<br><div><span class="gmail_quote">On 6/13/07, <b class="gmail_sendername">Robby Findler</b> <<a href="mailto:robby@cs.uchicago.edu">robby@cs.uchicago.edu</a>> wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"> <br>> It is kind of a combination of two intended aspects of the design of<br>> the PL. One: structs should be generative, so that no one can fake<br>> anyone else's structs. Two: modules should be separately compilable, <br>> which means that you have to have separate instantiations of modules<br>> used at compile time (which are then discarded). The second one is<br>> what's covered in the paper mentioned above.</blockquote> <div><br>Right - my point is that the combo is probably unexpected (even both work as intended), although the thread Ryan alluded to shows this is a known issue for a while.<br></div><br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"> > I'm not sure that there isn't a better pair of designs out there that<br>> interact better, tho.</blockquote><div><br>Understood - just want to raise this as something that other schemers might find useful in the future (and hopefully someone finds it an interesting puzzle to solve)... for now I will just go back to the old ways. <br><br>Thanks,<br>yc<br><br></div></div>