From: | Peter Geoghegan <peter(dot)geoghegan86(at)gmail(dot)com> |
---|---|
To: | Dave Page <dpage(at)pgadmin(dot)org> |
Cc: | pgadmin-hackers(at)postgresql(dot)org |
Subject: | Re: pgAdmin III commit: Revert the previous change that introduced sysSetti |
Date: | 2011-02-17 14:59:23 |
Message-ID: | AANLkTikYig59My-NJMeHbgV5KA9+tYGdpOc-pnL=_iZh@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgadmin-hackers |
Whoops. I had a false sense of security from ctlListView, where the
ambiguity between which is the preferred cast from the double and long
overloads prevents AppendItem(true) type calls from compiling at all.
I thought that there were no such calls to Write() overloads here -
however, there was an implicit cast to int that didn't occur with
ctlListView, which is what caused the problem.
I would like to produce a new patch where we deal with the problem
correctly, and create multiple, non-overloaded variants plus the
vanilla string function that is simply called write(), ala wx
perl/python bindings. That's probably not that hard to do, and is a
better, safer interface for us to use, considering that these calls
are fairly prevalent. Objections?
--
Regards,
Peter Geoghegan
From | Date | Subject | |
---|---|---|---|
Next Message | Dave Page | 2011-02-17 19:28:05 | Re: pgAdmin III commit: Revert the previous change that introduced sysSetti |
Previous Message | Dave Page | 2011-02-17 13:24:57 | Re: pgAdmin III commit: Revert the previous change that introduced sysSetti |