In 2020, the z/TPF parser was updated to use 64-bit ECB heap for parsing large XML and JSON documents (PJ46135).  The z/TPF parser is used not only by calling tpf_doc_parseDocument but by REST provider and REST consumer support.  As ever larger XML and JSON documents were being processed, the next limitation was being quickly approached in the size of the parse tree being created.  The tree structure consists of different types of nodes such as element, attribute, and content nodes.  The number of nodes for each node type is limited to 65,535.  A JSON property is represented by an element node while the JSON property value is represented by a content node.  XML or JSON documents that are several megabytes in size can exceed this limit, resulting in errors.  To allow for very large documents to be processed, the limit was changed from 65,535 to 4,294,967,295.  For more information see the APEDIT.