Release versioning inconsistency

Started by Marti Raudseppover 13 years ago12 messages
#1Marti Raudsepp
marti@juffo.org

Hi list,

The recent 9.2 beta releases have used a slightly different numbering
scheme than all previous releases.

It used to be that tarballs for version $VER were always available at:
http://ftp.postgresql.org/pub/source/v$VER/postgresql-$VER.tar.bz2

However, the new releases now use "v9.2.0beta2" for the directory
name, but "9.2beta2" in the tarball file. No big deal for most people,
but it will confuse people who have scripts to download PostgreSQL
tarballs automatically (e.g. packagers).

For example:
http://ftp.postgresql.org/pub/source/v9.2.0beta2/postgresql-9.2beta2.tar.bz2

Is there any reason behind this change?

Regards,
Marti

#2Magnus Hagander
magnus@hagander.net
In reply to: Marti Raudsepp (#1)
Re: Release versioning inconsistency

On Wed, Jun 20, 2012 at 10:28 AM, Marti Raudsepp <marti@juffo.org> wrote:

Hi list,

The recent 9.2 beta releases have used a slightly different numbering
scheme than all previous releases.

It used to be that tarballs for version $VER were always available at:
 http://ftp.postgresql.org/pub/source/v$VER/postgresql-$VER.tar.bz2

However, the new releases now use "v9.2.0beta2" for the directory
name, but "9.2beta2" in the tarball file. No big deal for most people,
but it will confuse people who have scripts to download PostgreSQL
tarballs automatically (e.g. packagers).

For example:
 http://ftp.postgresql.org/pub/source/v9.2.0beta2/postgresql-9.2beta2.tar.bz2

Is there any reason behind this change?

Not behind the first one. I believe that's just me and a bad memory -
I got it wrong. (I do believe that using the v9.2.0beta marker is
*better*, because then it sorts properly. But likely not enough much
better to be inconsistent with previous versions)

For beta2, the only reason was to keep it consistent with beta1.

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

#3Marti Raudsepp
marti@juffo.org
In reply to: Magnus Hagander (#2)
Re: Release versioning inconsistency

On Wed, Jun 20, 2012 at 12:18 PM, Magnus Hagander <magnus@hagander.net> wrote:

(I do believe that using the v9.2.0beta marker is
*better*, because then it sorts properly. But likely not enough much
better to be inconsistent with previous versions)

Good point. Maybe that's a reason to change the versioning scheme and
stick with "9.2.0betaX" everywhere. Including calling the final
release "9.2.0" instead of simply "9.2"?

Regards,
Marti

#4Magnus Hagander
magnus@hagander.net
In reply to: Marti Raudsepp (#3)
Re: Release versioning inconsistency

On Wed, Jun 20, 2012 at 11:23 AM, Marti Raudsepp <marti@juffo.org> wrote:

On Wed, Jun 20, 2012 at 12:18 PM, Magnus Hagander <magnus@hagander.net> wrote:

(I do believe that using the v9.2.0beta marker is
*better*, because then it sorts properly. But likely not enough much
better to be inconsistent with previous versions)

Good point. Maybe that's a reason to change the versioning scheme and
stick with "9.2.0betaX" everywhere. Including calling the final
release "9.2.0" instead of simply "9.2"?

That might actually be a good idea. We can't really change the way we
named the betas, but it's not too late to consider naming the actual
release as 9.2.0...

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

#5Dickson S. Guedes
listas@guedesoft.net
In reply to: Magnus Hagander (#4)
Re: Release versioning inconsistency

2012/6/20 Magnus Hagander <magnus@hagander.net>:

On Wed, Jun 20, 2012 at 11:23 AM, Marti Raudsepp <marti@juffo.org> wrote:

On Wed, Jun 20, 2012 at 12:18 PM, Magnus Hagander <magnus@hagander.net> wrote:

(I do believe that using the v9.2.0beta marker is
*better*, because then it sorts properly. But likely not enough much
better to be inconsistent with previous versions)

Good point. Maybe that's a reason to change the versioning scheme and
stick with "9.2.0betaX" everywhere. Including calling the final
release "9.2.0" instead of simply "9.2"?

