Re: Broken Urls present on the official PostgreSQL website (as reported by the Testing Harness Suite) - Mailing list pgsql-www

> On 25 Jul 2023, at 04:08, Jonathan S. Katz <jkatz@postgresql.org> wrote:
>
> On 7/24/23 2:22 PM, Akshat Jaimini wrote:
>
>>     > The broken links present in the log file attached below are:
>>     >
>>     > --- Links to external sites -----
>>     > 1. https://www.postgresql.org/docs/current/static/pgstandby.html
>>    <https://www.postgresql.org/docs/current/static/pgstandby.html>
>>     > <https://www.postgresql.org/docs/current/static/pgstandby.html
>>    <https://www.postgresql.org/docs/current/static/pgstandby.html>> :
>>    404 :
>>     > On Page https://www.postgresql.org/about/featurematrix/
>>    <https://www.postgresql.org/about/featurematrix/>
>>     > <https://www.postgresql.org/about/featurematrix/
>>    <https://www.postgresql.org/about/featurematrix/>>
>>     > 2. https://www.postgresql.org/docs/current/static/tsearch2.html
>>    <https://www.postgresql.org/docs/current/static/tsearch2.html>
>>     > <https://www.postgresql.org/docs/current/static/tsearch2.html
>>    <https://www.postgresql.org/docs/current/static/tsearch2.html>> : 404 :
>>     > On Page https://www.postgresql.org/about/featurematrix/
>>    <https://www.postgresql.org/about/featurematrix/>
>>     > <https://www.postgresql.org/about/featurematrix/
>>    <https://www.postgresql.org/about/featurematrix/>>
>
> I was unable to find either of these links when searching the feature matrix database. Is there any other info that
cannarrow down which entry these are in? Otherwise, these may have already been removed. 

I'm guessing this ran against the featurematrix/fixtures/data.json in the pgweb
repo which does contain these links, but which doesn't reflect the actual
website content due to being an initial dataset (in turn generated by the
script pgweb/dump_initial_data.sh).

The data.json being served on the production site is based off the database and
isn't committed to the repository.  A testing harness for the website code
should probably skip checking the validity of dynamic data unless it's running
against the live site for those checks.

--
Daniel Gustafsson