Re: Add statistics_collector_listen_addresses to fix hard-coding of "localhost"

On Fri, Oct 28, 2011 at 15:40, Robert Young <yayooo@gmail.com> wrote:
> Which wrong?
> 1.I got no money to buy a good machine to run both the services and datab=
ase.
> 2.I got no money to buy a good machine to run both the services and
> client applications.
> 3.Client applications hard-coding "localhost".
> 4.PG hard-coding "localhost".

Since They are equivalent,the answer is obvious:
Both #3 and #4 are wrong.

I admit, it is broken configuration.
But I said, broken configuration is just better than hard-coding.
Operating system designed flexible, give me the option to solve this
problem in broken configuration way.
Why database system wrote in hard-coding?

I admit, it is rare circumstance.
But I said, hard-coding is almost always right, NOT always right.
Just do the right thing is our rigorous attitude of work.

Database should be functional without underlying hostname or DNS facility.
So, I propose this patch to be applied.

On Fri, Oct 28, 2011 at 17:12, Robert Young <yayooo@gmail.com> wrote:
> It is client applications and services,NOT client applications and databa=
se.
> It just term's (client applications, services) misleading.
> To the system view,
> You should definitely known they are relationship between process and pro=
cess.
> Or I could still say some postgres process provide service,and some
> postgres process is client.
> PG process are just another couple of client and service.
> Why my client applications could not get so closer relationship with
> services,just like pg's client process and service's process ?
>
> Still,You got no knowledge about "client applications and services".
> What you said is your assumption.
> Without knowledge, you should consider them equivalent.
> PG got no priority.
>
> On Fri, Oct 28, 2011 at 16:35, Tom Lane <tgl@sss.pgh.pa.us> wrote:
>> Robert Young <yayooo@gmail.com> writes:
>>> You got no knowledge about "client applications".
>>> What you said is your assumption.
>>> Without knowledge, you should consider them equivalent.
>>> PG got no priority.
>>
>> Look, we will explain this once more. =C2=A0Postgres is entitled to assu=
me
>> that "localhost" means the local machine; there are Internet standards
>> saying so. =C2=A0On the other hand, client applications that assume the
>> database server is on the same machine they are on are definitely
>> broken, and need to be fixed.
>>
>> =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0regards, tom lane
>>
>