Talk:Structure theorem for finitely generated modules over a principal ideal domain

Primary ideals

Aren't primary ideals of a principal ideal domain simply ideals of the form $(p^{n})$ , $p$ prime? If so, I think invoking primary ideals seems unnecessarily general here, since IIRC primary ideals over non-PIDs can be much more complicated than prime powers. Functor salad 15:36, 10 October 2007 (UTC)[reply]

Good point. Primary ideal is the correct term, and it is called the primary decomposition, but I've added the clarification that here primary = prime power. (and also a note on the primary ideal page that in PIDs these coincide, but in general they don't, given a counter-example.) Thanks!

Nbarth (talk) 01:43, 23 November 2007 (UTC)[reply]

"Proofs" section

I commented out a part of the proofs section because I felt a) it contained misleading statements and errors b) it was unclear. I say unclear because I think it used terminology that is not readily found in Wikipedia. Misleading statements include:

"Hence, torsion free modules over PIDs are projective." (This is not true, you can find in Lam's Lectures on Modules and Rings that for any nonfield integral domain, the field of fractions is torsion free and nonprojective.)
"If M is projective, so is a direct summand of a free module F = M + N." (The author does not seem to use this, and it is nonsensical. Every summand of a free module is projective.)
"If M is torsion, it is the quotient of a free module." (Every module is a quotient of a free module, it has nothing to do with torsion.)

What is commented out needs heavy editing if it is to be useful, or else be replaced altogether. Rschwieb (talk) 19:21, 14 June 2011 (UTC)[reply]

"Uniqueness" section

The example module is not free, so it has no basis. I think that what you are triying to say is that the module is the direct sum of the cyclic submodules generated by these elements. Tama moana (talk) 15:09, 6 November 2015 (UTC)[reply]

Proofs missing uniqueness

I decided I'd try to finally learn a proof of this theorem, and was happy to find a proof section here. After following the link in the proof section to Smith normal form, I now understand why the decomposition exists. However, the proof section doesn't seem to concern itself with the uniqueness part, and the uniqueness section contains no proof. --Ørjan (talk) 18:09, 17 February 2019 (UTC)[reply]