Thread: Do Not Approve Release Announcement Yet

WWW Team,

I'll be posting the patch release announcement to postgresql.org later today 
or tommorrow AM.  Please do *not* approve it until you see Marc send out the 
e-mail version.

-- 
Josh Berkus
PostgreSQL @ Sun
San Francisco

On Sun, Sep 16, 2007 at 03:13:48PM -0700, Josh Berkus wrote:
> WWW Team,
> 
> I'll be posting the patch release announcement to postgresql.org later today 
> or tommorrow AM.  Please do *not* approve it until you see Marc send out the 
> e-mail version.

I'd like to make a suggestion around this:

Next time, how about we post it to the website *first*, and to -announce
later? Reasons:

1) Gives the web changes a chance to propagate out to all the mirrors
before people go looking for information.

2) The post on the web can be *changed and fixed* if typos are found when
more ppl look them over (like the download link this time)


We could then argue if the release announcement should be a duplicate of
the content, or just a link to it. But even if it's a duplicate of the
content, I think it makes a lot of sense to do it in that order.

It also needs to be better coordinated with loading of the new versions of
the documentation, something we completely forgot this time, but have
remembered for some previous releases. When the -announce mail went up,
the release notes weren't present on the website yet, but were referred to
in the announcement...


Thoughts & comments?

//Magnus

On Wed, Sep 19, 2007 at 04:38:49PM +0200, Magnus Hagander wrote:
> On Sun, Sep 16, 2007 at 03:13:48PM -0700, Josh Berkus wrote:
> > WWW Team,
> > 
> > I'll be posting the patch release announcement to postgresql.org later today 
> > or tommorrow AM.  Please do *not* approve it until you see Marc send out the 
> > e-mail version.
> 
> I'd like to make a suggestion around this:
> 
> Next time, how about we post it to the website *first*, and to -announce
> later? Reasons:
> 
> 1) Gives the web changes a chance to propagate out to all the mirrors
> before people go looking for information.
> 
> 2) The post on the web can be *changed and fixed* if typos are found when
> more ppl look them over (like the download link this time)
> 
> 
> We could then argue if the release announcement should be a duplicate of
> the content, or just a link to it. But even if it's a duplicate of the
> content, I think it makes a lot of sense to do it in that order.
> 
> It also needs to be better coordinated with loading of the new versions of
> the documentation, something we completely forgot this time, but have
> remembered for some previous releases. When the -announce mail went up,
> the release notes weren't present on the website yet, but were referred to
> in the announcement...
> 
> 
> Thoughts & comments?

Does the complete absense of comments mean that we'll do this next time?

//Magnus