OLE error code:80040E14

[neutralm]

Hi,

I am getting the following error:

OLE error code:80040E14 in Microsoft OLE DB Provider for SQL Server
Column 'tags.id' is invalid in the select list because it is not
contained in either an aggregate function or the GROUP BY clause.
when trying to execute the following query:

select tags.id, name, count(*) as count from taggings, tags where
tags.id = tag_id group by tag_id
The above query works fine on MySQL, but chokes on SQL Server.

Could anyone please help?

Thanks!

NM

[Erland Sommarskog]

(ne******@gmail.com) writes:

I am getting the following error:

OLE error code:80040E14 in Microsoft OLE DB Provider for SQL Server
Column 'tags.id' is invalid in the select list because it is not
contained in either an aggregate function or the GROUP BY clause.
when trying to execute the following query:

select tags.id, name, count(*) as count from taggings, tags where
tags.id = tag_id group by tag_id
The above query works fine on MySQL, but chokes on SQL Server.

SQL Server, like most DB engines, as well as ANSI SQL, that if your
SELECT list includes an aggregate such as COUNT(*), and there is no
OVER clause for the aggregate, then all unaggregated columns in the
SELECT list must appear in the GROUP BY list.

Change tag_id in the GROUP BY clause to tags.id or vice versa.

Apparently MySQL is lax on this point. As a matter of fact SQL Server
4.x also permitted columns to appear in the SELECT list, if they did
not appear in GROUP BY. Sometimes the result made sense, as here
where tags.id is one-to-one with tags_id. Sometimes you got screenfulls
of garbage when you expected two lines, because you had left out a
column in the GROUP BY clause. The feature was removed in SQL Server
6.0 (and Sybase System 10), missed by few.

--
Erland Sommarskog, SQL Server MVP, es****@sommarskog.se

Books Online for SQL Server 2005 at
http://www.microsoft.com/technet/pro...ads/books.mspx
Books Online for SQL Server 2000 at
http://www.microsoft.com/sql/prodinf...ons/books.mspx

[neutralm]

Thanks for your prompt reply, Erland. Pardon my ignorance, but I'm
still not sure if I understood how to solve the problem (although I
think I understand what the problem is from your explanation).

I have two tables:

1. tags (with the primary key 'id' and an attribute 'name')
2. taggings (the primary key is 'id', the foreign key is 'tag_id')

The query string I'm using is:

select tags.id, taggings.tag_id, name, count(*) as count from taggings,
tags where tags.id = taggings.tag_id group by taggings.tag_id

How should the correct query look like?

Thanks so much in advance!
Erland Sommarskog wrote:

(ne******@gmail.com) writes:

I am getting the following error:

OLE error code:80040E14 in Microsoft OLE DB Provider for SQL Server
Column 'tags.id' is invalid in the select list because it is not
contained in either an aggregate function or the GROUP BY clause.
when trying to execute the following query:

select tags.id, name, count(*) as count from taggings, tags where
tags.id = tag_id group by tag_id
The above query works fine on MySQL, but chokes on SQL Server.

SQL Server, like most DB engines, as well as ANSI SQL, that if your
SELECT list includes an aggregate such as COUNT(*), and there is no
OVER clause for the aggregate, then all unaggregated columns in the
SELECT list must appear in the GROUP BY list.

Change tag_id in the GROUP BY clause to tags.id or vice versa.

Apparently MySQL is lax on this point. As a matter of fact SQL Server
4.x also permitted columns to appear in the SELECT list, if they did
not appear in GROUP BY. Sometimes the result made sense, as here
where tags.id is one-to-one with tags_id. Sometimes you got screenfulls
of garbage when you expected two lines, because you had left out a
column in the GROUP BY clause. The feature was removed in SQL Server
6.0 (and Sybase System 10), missed by few.

--
Erland Sommarskog, SQL Server MVP, es****@sommarskog.se

Books Online for SQL Server 2005 at
http://www.microsoft.com/technet/pro...ads/books.mspx
Books Online for SQL Server 2000 at
http://www.microsoft.com/sql/prodinf...ons/books.mspx

[Ed Murphy]

ne******@gmail.com wrote:

The query string I'm using is:

select tags.id, taggings.tag_id, name, count(*) as count from taggings,
tags where tags.id = taggings.tag_id group by taggings.tag_id

How should the correct query look like?

select taggings.tag_id, name, count(*) as tag_id_count
from taggins join tags on taggings.tag_id = tags.id
group by taggings.tag_id, name

Explanations:

1) GROUP BY must include all unaggregated columns from the SELECT,
i.e. everything that is not a COUNT(), SUM(), etc. (Why doesn't
it implicitly assume this? Apparently, it used to let you leave
things out, but that caused more trouble than it was worth. The
short answer is "just give it what it wants".)

2) tags.id and taggings.tag_id are forced to be equal, so you only need
to include one of them. Optional but recommended, as it's simpler
and conserves bandwidth.

3) The join is changed from SELECT ... FROM A, B WHERE A.X = B.Y
to SELECT ... FROM A JOIN B ON A.X = B.Y
Optional but recommended, as it keeps join conditions separate from
each other, and from other restrictions (e.g. NAME LIKE '%ABC%'),
all of which makes the query easier to understand.

[neutralm]

Thank you very much, Ed. I really appreciate how quickly you've help me
fix this problem!

Ed Murphy wrote:

ne******@gmail.com wrote:

The query string I'm using is:

select tags.id, taggings.tag_id, name, count(*) as count from taggings,
tags where tags.id = taggings.tag_id group by taggings.tag_id

How should the correct query look like?

select taggings.tag_id, name, count(*) as tag_id_count
from taggins join tags on taggings.tag_id = tags.id
group by taggings.tag_id, name

Explanations:

1) GROUP BY must include all unaggregated columns from the SELECT,
i.e. everything that is not a COUNT(), SUM(), etc. (Why doesn't
it implicitly assume this? Apparently, it used to let you leave
things out, but that caused more trouble than it was worth. The
short answer is "just give it what it wants".)

2) tags.id and taggings.tag_id are forced to be equal, so you only need
to include one of them. Optional but recommended, as it's simpler
and conserves bandwidth.

3) The join is changed from SELECT ... FROM A, B WHERE A.X = B.Y
to SELECT ... FROM A JOIN B ON A.X = B.Y
Optional but recommended, as it keeps join conditions separate from
each other, and from other restrictions (e.g. NAME LIKE '%ABC%'),
all of which makes the query easier to understand.