Legend: Bold face encodes 'need more information'; italics encodes issues that are (semi)settled in the sense that they are the basis for future discussions; Underlined Text are action items.

• Requirement on r:
  • Need to motivate the target. We can't have a fishing expedition. 'Let's get as low as we can' won't work as a justification. There needs to be a well motivated requirement. Raphael says that it would be hard to motivate the level of r based on specific models, but one should be able to generate motivations based on fundamental physics and classes of theories. In the proposal we argued for particular goals (and were limited by foregrounds). Raphael (or Lloyd?) will organize a group of people who will work on a particular target. The S4 science book targets r<0.002 at 95%.
  • The connection between limits on r and resolution: does the limit on r depends on delensing, and if so, how much delensing, and how does that relate to the resolution? The probe is unlikely to have an aperture larger than 2 m. Amy says it is likely to be less than 1.5 m. Charles argues that delensing will be complicated by foregrounds. Shaul suggests first understanding how much delensing is necessary, and its connection to resolution without the complication of foregrounds. All agree that foreground complications will have to be addressed as well.
  • Alex Van Engelen and Blake Sherwin are mentioned as some who can contribute analysis.
  • Lloyd (?) will organize study of delensing. Question: “what is the level of r that can be achieved as a function of resolution and depth?”