That might actually be a good idea. We can't really change the way we
named the betas, but it's not too late to consider naming the actual
release as 9.2.0...

May be a symlink could be created just do fit the same pattern that other
versions do and keeps the actual links (for beta) working.

I'm using the same pattern in `pgvm` [1]https://github.com/guedes/pgvm/blob/master/include/sites and it is failing to fetch
beta versions :(

[1]: https://github.com/guedes/pgvm/blob/master/include/sites

regards
--
Dickson S. Guedes
mail/xmpp: guedes@guedesoft.net - skype: guediz
http://guedesoft.net - http://www.postgresql.org.br

#6Peter Eisentraut
peter_e@gmx.net
In reply to: Magnus Hagander (#4)
Re: Release versioning inconsistency

On ons, 2012-06-20 at 13:26 +0200, Magnus Hagander wrote:

On Wed, Jun 20, 2012 at 11:23 AM, Marti Raudsepp <marti@juffo.org> wrote:

On Wed, Jun 20, 2012 at 12:18 PM, Magnus Hagander <magnus@hagander.net> wrote:

(I do believe that using the v9.2.0beta marker is
*better*, because then it sorts properly. But likely not enough much
better to be inconsistent with previous versions)

Good point. Maybe that's a reason to change the versioning scheme and
stick with "9.2.0betaX" everywhere. Including calling the final
release "9.2.0" instead of simply "9.2"?

That might actually be a good idea. We can't really change the way we
named the betas, but it's not too late to consider naming the actual
release as 9.2.0...

The final release was always going to be called 9.2.0, but naming the
beta 9.2.0betaX is wrong. There was a previous discussion about that
particular point.

#7Tom Lane
tgl@sss.pgh.pa.us
In reply to: Peter Eisentraut (#6)
Re: Release versioning inconsistency

Peter Eisentraut <peter_e@gmx.net> writes:

On ons, 2012-06-20 at 13:26 +0200, Magnus Hagander wrote:

That might actually be a good idea. We can't really change the way we
named the betas, but it's not too late to consider naming the actual
release as 9.2.0...

The final release was always going to be called 9.2.0, but naming the
beta 9.2.0betaX is wrong. There was a previous discussion about that
particular point.

Yes. There is no reason to change the naming scheme we have been using
for years now (at least since version_stamp.pl was invented for 7.4).
The only problem is that somebody got the name of the directory wrong on
the FTP server.

regards, tom lane

#8Magnus Hagander
magnus@hagander.net
In reply to: Tom Lane (#7)
Re: Release versioning inconsistency

On Wed, Jun 20, 2012 at 5:20 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:

Peter Eisentraut <peter_e@gmx.net> writes:

On ons, 2012-06-20 at 13:26 +0200, Magnus Hagander wrote:

That might actually be a good idea. We can't really change the way we
named the betas, but it's not too late to consider naming the actual
release as 9.2.0...

The final release was always going to be called 9.2.0, but naming the
beta 9.2.0betaX is wrong.  There was a previous discussion about that
particular point.

Yes.  There is no reason to change the naming scheme we have been using
for years now (at least since version_stamp.pl was invented for 7.4).
The only problem is that somebody got the name of the directory wrong on
the FTP server.

If that wasn't clear, then yes, that was me.

I don't recall the reason why using 9.2.0betax was actually wrong - i
realize that's not the name of the version, so thereby the directory
was wrong. But in what way would it be wrong to call the version that?
Given that it would help with sorting. (And yes, this is a very
long-forward question, more about 9.3, since we can't really go back
and change the current filename..)

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

#9Magnus Hagander
magnus@hagander.net
In reply to: Dickson S. Guedes (#5)
Re: Release versioning inconsistency

On Wed, Jun 20, 2012 at 1:35 PM, Dickson S. Guedes <listas@guedesoft.net> wrote:

2012/6/20 Magnus Hagander <magnus@hagander.net>:

On Wed, Jun 20, 2012 at 11:23 AM, Marti Raudsepp <marti@juffo.org> wrote:

On Wed, Jun 20, 2012 at 12:18 PM, Magnus Hagander <magnus@hagander.net> wrote:

(I do believe that using the v9.2.0beta marker is
*better*, because then it sorts properly. But likely not enough much
better to be inconsistent with previous versions)

Good point. Maybe that's a reason to change the versioning scheme and
stick with "9.2.0betaX" everywhere. Including calling the final
release "9.2.0" instead of simply "9.2"?

