Versions RSS page is missing version(s)

(adding pgsql-www, isn't this more a www question than a general
postgresql usage question?)

2010/1/29 Greg Sabino Mullane <greg@turnstep.com>:

While looking into a failed check_postgres check, I found a problem
with the canonical versions page here:

http://www.postgresql.org/versions.rss

It only goes back to 8.0, but as far as I know, 7.4 is not unsupported

Hmm. Seems 7.4 was removed by Dave when we released 8.4:

https://pgweb.postgresql.org/changeset/2520

Dave, do you recall why?

yet, so that page should be listing 7.4.27. Further, shouldn't we be keeping
even 'unsupported' versions on this page, so (e.g. case of check_postgres.pl)
clients can check if they have the latest revision, even if the major/minor
combo is super old?

No, I don't think we should. We should list supported versions only.
And check_postgres could be advised to throw a warning at least if
you're running an unsupported version ;)

--
Magnus Hagander
Me: http://www.hagander.net/
Work: http://www.redpill-linpro.com/

Le 29/01/2010 15:51, Magnus Hagander a �crit :

(adding pgsql-www, isn't this more a www question than a general
postgresql usage question?)

2010/1/29 Greg Sabino Mullane <greg@turnstep.com>:

While looking into a failed check_postgres check, I found a problem
with the canonical versions page here:

http://www.postgresql.org/versions.rss

It only goes back to 8.0, but as far as I know, 7.4 is not unsupported

Hmm. Seems 7.4 was removed by Dave when we released 8.4:

https://pgweb.postgresql.org/changeset/2520

Dave, do you recall why?

Perhaps because you only display five releases in the "Latest Releases"
part of the front page? which means 8.0 to 8.4.

--
Guillaume.
http://www.postgresqlfr.org
http://dalibo.com

On Fri, Jan 29, 2010 at 2:53 PM, Guillaume Lelarge
<guillaume@lelarge.info> wrote:

Le 29/01/2010 15:51, Magnus Hagander a écrit :

(adding pgsql-www, isn't this more a www question than a general
postgresql usage question?)

2010/1/29 Greg Sabino Mullane <greg@turnstep.com>:

While looking into a failed check_postgres check, I found a problem
with the canonical versions page here:

http://www.postgresql.org/versions.rss

It only goes back to 8.0, but as far as I know, 7.4 is not unsupported

Hmm. Seems 7.4 was removed by Dave when we released 8.4:

https://pgweb.postgresql.org/changeset/2520

Dave, do you recall why?

Perhaps because you only display five releases in the "Latest Releases"
part of the front page? which means 8.0 to 8.4.

Yes. iirc, that was the reason.

--
Dave Page
EnterpriseDB UK: http://www.enterprisedb.com

On Fri, 2010-01-29 at 15:08 +0000, Dave Page wrote:

Perhaps because you only display five releases in the "Latest

Releases"

part of the front page? which means 8.0 to 8.4.

Yes. iirc, that was the reason.

Is it possible to add all (7.3+) versions to versions.rss, but show only
top 5 versions?

Regards,
--
Devrim GÜNDÜZ, RHCE
Command Prompt - http://www.CommandPrompt.com
devrim~gunduz.org, devrim~PostgreSQL.org, devrim.gunduz~linux.org.tr
http://www.gunduz.org Twitter: http://twitter.com/devrimgunduz

On Fri, 2010-01-29 at 17:49 +0200, Devrim GÜNDÜZ wrote:

On Fri, 2010-01-29 at 15:08 +0000, Dave Page wrote:

Perhaps because you only display five releases in the "Latest

Releases"

part of the front page? which means 8.0 to 8.4.

Yes. iirc, that was the reason.

Is it possible to add all (7.3+) versions to versions.rss, but show only
top 5 versions?

7.3 is not supported. It would be 7.4+ and even that, we need to put
that under a different category... Something like:

Old versions you shouldn't use:

8.0
7.4

Joshua D. Drake

Regards,

--
PostgreSQL.org Major Contributor
Command Prompt, Inc: http://www.commandprompt.com/ - 503.667.4564
Consulting, Training, Support, Custom Development, Engineering
Respect is earned, not gained through arbitrary and repetitive use or Mr. or Sir.

On Fri, 2010-01-29 at 08:44 -0800, Joshua D. Drake wrote:

Is it possible to add all (7.3+) versions to versions.rss, but show

only

top 5 versions?

7.3 is not supported.

I know. But it would be a chance to remind people to upgrade their
installations.
--
Devrim GÜNDÜZ, RHCE
Command Prompt - http://www.CommandPrompt.com
devrim~gunduz.org, devrim~PostgreSQL.org, devrim.gunduz~linux.org.tr
http://www.gunduz.org Twitter: http://twitter.com/devrimgunduz

2010/1/29 Devrim GÜNDÜZ <devrim@gunduz.org>:

On Fri, 2010-01-29 at 15:08 +0000, Dave Page wrote:

Perhaps because you only display five releases in the "Latest

Releases"

part of the front page? which means 8.0 to 8.4.

Yes. iirc, that was the reason.

