Greetings,
I've written a nice recursive walk through my asset-database to generate a document which either shows which assets are impacted by a certain asset or (other direction, same thing) which assets are dependent on a certain asset.
So far all cool. I've managed to do everything in DPML.

Now I want to improve performance by launching my subrequests (due to SQL requirements I can only do one level at a time, hence the need for recursiveness) async. By serializing the output-writing (golden thread) I should be able to pull that off.
So far still good.

The change to async of the DPML request is easy. However, this request is inside a while loop. So I have to keep track of x-number requests in order to join all of them after the loop (otherwise there's no benefit).
In a script language that would be easy. Enter a table. Build it as you run through the while, process it after.
In DPML however ... how do I keep track of the requests ?

Has anyone done this (and succeeded) before ?
Regards,
Tom