Re: ESC closes window without warning

On Thu, Feb 7, 2013 at 10:20 AM, Michal Kozusznik
<kozusznik(dot)michal(at)ifortuna(dot)cz> wrote:
> On 6.2.2013 18:19, Dave Page wrote:
>>
>> It won't be fixed in 1.16.2 I doubt, as it's a change in behaviour
>> that's been established for over 10 years with no previous complaints
>> that I can recall.
>
>
> I don't want to be rude, but maybe no one has used this feature due bugs
> appearing from time to time. pgAdmin is full of inconsistent resolutions
> established for over 10 years (in case of requesters, some appears even if
> are not needed at all why in other places there is no any attempt to confirm
> action). But this is not a reason to stay with it for next 10.
> It is quite clear that reported lack of confirmation, opens a way to lose
> the work. I have feeling that even you agree with this. So why the hell do
> not agree to improve it? You can even make it optional. I believe adding
> another checkbox to options tab is not impossible mission (sorry but some
> answers sounds like this)

You are missing the point. I have no problem changing this for the
next major version, but it's a change in functionality and requires
additional translation work, neither of which are things we do for
minor releases except where necessary to fix serious bugs - which this
is not.

> Thanx for the patch. It's presence proving that this feature is needed.
> However there are 2 issues:
> - most of pgAdmin users don't want to maintain own fork of pgadmin,
> recompile any new version applying patches etc

You have missed the point again. I'm not expecting anyone to apply the
patch themselves except other developers, but am presenting it as an
option for consideration. That's how open source work is done -
patches are written, and the pros/cons and alternate approaches
discussed before a solution is settled upon.

> - as you noted, the behavior is not quite correct. Confirmation should
> appear when value of any single form item has been changed (doesn't matter
> if dialog is happy with it or not). I believe it is not hard to implement
> such checking for any window.

Well that is exactly the point I was raising. I think it would be a
lot of work to add the check in that way (though it is undoubtedly the
best way to do it). If you have a simple way to make it work, I'd love
to see a proposed patch because at the moment, I don't see a way to do
it that wouldn't take at least a few days of work.

--
Dave Page
Blog: http://pgsnake.blogspot.com
Twitter: @pgsnake

EnterpriseDB UK: http://www.enterprisedb.com
The Enterprise PostgreSQL Company