Daniel Pritchard

To continue the discussion, an icon for Google Scholar and Mendeley would be great. Carl's suggestion for Mendeley is a good one. [GS's favicon](http://scholar.google.com/favicon.ico) would be a good way to...

FYI, I have implemented some basic handling for lists and "everything else" `signature(ANY, ...)` over here: https://github.com/dpritchard/validate/commit/912b87afff8f61131f58a53e0b5a6bf3734e4875 But I don't think this is going to be so simple if reference...

OK, so S4 dispatch is not too hard! This works as (I) expect: https://github.com/dpritchard/validate/commit/4bc15df58ba749faac80ac64bdfca81d4124411a I think that existing (`data.frame`, based) method signatures should also be updated in a similar way......

> OK, so S4 dispatch is not too hard! Well, maybe it's harder than I thought! Introducing a method for a `list` creates a method dispatch ambiguity (for `data.frames`), obviously....

Thanks for the reply. I think it best to leave those design choices to you and the other package authors. For now the changes I have made to my local...

Would a pull request that makes the following changes be useful? 1. Prevent dispatch on unsupported reference types and/or 2. Allows lists and other types for the simple (non-reference) case...