Reread the thread, you're wrong and that's what you have to learn with this question.

One of us missed the requirements of the whole project.

Can you repost the question with sample data and the required output please?

Is the clustered index on the dt column?

replace(replace (column, char(10), ''), char(13), '')

why don't you want to keep that information in the db??

I feel your pain.

TSQL is great, the design is awfull, that's where you're having problems.

oh boy thats sad ((

Yup

If I can make a proc work like a function but can't use it results anywhere it what's the point

The point of the function...

Would've loved to see their face after you gave them that argument.

Someone smart here said this once : "When you see yourself in the face of an impossibly complicated task, look at a way to simplify the task or avoid it...

We should ask Bill Gates to make a public announcement on this one... so we can move on to other things.

So what is it you do not understand?

Search this site for Change Management. There are quite a few threads/articles on that topic.

The relationships you make in the views do not propagate to the schema.

... but the ones in the diagrams do.

I think I would make a new job. This job first step would to check for the file. If it's there, then start the current job (sp_startjob).

How about check for the file first, then start the job?