standardizing factoring loading with different options in SEM

[Yu-Ting Chin]

Hi Statalists,
I am working on a CFA model in the structural equation model using Stata 17.0 for Windows.
The following codes confused me.

1. sem(M-> x1 x2 x3 x4 x5),

2. sem(M-> x1 x2 x3 x4 x5), standarized

3. sem(M-> x1 x2 x3 x4 x5), var(M@1)

4. sem(M-> x1 x2 x3 x4 x5), standarized var(M@1)

To my understanding,
code 1. uses x1 as reference; thus the factor loading of x1 will be 1.
code 2. also uses x1 as reference; however, all the indicators will be scaleless. So the factor loading of x1 can be identified.
code 3. constrains the variance of M to be 1 in order to identify the parameters of x1-x5.

My questions are

a. Is code 4. reasonable?
b. If not, what should be done if I'd like to constrain the variance of M while having all the factor loadings scaleless.

Thanks.