pgsql: Remove pre-7.4 documentaiton mentions, now that 8.0 is the oldest

Bruce Momjian wrote:

Log Message:
-----------
Remove pre-7.4 documentaiton mentions, now that 8.0 is the oldest
supported release.

per http://wiki.postgresql.org/wiki/PostgreSQL_Release_Support_Policy
7.4 is still supported for a few months to come (and will be EOL'd
together with 8.0). I'm also not really sure why we need to change stuff
like that, this kind of information might still be useful for somebody
trying to upgrade from an unsupported release to a supported one.

Stefan

On Wed, Feb 24, 2010 at 5:03 AM, Stefan Kaltenbrunner
<stefan@kaltenbrunner.cc> wrote:

Bruce Momjian wrote:

Log Message:
-----------
Remove pre-7.4 documentaiton mentions, now that 8.0 is the oldest
supported release.

per http://wiki.postgresql.org/wiki/PostgreSQL_Release_Support_Policy 7.4 is
still supported for a few months to come (and will be EOL'd together with
8.0). I'm also not really sure why we need to change stuff like that, this
kind of information might still be useful for somebody trying to upgrade
from an unsupported release to a supported one.

Yeah.

...Robert

Robert Haas wrote:

On Wed, Feb 24, 2010 at 5:03 AM, Stefan Kaltenbrunner
<stefan@kaltenbrunner.cc> wrote:

Bruce Momjian wrote:

Log Message:
-----------
Remove pre-7.4 documentaiton mentions, now that 8.0 is the oldest
supported release.

per http://wiki.postgresql.org/wiki/PostgreSQL_Release_Support_Policy 7.4 is
still supported for a few months to come (and will be EOL'd together with
8.0). I'm also not really sure why we need to change stuff like that, this
kind of information might still be useful for somebody trying to upgrade
from an unsupported release to a supported one.

Yeah.

Well, the documentation still exists in the old releases, even 8.4. The
big question is how much back-version information we should keep in our
docs, and does it make sense to keep paragraphs around that are only
meaningful to < 1% of people reading it. Some people are saying keep
more, some are saying keep less, so I am betting I have hit the proper
balance. ;-)

-- 
  Bruce Momjian  <bruce@momjian.us>        http://momjian.us
  EnterpriseDB                             http://enterprisedb.com
  PG East:  http://www.enterprisedb.com/community/nav-pg-east-2010.do
  + If your life is a hard drive, Christ can be your backup. +

Bruce Momjian wrote:

Robert Haas wrote:

On Wed, Feb 24, 2010 at 5:03 AM, Stefan Kaltenbrunner
<stefan@kaltenbrunner.cc> wrote:

Bruce Momjian wrote:

Log Message:
-----------
Remove pre-7.4 documentaiton mentions, now that 8.0 is the oldest
supported release.

per http://wiki.postgresql.org/wiki/PostgreSQL_Release_Support_Policy 7.4 is
still supported for a few months to come (and will be EOL'd together with
8.0). I'm also not really sure why we need to change stuff like that, this
kind of information might still be useful for somebody trying to upgrade
from an unsupported release to a supported one.

Yeah.

Well, the documentation still exists in the old releases, even 8.4. The
big question is how much back-version information we should keep in our
docs, and does it make sense to keep paragraphs around that are only
meaningful to < 1% of people reading it. Some people are saying keep
more, some are saying keep less, so I am betting I have hit the proper
balance. ;-)

Well even if it is useful information for only 1% of our readers (which
given the access stats on the html logs is a HUGE number) and we don't
have any real maintenance overhead with keeping it (which I kinda doubt
we have).

And just from looking at some of the hunks in more detail I think we are
actually removing fairly reasonable information (especially if we are
talking stuff like behaviour changes) :/

Stefan

Stefan Kaltenbrunner wrote:

Bruce Momjian wrote:

Robert Haas wrote:

On Wed, Feb 24, 2010 at 5:03 AM, Stefan Kaltenbrunner
<stefan@kaltenbrunner.cc> wrote:

Bruce Momjian wrote:

Log Message:
-----------
Remove pre-7.4 documentaiton mentions, now that 8.0 is the oldest
supported release.

per http://wiki.postgresql.org/wiki/PostgreSQL_Release_Support_Policy 7.4 is
still supported for a few months to come (and will be EOL'd together with
8.0). I'm also not really sure why we need to change stuff like that, this
kind of information might still be useful for somebody trying to upgrade
from an unsupported release to a supported one.

Yeah.

