 So we were discussing the content of the project and we focused our discussion around gaps in terms of the content of the project and also some lessons that we've already learned which could be fed into this project or into new projects. So some of the gaps were things like partly related to the survey tools that we're using, things like policy influences on crop residue management, the whole institutional setup, the analysis of those is somehow not very well captured in this project. Because we have tended to focus all our efforts on the farm level, the household level and the village level, then some of those higher level drivers are not really very well captured in this project and maybe that's something we need to think about. Another thing related to that is this whole area of farmer decision making, farmer preferences, allocation decisions which are really at the heart of this project and perhaps the way in which we've structured some of our survey tools to be very specific and quantitative have maybe not left enough room for capturing open questions about how farmers make decisions. Another issue we discussed was the whole novelty of this project because there have been many studies in mixed crop livestock systems in terms of how crop residues are allocated over the last two or three decades. So what's really new about what we're doing? Something that we felt was new was that previous work has tended to be focused on technical aspects of crop residue management whereas we are focusing a lot more on this project on some economic assessment of farmer decision making and also the use of multiple sites around the world which are at different stages of intensification allows us to say something about trajectories of intensification and how decision making changes as systems intensify. So maybe that's something new about this project too. Some lessons, one lesson I think that we've learned or perhaps we need to learn is that we need to integrate the social with the technical especially in the modelling component where we could get very we could turn out some very mechanistic biophysical models which don't really capture some of the social issues so that's something we need to think about. Other lessons, we need to keep our eye on the big drivers, the global drivers. There's already been this SLP driver study but we need to feed some of that thinking into this project and maybe that's something that could be new about this project and I think that's probably what captures most of what we've discussed.