On Mon, May 4, 2009 at 11:13 AM, Abdulaziz Ghuloum <[email protected]> wrote: > > How did you come to expect that Ikarus should behave exactly > like PLT or Larceny?
When I started this experiment I did not know what to expect. I had an idea of the purpose of the dummy macro trick, but it did not work on PLT, so I though I misunderstood. When Matthew Flatt said that it was a bug, and Andre' explained the visiting mechanism, I decided that I was right the first time, after all, and I expected Ikarus to be consistent too. But remember that the whole point of the exercise was to discover what is portable and what is not. It seems that Matthew Flatt's trick to bring static information at run-time is not portable. Fine, this is an important discovery per se, something I did not know in advance. >> If Ikarus does not work this way it seems that it is >> impossible to rely on side effects to survive separate >> compilation. > > It's not impossible, just hard. Now I can probably improve the > situation a bit, but I don't really see much point in doing so. If it cannot be done portably, I see little point in doing so too.
