First of all, the topic should be something like "sharing a sequential dataset by many jobs," not "job card."
Second, when I read the first post I got the impression syamcs wanted multiple jobs to write to the dataset. This is ill advised. The usual idea of writing a sequential dataset is to entirely replace any previous content of the dataset. While it is certainly possible to write programs that update a sequential dataset, this requires programming skills and methods not readily available to most Cobol programmers, and I think we would require much more information than syamcs has disclosed before we could discuss solutions.