Yeah agree with Ash. I think I got bitten by it a couple of times at the very least where the dag didn't show up as I just had "with DAG(...): "
On Wed, 27 Apr 2022 at 21:28, Jarek Potiuk <[email protected]> wrote: > That looks like a nice improvement. > > J. > > On Wed, Apr 27, 2022 at 10:04 PM Malthe <[email protected]> wrote: > > > > On Wed, 27 Apr 2022 at 15:27, Constance Martineau > > <[email protected]> wrote: > > > Am intrigued. Curious about dynamic dag pattern, where you create the > DAG object in a create_dag function and adding the DAG to globals. Would > this new way prevent someone from modifying the dag object within the > function, or returning it? > > > > It would not change the current way of programming DAGs. The only > > change is that if you call the DAG constructor (or a function that > > does so including one using the @dag decorator), then that DAG will be > > made available regardless of whether you assign it a top-level > > variable or not. > > > > Cheers >
