Another Modest Proposal: Platforms

On 9/22/10 1:17 PM, David Fetter wrote:

While it's interesting to note, in an historical sense, that a
platform most recently updated when 1999 was still in the future, I
think it's time we did a little pruning.

It is unclear to me what problem you're trying to solve.

--
-- Josh Berkus
PostgreSQL Experts Inc.
http://www.pgexperts.com

On Wed, Sep 22, 2010 at 01:17:54PM -0700, David Fetter wrote:

Folks,

While it's interesting to note, in an historical sense, that a
platform most recently updated when 1999 was still in the future, I
think it's time we did a little pruning.

We can start by supporting only platforms git runs on, this being the
first in what I'd picture as a set of base requirements.

What say?

Cheers,
David.

Given the amount of trouble I had to get a git for a Solaris 8
system, I am not too keen on this definition for platform. PostgreSQL
runs very well on the same system, along with SVN and CVS.

Cheers,
Ken

On Wed, 2010-09-22 at 13:17 -0700, David Fetter wrote:

Folks,

While it's interesting to note, in an historical sense, that a
platform most recently updated when 1999 was still in the future, I
think it's time we did a little pruning.

We can start by supporting only platforms git runs on, this being the
first in what I'd picture as a set of base requirements.

What say?

I say that Bruce got Git to run on BSD/OS 4 or something like that. I
suggest that it won't matter what we say :P

JD

Cheers,
David.
--
David Fetter <david@fetter.org> http://fetter.org/
Phone: +1 415 235 3778 AIM: dfetter666 Yahoo!: dfetter
Skype: davidfetter XMPP: david.fetter@gmail.com
iCal: webcal://www.tripit.com/feed/ical/people/david74/tripit.ics

Remember to vote!
Consider donating to Postgres: http://www.postgresql.org/about/donate

--
PostgreSQL.org Major Contributor
Command Prompt, Inc: http://www.commandprompt.com/ - 509.416.6579
Consulting, Training, Support, Custom Development, Engineering
http://twitter.com/cmdpromptinc | http://identi.ca/commandprompt

David Fetter <david@fetter.org> writes:

We can start by supporting only platforms git runs on, this being the
first in what I'd picture as a set of base requirements.

Sounds like allowing the tail to wag the dog.

regards, tom lane

On Wed, Sep 22, 2010 at 4:17 PM, David Fetter <david@fetter.org> wrote:

What say?

"No." :-)

I'd be fine with dropping support for ancient platforms if it lets us
do something cool that we can't otherwise do, but there's no value in
doing it just because we can.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise Postgres Company

On 09/22/2010 04:38 PM, Robert Haas wrote:

On Wed, Sep 22, 2010 at 4:17 PM, David Fetter<david@fetter.org> wrote:

What say?

"No." :-)

I'd be fine with dropping support for ancient platforms if it lets us
do something cool that we can't otherwise do, but there's no value in
doing it just because we can.

Couldn't have said it better.

cheers

andrew

On Wed, Sep 22, 2010 at 04:28:04PM -0400, Tom Lane wrote:

David Fetter <david@fetter.org> writes:

We can start by supporting only platforms git runs on, this being
the first in what I'd picture as a set of base requirements.

Sounds like allowing the tail to wag the dog.

"Runs git" is actually not a bad proxy for "has modern developer
tools."

This would be the first, as I mentioned, of a list of base functional
requirements.

Cheers,
David.
--
David Fetter <david@fetter.org> http://fetter.org/
Phone: +1 415 235 3778 AIM: dfetter666 Yahoo!: dfetter
Skype: davidfetter XMPP: david.fetter@gmail.com
iCal: webcal://www.tripit.com/feed/ical/people/david74/tripit.ics

Remember to vote!
Consider donating to Postgres: http://www.postgresql.org/about/donate

On Wed, 2010-09-22 at 13:58 -0700, David Fetter wrote:

On Wed, Sep 22, 2010 at 04:28:04PM -0400, Tom Lane wrote:

David Fetter <david@fetter.org> writes:

We can start by supporting only platforms git runs on, this being
the first in what I'd picture as a set of base requirements.

Sounds like allowing the tail to wag the dog.

"Runs git" is actually not a bad proxy for "has modern developer
tools."

This would be the first, as I mentioned, of a list of base functional
requirements.

David,

Perhaps you could suggest some more specific ideas of your proposal? I
mean I am with you on the idea of reducing our supported platforms. I
see no reason to support

<Solaris 9
<NetBSD in any form
<FreeBSD <6
<Any Linux not supported by its own community (e.g; FC9)
<Irix
<> SCO

Note: I am not actually advocating this as much as stating my own
personal opinion.

Joshua D. Drake

--
PostgreSQL.org Major Contributor
Command Prompt, Inc: http://www.commandprompt.com/ - 509.416.6579
Consulting, Training, Support, Custom Development, Engineering
http://twitter.com/cmdpromptinc | http://identi.ca/commandprompt

On Wed, Sep 22, 2010 at 02:02:18PM -0700, Joshua D. Drake wrote:

On Wed, 2010-09-22 at 13:58 -0700, David Fetter wrote:

On Wed, Sep 22, 2010 at 04:28:04PM -0400, Tom Lane wrote:

David Fetter <david@fetter.org> writes:

We can start by supporting only platforms git runs on, this being
the first in what I'd picture as a set of base requirements.