Is it possible to add all (7.3+) versions to versions.rss, but show only
top 5 versions?

We certainly don't want 7.3 in there - that's not supported at all.

--
Dave Page
EnterpriseDB UK: http://www.enterprisedb.com

On Fri, 2010-01-29 at 18:50 +0200, Devrim GÜNDÜZ wrote:

On Fri, 2010-01-29 at 08:44 -0800, Joshua D. Drake wrote:

Is it possible to add all (7.3+) versions to versions.rss, but show

only

top 5 versions?

7.3 is not supported.

I know. But it would be a chance to remind people to upgrade their
installations.

An arbitrary listing won't do that.

Joshua D. Drake

--
PostgreSQL.org Major Contributor
Command Prompt, Inc: http://www.commandprompt.com/ - 503.667.4564
Consulting, Training, Support, Custom Development, Engineering
Respect is earned, not gained through arbitrary and repetitive use or Mr. or Sir.

-----BEGIN PGP SIGNED MESSAGE-----
Hash: RIPEMD160

yet, so that page should be listing 7.4.27. Further, shouldn't we be keeping
even 'unsupported' versions on this page, so (e.g. case of check_postgres.pl)
clients can check if they have the latest revision, even if the major/minor
combo is super old?

No, I don't think we should. We should list supported versions only.
And check_postgres could be advised to throw a warning at least if
you're running an unsupported version ;)

I'm not sure how useful that is. Surely while we encourage people to run
a recent major version, we also want to encourage people who will not
or cannot upgrade to at least be running the latest revision of a branch,
no matter how old it is?

How about a compromise? We add a new field to that XML so we can state
that it is unsupported, but leave it in there. That way, programs such
as check_postgres can not only distinguish between old but valid versions
and invalid versions (e.g. "7.typo.oops") but can act in a more intelligent
way for unsupported versions. Heck, maybe an estimated end-of-life date
field for all versions as well?

Either way, please add 7.4 back in. :)

- --
Greg Sabino Mullane greg@turnstep.com
End Point Corporation http://www.endpoint.com/
PGP Key: 0x14964AC8 201001291229
http://biglumber.com/x/web?pk=2529DF6AB8F79407E94445B4BC9B906714964AC8
-----BEGIN PGP SIGNATURE-----

iEYEAREDAAYFAktjHAoACgkQvJuQZxSWSsgCVQCguBaAiZQLoH9Q5nE4UEDymyp0
SQcAn1pf1rgEKDH6LrmmngBPtxketfA0
=XJcP
-----END PGP SIGNATURE-----

On Fri, 2010-01-29 at 09:25 -0800, Joshua D. Drake wrote:

I know. But it would be a chance to remind people to upgrade their
installations.

An arbitrary listing won't do that.

Depends.
--
Devrim GÜNDÜZ, RHCE
Command Prompt - http://www.CommandPrompt.com
devrim~gunduz.org, devrim~PostgreSQL.org, devrim.gunduz~linux.org.tr
http://www.gunduz.org Twitter: http://twitter.com/devrimgunduz

On Fri, Jan 29, 2010 at 18:34, Greg Sabino Mullane <greg@turnstep.com> wrote:

yet, so that page should be listing 7.4.27. Further, shouldn't we be keeping
even 'unsupported' versions on this page, so (e.g. case of check_postgres.pl)
clients can check if they have the latest revision, even if the major/minor
combo is super old?

No, I don't think we should. We should list supported versions only.
And check_postgres could be advised to throw a warning at least if
you're running an unsupported version ;)

I'm not sure how useful that is. Surely while we encourage people to run
a recent major version, we also want to encourage people who will not
or cannot upgrade to at least be running the latest revision of a branch,
no matter how old it is?

We don't support 7.3. Not even if you run the latest version.

How about a compromise? We add a new field to that XML so we can state
that it is unsupported, but leave it in there. That way, programs such
as check_postgres can not only distinguish between old but valid versions
and invalid versions (e.g. "7.typo.oops") but can act in a more intelligent
way for unsupported versions. Heck, maybe an estimated end-of-life date
field for all versions as well?

How do you add that field in a backwards compatible way? Meaning that
people or tools relying on it should *not* see 7.3 or 6.1 or whatever.
And it needs to be done within the RSS spec (which does allow custom
namespaces though, so that may not be a problem)

As for an estimated end-of-life, yes, we could definitely add that.
Now that we finally have it :-)

Either way, please add 7.4 back in. :)

Done, will be on in the next site rebuild.

--
Magnus Hagander
Me: http://www.hagander.net/
Work: http://www.redpill-linpro.com/

-----BEGIN PGP SIGNED MESSAGE-----
Hash: RIPEMD160

I'm not sure how useful that is. Surely while we encourage people to run
a recent major version, we also want to encourage people who will not
or cannot upgrade to at least be running the latest revision of a branch,
no matter how old it is?

We don't support 7.3. Not even if you run the latest version.

No, but I imagine we still would encourage people to run the latest revision
of it. Come this time next year, I hope that we'll tell people on 7.4.2 to
upgrade to 9.0 as soon as possible, but to upgrade to 7.4.27 *immaediately*.

