Today I needed one for the first time, and why? I did my query directly into the mysql terminal (I rarely do this).

I had to use:
SELECT
count(t1.id_address), min(t1.id_address)
FROM
address t1
WHERE
id_address >= (SELECT
min(id_address)
FROM
applicant t2
WHERE
dt_complete > CURDATE() – INTERVAL 1 YEAR)

It’s so messy. But it was a quick hack. I still see no alternative, since address is used by like 4 other tables.

SELECT DISTINCT children.id_section, children.id_parent, children.barnav, children.lastmod
FROM sections children LEFT JOIN sections parents ON children.id_parent=parent.id_section
WHERE children.active=1
AND parents.active=1
ORDER BY children.id_parent;

Worked seemlessly.
Thanks -specially- for the wisdom within, man.

That query can be easily rewritten using an inner join. There’s no rule saying that you cannot join a table on itself. It would be a worthwhile exercise for you to figure out how to rewrite the query without using a nested join.

SELECT id_section, id_parent AS aka_parent, barnav, lastmod
FROM sections
WHERE active=1
AND (SELECT active FROM sections WHERE id_section=aka_parent)=1
ORDER BY aka_parent;

There you need to use INSIDE the nested query a value retrieved as an alias in the main query.
So, if you are looping this, and you go the distinct query way you may not get the same results. Isn’t it?

Thanks!
-Scott