Re: missing docs for new minor releases

On Thu, Feb 9, 2023 at 6:53 PM Jonathan S. Katz <jkatz(at)postgresql(dot)org>
wrote:

> On 2/9/23 12:45 PM, Magnus Hagander wrote:
> > On Thu, Feb 9, 2023 at 6:04 PM Jonathan S. Katz <jkatz(at)postgresql(dot)org
> > <mailto:jkatz(at)postgresql(dot)org>> wrote:
> >
> > On 2/9/23 11:37 AM, Jonathan S. Katz wrote:
> > > On 2/9/23 10:34 AM, Robert Treat wrote:
> > >> I reckon someone may already be working on this, but in case
> > not, FYI
> > >> that the urls linked from the docs section of the website are
> > giving
> > >> 404's...
> > >> ie.
> > >> from-> https://www.postgresql.org/docs/release/
> > <https://www.postgresql.org/docs/release/>
> > >> <https://www.postgresql.org/docs/release/
> > <https://www.postgresql.org/docs/release/>>
> > >> to-> https://www.postgresql.org/docs/release/14.7/
> > <https://www.postgresql.org/docs/release/14.7/>
> > >> <https://www.postgresql.org/docs/release/14.7/
> > <https://www.postgresql.org/docs/release/14.7/>>
> > >> as well as the other new minor releases.
> > >
> > > This was a caching issue and should now be resolved.
> >
> > Barring exceptional circumstances, this two-liner should be the fix
> for
> > this issue, i.e. operator error.
> >
> > When a version record is updated, this will purge from the cache all
> > the
> > release notes associated with that version. We could further limit
> > it to
> > the latest major/minor, but I do want to anticipate any errors that
> > could occur with setting the version record, e.g. an overincrement.
> >
> > This also purges the "/docs/release/" page itself, as that needs to
> > contain the updated list of release notes.
> >
> >
> > It should probably have a trailing slash at the version specific one,
> > and probably a terminating dollar?
>
> Well, that was my point. We could make it major/minor version specific,
> but actually we do link to all of the minor versions within a tree, e.g.
> the right side bar:
>
> https://www.postgresql.org/docs/release/12.13/
>
> So I think it's correct as is.
>

Huh, dangit. I thought it meant something else, I should've better reviewed
what our actual URL scheme is :)

To fight off similar comments in the future, maybe add a code comment about
why we do that. But yes, thinking it over again, I agree the code is
correct.

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