How about a compromise? We add a new field to that XML so we can state
that it is unsupported, but leave it in there. That way, programs such
as check_postgres can not only distinguish between old but valid versions
and invalid versions (e.g. "7.typo.oops") but can act in a more intelligent
way for unsupported versions. Heck, maybe an estimated end-of-life date
field for all versions as well?

How do you add that field in a backwards compatible way? Meaning that
people or tools relying on it should *not* see 7.3 or 6.1 or whatever.
And it needs to be done within the RSS spec (which does allow custom
namespaces though, so that may not be a problem)

Well I don't know what people are reading the XML, so let's discuss tools.
Do you have a use case in mind where adding old versions would break something?
Has this always been advertised as a list of *supported* versions, or as a list
of the *latest* revisions? I've always assumed the latter was more important
that the former.

As for an estimated end-of-life, yes, we could definitely add that.
Now that we finally have it :-)

+1

Either way, please add 7.4 back in. :)

Done, will be on in the next site rebuild.

Thanks much, and thanks to Devrim for originally spotting the bug.

- --
Greg Sabino Mullane greg@turnstep.com
End Point Corporation http://www.endpoint.com/
PGP Key: 0x14964AC8 201002010931
http://biglumber.com/x/web?pk=2529DF6AB8F79407E94445B4BC9B906714964AC8
-----BEGIN PGP SIGNATURE-----

iEYEAREDAAYFAktm5hIACgkQvJuQZxSWSshVwwCeINTRgE7L5UWHJBIJgKDq3GIe
X/gAoOivHWlQaVI3nI+TWjUkwxTlicUx
=d+Yp
-----END PGP SIGNATURE-----

On Mon, Feb 1, 2010 at 15:33, Greg Sabino Mullane <greg@turnstep.com> wrote:

I'm not sure how useful that is. Surely while we encourage people to run
a recent major version, we also want to encourage people who will not
or cannot upgrade to at least be running the latest revision of a branch,
no matter how old it is?

We don't support 7.3. Not even if you run the latest version.

No, but I imagine we still would encourage people to run the latest revision
of it. Come this time next year, I hope that we'll tell people on 7.4.2 to

Do we really, officially, care?

upgrade to 9.0 as soon as possible, but to upgrade to 7.4.27 *immaediately*.

We should be, and afaik are, telling people to upgrade away from 7.4
immidiately *already*.

How about a compromise? We add a new field to that XML so we can state
that it is unsupported, but leave it in there. That way, programs such
as check_postgres can not only distinguish between old but valid versions
and invalid versions (e.g. "7.typo.oops") but can act in a more intelligent
way for unsupported versions. Heck, maybe an estimated end-of-life date
field for all versions as well?

How do you add that field in a backwards compatible way? Meaning that
people or tools relying on it should *not* see 7.3 or 6.1 or whatever.
And it needs to be done within the RSS spec (which does allow custom
namespaces though, so that may not be a problem)

Well I don't know what people are reading the XML, so let's discuss tools.
Do you have a use case in mind where adding old versions would break something?

I don't know what tools people use. That's the point of using RSS,
people can use whatever tool they want.

Has this always been advertised as a list of *supported* versions, or as a list
of the *latest* revisions? I've always assumed the latter was more important
that the former.

The *meaning* has always been supported versions, but if you read the
contents of the feed it does say latest.

Does anybody know if it's actually supported to have multiple channels
in one RSS feed? If it is, we could add a second channel with
unsupported versions, still listing the latest version of them.

--
Magnus Hagander
Me: http://www.hagander.net/
Work: http://www.redpill-linpro.com/

-----BEGIN PGP SIGNED MESSAGE-----
Hash: RIPEMD160

No, but I imagine we still would encourage people to run the latest revision
of it. Come this time next year, I hope that we'll tell people on 7.4.2 to

Do we really, officially, care?

Well, yes, we certainly should. Just because a branch is no longer officially
supported doesn't mean we want to discourage people from running the latest
available revision.

upgrade to 9.0 as soon as possible, but to upgrade to 7.4.27 *immaediately*.

We should be, and afaik are, telling people to upgrade away from 7.4
immidiately *already*.

Well, sure, but there's a world of difference from upgrading from 7.4.2 to
7.4.27 and from upgrading from 7.4.2 to 8.4.2.

The *meaning* has always been supported versions, but if you read the
contents of the feed it does say latest.

Well, I'd prefer to have the old versions, but I can handle the status quo. As
long nobody pulls versions before they are really dead again. :)

- --
Greg Sabino Mullane greg@turnstep.com
End Point Corporation http://www.endpoint.com/
PGP Key: 0x14964AC8 201002111128
http://biglumber.com/x/web?pk=2529DF6AB8F79407E94445B4BC9B906714964AC8
-----BEGIN PGP SIGNATURE-----

iEYEAREDAAYFAkt0MCUACgkQvJuQZxSWSsjKlQCgr1H+rp14YYhnByAOz2CaGqCp
+IAAoN00KX9OVwnxAOZIJpAyGgx6qCv1
=1D4+
-----END PGP SIGNATURE-----