Years ago I wouldn't dream of writing a tip about copying datasets using ISPF. Since the year 2000 it has become a lot easier than in the previous century. The new ISPF releases for z/OS has made it quite easy to copy datasets. Maybe you are still doing like this: first you allocate the dataset and then you copy data to it. You don't have to do this any more.

Try using ISPF MOVE/COPY (also known as 3.3 just like DSLIST is known as 3.4) Type in the name of the dataset, you want to copy from and press Enter. On the next panel you type in the name of the dataset you want to copy to. Let this name be the name of a non-existing dataset. Press Enter and now ISPF will kindly ask you if you want to allocate a new dataset using the same dataset attributes as the original one. Of course you want that. Type in the correct answer and press Enter again. Then ISPF does it all, allocating and copying.

When you are using ISPF EDIT/VIEW and uses the CREATE/REPLACE command for a non-existing dataset, the same thing happens. ISPF kindly ask for permission to create a new dataset. I simply cannot use this new feature enough. I have stopped using the ISPF DATASET ALLOCATION feature (3.2). All my new datasets are created using either ISPF EDIT/VIEW or ISPF COPY/MOVE (mainly for load libraries).

But do not feel too happy about this sudden user friendliness from IBM. I noticed that in older z/OS ISPF releases the CO line command in DSLIST (3.4) didn't ask polite questions but refused to copy to non-existing datasets. I haven't tried in the newest ISPF release, but I don't think they have made it work like the other COPY commands, but try it. You might be surprised.