: psycopg

On Saturday 21 July 2012, Daniele Varrazzo wrote:
> Why
> doesn't the cursor_bin (as you call it) always use only PQexecParams
> and the ISQLParam protocol for adaptation?
> What is a case in which people would want to use the same cursor for
> the two param styles?
>
> No automatic switch between the two:...

In my case, I want an improved cursor, one that will simply be able to be
"switched on" on existing applications. It is optional anyway and may be
marked 'experimental' until its functionality is almost identical to the
existing text cursors.

Using different classes, different adapters and in-transaction decisions (based
on the detected psycopg2 version, too) inside the applications is IMHO much
more messy (and expensive) than doing an automatic failover in the C library.

--
Say NO to spam and viruses. Stop using Microsoft Windows!

In response to

Browse psycopg by date

  From Date Subject
Next Message Raheem Sarcar 2012-07-26 06:11:33 Inconsistent results when calculating "age" of db records
Previous Message Daniele Varrazzo 2012-07-20 23:57:06 Re: binary protocol, again