Well, the documentation still exists in the old releases, even 8.4. The
big question is how much back-version information we should keep in our
docs, and does it make sense to keep paragraphs around that are only
meaningful to < 1% of people reading it. Some people are saying keep
more, some are saying keep less, so I am betting I have hit the proper
balance. ;-)

Well even if it is useful information for only 1% of our readers (which
given the access stats on the html logs is a HUGE number) and we don't
have any real maintenance overhead with keeping it (which I kinda doubt
we have).

And just from looking at some of the hunks in more detail I think we are
actually removing fairly reasonable information (especially if we are
talking stuff like behaviour changes) :/

The issue isn't the cost of us maintaining the SGML, it is the cost of
individuals reading it, deciding if it applies to them, and then
skipping to the next paragraph or section. If we had all this stuff in
one isolated place, we would not have that issue and we could leave it
there forever, but then few people would find it when they needed it.

Also, this stuff is all in the release notes too, which is what people
should be reading for ugprades anyway. The question is what should be
in the general documentation that everyone reads.

-- 
  Bruce Momjian  <bruce@momjian.us>        http://momjian.us
  EnterpriseDB                             http://enterprisedb.com
  PG East:  http://www.enterprisedb.com/community/nav-pg-east-2010.do
  + If your life is a hard drive, Christ can be your backup. +

Bruce Momjian <bruce@momjian.us> writes:

Robert Haas wrote:

On Wed, Feb 24, 2010 at 5:03 AM, Stefan Kaltenbrunner
<stefan@kaltenbrunner.cc> wrote:

per http://wiki.postgresql.org/wiki/PostgreSQL_Release_Support_Policy 7.4 is
still supported for a few months to come (and will be EOL'd together with
8.0). I'm also not really sure why we need to change stuff like that, this
kind of information might still be useful for somebody trying to upgrade
from an unsupported release to a supported one.

Yeah.

Well, the documentation still exists in the old releases, even 8.4. The
big question is how much back-version information we should keep in our
docs, and does it make sense to keep paragraphs around that are only
meaningful to < 1% of people reading it. Some people are saying keep
more, some are saying keep less, so I am betting I have hit the proper
balance. ;-)

I didn't really agree with what you took out before, and I am definitely
going to object to this latest set of diffs. What it appears to me
you have done is a search-and-destroy on any paragraph mentioning "7.x",
without any consideration of whether that removes important information
from the overall presentation. Those paras are generally comparing old
and new behavior, and even if you don't care specifically what the old
behavior was, they present useful explanation of the new behavior.

I also agree with the objection that there are still lots of people who
are going to be trying to port old apps to 9.0.

regards, tom lane

Tom Lane wrote:

Bruce Momjian <bruce@momjian.us> writes:

Robert Haas wrote:

On Wed, Feb 24, 2010 at 5:03 AM, Stefan Kaltenbrunner
<stefan@kaltenbrunner.cc> wrote:

per http://wiki.postgresql.org/wiki/PostgreSQL_Release_Support_Policy 7.4 is
still supported for a few months to come (and will be EOL'd together with
8.0). I'm also not really sure why we need to change stuff like that, this
kind of information might still be useful for somebody trying to upgrade
from an unsupported release to a supported one.

Yeah.

Well, the documentation still exists in the old releases, even 8.4. The
big question is how much back-version information we should keep in our
docs, and does it make sense to keep paragraphs around that are only
meaningful to < 1% of people reading it. Some people are saying keep
more, some are saying keep less, so I am betting I have hit the proper
balance. ;-)

I didn't really agree with what you took out before, and I am definitely
going to object to this latest set of diffs. What it appears to me
you have done is a search-and-destroy on any paragraph mentioning "7.x",
without any consideration of whether that removes important information
from the overall presentation. Those paras are generally comparing old
and new behavior, and even if you don't care specifically what the old
behavior was, they present useful explanation of the new behavior.

Yea, let me try again and rephrase some of it to highlight the behavior
and not the version change.

I also agree with the objection that there are still lots of people who
are going to be trying to port old apps to 9.0.

Well, I stand by my statement that it is a judgement call on how much we
keep, and there is a cost to readers to keep it, but there isn't very
much of it. Are the people who wanted more aggressive removal OK with
putting back the pre-7.4 documentation mentions?

-- 
  Bruce Momjian  <bruce@momjian.us>        http://momjian.us
  EnterpriseDB                             http://enterprisedb.com
  PG East:  http://www.enterprisedb.com/community/nav-pg-east-2010.do
  + If your life is a hard drive, Christ can be your backup. +

On Wed, 24 Feb 2010, Bruce Momjian wrote:

Well, I stand by my statement that it is a judgement call on how much we
keep, and there is a cost to readers to keep it, but there isn't very
much of it. Are the people who wanted more aggressive removal OK with
putting back the pre-7.4 documentation mentions?

