In old Caché versions it was possible to create a new role based on predefined %Developer by copying it and adding some resources as needed. It was true at least from 2010.1 to 2015.1.

After upgrade from 2015.1.4 to 2017.2.1 it turned that it's only partially true now. User with a "New-Developer" role can enter Studio and open existing cls/mac/etc for editing and everything is OK unless he tries to create something new (Ctrl-N), than he gets a pop-up with %msg: <User xxx does not have enough privilege to execute stored procedure %CSP.StudioTemplateMgr_Templates>

In part 1 and part 2 of this series we set up two user types, simpleuser and poweruser. In this part of the tutorial we create one last user type having privileges typically needed by an administrator/developer in analytics.

In part 1, part 2, and part 3 parts of this series we set up three user types. In part 4 we saw how to secure model elements and DeepSee items. In this last part of the tutorial we conclude with some remarks on DeepSee security and troubleshooting tips. In particular, we see how pivot tables in User Portal can be "hidden".

In part 1, part 2, and part 3 of this series we set up three user types. In this part of the tutorial we see how to secure model elements (such as DeepSee cubes) and DeepSee items (such as a folder containing pivot tables and dashboards in the DeepSee User Portal).

The following code walks a DOM using %XML.Node. It also prevents %XML.Writer to change whitespace. Run the code using the class method "test":

Class objectscript.walkDOM Extends %Persistent
{
    ClassMethod dfs(node As %XML.Node)
    {
        s entrynode=node.NodeId
        do {
        //element nodes with one whitespacetyped child are the ones we want to change
        if (node.NodeType=$$$xmlELEMENTNODE){
            s snode=node.NodeId     
            if (node.MoveToFirstChild())            
                {
                    i ('node.MoveToNextSibling()){
                        i (node.NodeType=$$$xmlWHITESPACENODE){
                            s node.NodeType=$$$xmlTEXTNODE
                            s node.NodeId=snode
                        }
                    }
            }
            s node.NodeId=snode     
        }   
        if (node.HasChildNodes()){
            d node.MoveToFirstChild()
            d ..dfs(node)
        }
        } while (node.NodeType'="" && node.MoveToNextSibling())
        s node.NodeId=entrynode
         
    }
     
    ClassMethod test()
    {
      set xml = "abcdefg<![CDATA[   ]]>"
     
      s reader=##class(%XML.Reader).%New()
      do reader.OpenString(xml)  
      set writer = ##class(%XML.Writer).%New()
      //do some magic
      d ..dfs(reader.Document)
       
      w !,"with indent=1:",!
      set writer.Indent = 1
      do writer.OutputToString()
      do writer.Document(reader.Document)
      w writer.GetXMLString()
      set writer.Indent = 0
      w !,"with indent=0:",!
      do writer.OutputToString()
      do writer.Document(reader.Document)
      w writer.GetXMLString()
    }
}

Here's a link to the code on GitHub: https://github.com/intersystems-community/code-snippets/blob/master/src/...

Here is a snippet that I learned yesterday

You can define an index on a collection property but when I tried to use it, I failed. I was using

     Select ….. where …. :xx %INLIST collproperty

But this will not use an index, but the equivalent syntax

     SELECT .. WHERE ... FOR SOME %ELEMENT(collproperty) (%VALUE=:xx)

will use the index

Check out

http://docs.intersystems.com/latest/csp/docbook/DocBook.UI.Page.cls?KEY=...

This data set demonstrates a basic M:N relationship between 2 tables
The dataset is targeted to show a slim implementation of M:N
It's no question that other implementations exist. But at significantly more storage consumption.