dbus on win32

[Tor Lillqvist]

Markus Fritsche writes:
 > I found dbus today (a link in squeak-dev proposing dbus for callbacks 
 > and as interlanguage-ipc). How is the win32 port going?

I worked on it in January, but not since. Nothing of my work got
committed, and the original reason for me working on the port to
Windows kinda fizzled out. The changes are now just sitting on my
disk. (I do have an offline copy of the diffs (I think there is a
pointer in the mailing list archives), so no risk of it getting lost
in case my machine's disk crashes...) The CVS codebase has presumably
changed a lot since, so the Win32 work needs at least a thorough
review, and testing, possibly even needs to be partially redone.

 > And, if I understood correctly, changing from local sockets to
 > network ones shouldn't be to hard?

Hmm, you mean changing from using Unix domain sockets to TCP sockets?
As far as I can say, that's true. (There are no Unix domain sockets on
Windows (no surprise), so my Win32 port used only TCP sockets instead,
although bound to the loopback interface.)

 > If so, could dbus be an alternative to xmlrpc, or am I completely
 > wrong about the potential of dbus?

Somebody else, who understands the architectural issues, will have to
answer that.

--tml