That might actually be a good idea. We can't really change the way we
named the betas, but it's not too late to consider naming the actual
release as 9.2.0...

May be a symlink could be created just do fit the same pattern that other
versions do and keeps the actual links (for beta) working.

That we can do - in fact, done.

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

#10Dickson S. Guedes
listas@guedesoft.net
In reply to: Magnus Hagander (#9)
Re: Release versioning inconsistency

2012/6/21 Magnus Hagander <magnus@hagander.net>:

On Wed, Jun 20, 2012 at 1:35 PM, Dickson S. Guedes <listas@guedesoft.net> wrote:

2012/6/20 Magnus Hagander <magnus@hagander.net>:

On Wed, Jun 20, 2012 at 11:23 AM, Marti Raudsepp <marti@juffo.org> wrote:

On Wed, Jun 20, 2012 at 12:18 PM, Magnus Hagander <magnus@hagander.net> wrote:

(I do believe that using the v9.2.0beta marker is
*better*, because then it sorts properly. But likely not enough much
better to be inconsistent with previous versions)

Good point. Maybe that's a reason to change the versioning scheme and
stick with "9.2.0betaX" everywhere. Including calling the final
release "9.2.0" instead of simply "9.2"?

That might actually be a good idea. We can't really change the way we
named the betas, but it's not too late to consider naming the actual
release as 9.2.0...

May be a symlink could be created just do fit the same pattern that other
versions do and keeps the actual links (for beta) working.

That we can do - in fact, done.

It works fine here, thanks!

[]s
--
Dickson S. Guedes
mail/xmpp: guedes@guedesoft.net - skype: guediz
http://guedesoft.net - http://www.postgresql.org.br

#11Peter Eisentraut
peter_e@gmx.net
In reply to: Magnus Hagander (#9)
Re: Release versioning inconsistency

On tor, 2012-06-21 at 16:19 +0200, Magnus Hagander wrote:

On Wed, Jun 20, 2012 at 1:35 PM, Dickson S. Guedes <listas@guedesoft.net> wrote:

2012/6/20 Magnus Hagander <magnus@hagander.net>:

On Wed, Jun 20, 2012 at 11:23 AM, Marti Raudsepp <marti@juffo.org> wrote:

On Wed, Jun 20, 2012 at 12:18 PM, Magnus Hagander <magnus@hagander.net> wrote:

(I do believe that using the v9.2.0beta marker is
*better*, because then it sorts properly. But likely not enough much
better to be inconsistent with previous versions)

Good point. Maybe that's a reason to change the versioning scheme and
stick with "9.2.0betaX" everywhere. Including calling the final
release "9.2.0" instead of simply "9.2"?

That might actually be a good idea. We can't really change the way we
named the betas, but it's not too late to consider naming the actual
release as 9.2.0...

May be a symlink could be created just do fit the same pattern that other
versions do and keeps the actual links (for beta) working.

That we can do - in fact, done.

Why not just move the whole thing and not increase the confusion? The
press releases don't refer to the directory directly.

#12Peter Eisentraut
peter_e@gmx.net
In reply to: Magnus Hagander (#8)
Re: Release versioning inconsistency

On tor, 2012-06-21 at 16:17 +0200, Magnus Hagander wrote:

I don't recall the reason why using 9.2.0betax was actually wrong - i
realize that's not the name of the version, so thereby the directory
was wrong. But in what way would it be wrong to call the version that?

It's not the beta for 9.2.0, it's the beta for the 9.2 series. There is
not 9.2.1betaX, after all.

Given that it would help with sorting.

How does it help? 9.2.0, 9.2.0betaX, 9.2.1, ... ???