Sounds like allowing the tail to wag the dog.

"Runs git" is actually not a bad proxy for "has modern developer
tools."

This would be the first, as I mentioned, of a list of base functional
requirements.

David,

Perhaps you could suggest some more specific ideas of your proposal? I
mean I am with you on the idea of reducing our supported platforms. I
see no reason to support

<Solaris 9
<NetBSD in any form
<FreeBSD <6
<Any Linux not supported by its own community (e.g; FC9)
<Irix
<> SCO

Note: I am not actually advocating this as much as stating my own
personal opinion.

Joshua D. Drake

It's not about naming platforms for exclusion. It's about requiring
functionalities for *in*clusion.

Cheers,
David.
--
David Fetter <david@fetter.org> http://fetter.org/
Phone: +1 415 235 3778 AIM: dfetter666 Yahoo!: dfetter
Skype: davidfetter XMPP: david.fetter@gmail.com
iCal: webcal://www.tripit.com/feed/ical/people/david74/tripit.ics

Remember to vote!
Consider donating to Postgres: http://www.postgresql.org/about/donate

On Wed, 2010-09-22 at 14:08 -0700, David Fetter wrote:

It's not about naming platforms for exclusion. It's about requiring
functionalities for *in*clusion.

I repeat:

Perhaps you could suggest some more specific ideas of your proposal?

I mean, it took us forever to require Perl 5.8.

Joshua D. Drake

--
PostgreSQL.org Major Contributor
Command Prompt, Inc: http://www.commandprompt.com/ - 509.416.6579
Consulting, Training, Support, Custom Development, Engineering
http://twitter.com/cmdpromptinc | http://identi.ca/commandprompt

On Sep 22, 2010, at 2:08 PM, David Fetter wrote:

It's not about naming platforms for exclusion. It's about requiring
functionalities for *in*clusion.

Passes all tests.

David

"Joshua D. Drake" <jd@commandprompt.com> writes:

I mean, it took us forever to require Perl 5.8.

... and we still make a point of not having a hard requirement for
that. If you don't want plperl, you can build from a tarball with
no perl at all.

Given the project history, I can't see us turning a dependency
we just added this week into a hard requirement anytime soon.

Now having said that, if you define "supported platform" to mean
"gets tested on the buildfarm", we do require Perl. And CVS,
which will soon get replaced by a requirement for Git. But I'm
not going to tell someone to get lost if they file a portability
bug report without having set up a buildfarm animal first.

regards, tom lane

On Wed, Sep 22, 2010 at 06:03:16PM -0400, Tom Lane wrote:

"Joshua D. Drake" <jd@commandprompt.com> writes:

I mean, it took us forever to require Perl 5.8.

... and we still make a point of not having a hard requirement for
that. If you don't want plperl, you can build from a tarball with
no perl at all.

Given the project history, I can't see us turning a dependency we
just added this week into a hard requirement anytime soon.

Now having said that, if you define "supported platform" to mean
"gets tested on the buildfarm", we do require Perl. And CVS, which
will soon get replaced by a requirement for Git. But I'm not going
to tell someone to get lost if they file a portability bug report
without having set up a buildfarm animal first.

I agree that "get lost" is not a reasonable first reaction, but as
with platforms like AIX, "It would help us enormously for you to put
up a buildfarm animal with your development environment on it" isn't.

Cheers,
David.
--
David Fetter <david@fetter.org> http://fetter.org/
Phone: +1 415 235 3778 AIM: dfetter666 Yahoo!: dfetter
Skype: davidfetter XMPP: david.fetter@gmail.com
iCal: webcal://www.tripit.com/feed/ical/people/david74/tripit.ics

Remember to vote!
Consider donating to Postgres: http://www.postgresql.org/about/donate

On Wed, Sep 22, 2010 at 6:56 PM, David Fetter <david@fetter.org> wrote:

On Wed, Sep 22, 2010 at 06:03:16PM -0400, Tom Lane wrote:

"Joshua D. Drake" <jd@commandprompt.com> writes:

I mean, it took us forever to require Perl 5.8.

... and we still make a point of not having a hard requirement for
that.  If you don't want plperl, you can build from a tarball with
no perl at all.

Given the project history, I can't see us turning a dependency we
just added this week into a hard requirement anytime soon.

Now having said that, if you define "supported platform" to mean
"gets tested on the buildfarm", we do require Perl.  And CVS, which
will soon get replaced by a requirement for Git.  But I'm not going
to tell someone to get lost if they file a portability bug report
without having set up a buildfarm animal first.

I agree that "get lost" is not a reasonable first reaction, but as
with platforms like AIX, "It would help us enormously for you to put
up a buildfarm animal with your development environment on it" isn't.

I feel like we do that already, as the occasion demands... so this
isn't really a change in policy from that point of view.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise Postgres Company