Why are those that don't want it determining what those that do want it
have access to? Easier for those that don't want it to 'skip it' then it
is for those that do to go searching in older releases for it ...

----
Marc G. Fournier Hub.Org Hosting Solutions S.A.
scrappy@hub.org http://www.hub.org

Yahoo:yscrappy Skype: hub.org ICQ:7615664 MSN:scrappy@hub.org

Bruce Momjian wrote:

Tom Lane wrote:

Bruce Momjian <bruce@momjian.us> writes:

Robert Haas wrote:

On Wed, Feb 24, 2010 at 5:03 AM, Stefan Kaltenbrunner
<stefan@kaltenbrunner.cc> wrote:

per http://wiki.postgresql.org/wiki/PostgreSQL_Release_Support_Policy 7.4 is
still supported for a few months to come (and will be EOL'd together with
8.0). I'm also not really sure why we need to change stuff like that, this
kind of information might still be useful for somebody trying to upgrade
from an unsupported release to a supported one.

Yeah.

Well, the documentation still exists in the old releases, even 8.4. The
big question is how much back-version information we should keep in our
docs, and does it make sense to keep paragraphs around that are only
meaningful to < 1% of people reading it. Some people are saying keep
more, some are saying keep less, so I am betting I have hit the proper
balance. ;-)

I didn't really agree with what you took out before, and I am definitely
going to object to this latest set of diffs. What it appears to me
you have done is a search-and-destroy on any paragraph mentioning "7.x",
without any consideration of whether that removes important information
from the overall presentation. Those paras are generally comparing old
and new behavior, and even if you don't care specifically what the old
behavior was, they present useful explanation of the new behavior.

Yea, let me try again and rephrase some of it to highlight the behavior
and not the version change.

Well the behaviour changed with a given version which is crucial
information for somebody doing a migration... It is also useful
historical information for people reading the manual - it is not
impossible that this could effect on the application design...

I also agree with the objection that there are still lots of people who
are going to be trying to port old apps to 9.0.

Well, I stand by my statement that it is a judgement call on how much we
keep, and there is a cost to readers to keep it, but there isn't very
much of it. Are the people who wanted more aggressive removal OK with
putting back the pre-7.4 documentation mentions?

Who actually are those people? I don't recall anybody complaining that
we have too much information in our docs (maybe that they wnat better
search or a better structure).

Stefan

Stefan Kaltenbrunner wrote:

Yea, let me try again and rephrase some of it to highlight the behavior
and not the version change.

Well the behaviour changed with a given version which is crucial
information for somebody doing a migration... It is also useful
historical information for people reading the manual - it is not
impossible that this could effect on the application design...

I also agree with the objection that there are still lots of people who
are going to be trying to port old apps to 9.0.

Well, I stand by my statement that it is a judgement call on how much we
keep, and there is a cost to readers to keep it, but there isn't very
much of it. Are the people who wanted more aggressive removal OK with
putting back the pre-7.4 documentation mentions?

Who actually are those people? I don't recall anybody complaining that
we have too much information in our docs (maybe that they wnat better
search or a better structure).

Well, by that argument, should we have Postgres 6.3 information in our
documentation? I doubt anyone would explicitly complain about it, but
it would serve very little useful purpose and make our documentation
harder to read.

I don't really care if we remove the old stuff or not --- removing it,
or at least reviewing possible removal stuff, is a standard practice for
every major relesae, so I did it. If people want nothing removed, that
is fine with me.

In fact, I have heard enough complaints. I am reversing my removals and
if someone else wants to do the job, go ahead.

-- 
  Bruce Momjian  <bruce@momjian.us>        http://momjian.us
  EnterpriseDB                             http://enterprisedb.com
  PG East:  http://www.enterprisedb.com/community/nav-pg-east-2010.do
  + If your life is a hard drive, Christ can be your backup. +

Bruce Momjian <bruce@momjian.us> writes:

The issue isn't the cost of us maintaining the SGML, it is the cost of
individuals reading it, deciding if it applies to them, and then
skipping to the next paragraph or section. If we had all this stuff in
one isolated place, we would not have that issue and we could leave it
there forever, but then few people would find it when they needed it.

That's a pretty lame argument when you are talking about removing
a few paragraphs out of 7.5MB (and counting) of documentation.
The incremental savings for readers who don't care is negligible. The
incremental cost for readers who needed the info could be very high.

regards, tom lane