Marc Weber wrote:
Thanks for the patches! Neither one seems consistent to me with my
vision for how Ur/Web should be set up, so I'll leave them as something
folks can apply optionally.
It doesn't make sense to me that you call Haskell's "implicit mutual
recursion spread across files" an entangled mess because you want to
read code top-down to understand what's happening but don't apply this
pattern to urweb .. because the http, fastcgi, cgi targets register
themselves?
The case of plugin systems is different than usual. I'm not fully
satisfied with the imperative solution that the Ur/Web compiler uses
now, but I'm also not so bothered by it (and don't know a better
approach that works in ML). The way the code is set up now, it's easy
to define your own variant of the Ur/Web compiler that adds a new
protocol, with no need to touch the main urweb repository; I've done
this to good effect in the past (no code of this kind is publicly
available, though).
I used this pattern for single line comments forcing a space because #
is used by labels and chars:
line_comment = [#][ ][^\n]*\n;
However this only works using sml/nj but not mlton?
Any idea what could be the cause before I start debugging this issue?
I don't know, though you might try escaping more characters with
backslashes.
_______________________________________________
Ur mailing list
[email protected]
http://www.impredicative.com/cgi-bin/mailman/listinfo/ur
