gnome-pilot 0.1.65 and pilot-link 0.11.1

["Brett A. Thomas" <quark baz com>]

All,

OK, I've been through the archives.  I've read Peter Robinson's post in
May about getting all this to work, and it still *just doesn't*, and I'm
pretty stuck and frustrated.  So, I apologize for wasting all your time
on what is probably a misconfiguration on my part, but any help would be
appreciated.  I also hope that, if we can document the fix, anyone else
in this boat will be better off.

So, my situation is, I have a Handspring Treo, which won't work with
older versions of pilot-link, so just grabbing a binary compile of
everything isn't an option.  Theoretically I know what I'm doing, but
it's been a long time since I've messed around with autoconf and the
like.

First of all, I'm running RedHat Linux 7.2.  I carefully removed all
traces of the stock gnome-pilot and pilot-link packages (with rpm -ev). 
Next, I grabbed pilot-link 0.11.1.  I executed configure with a --prefix
of /usr, because I'd discovered that gnome-pilot's .configure seemed
very unhappy if it was in a different place than the gnome stuff - at
least with JPR's modern pilot-link patch applied.  I compiled and
installed pilot-link without incident, ran ldconfig, and was able to use
pilot-xfer to list the applications on my Treo and do a complete backup.

After that, I got gnome-pilot 0.1.65, and applied JPR's patches with
'patch -p0'.  Here's where it gets a little ugly.  I recognized from my
many-years ago time as a Unix C programmer (now I'm a PHB) that autoconf
or somesuch needed to be run, because the changes to add libpisync were
to a .m4 file, and I knew I needed to regenerate configure.  The README
suggested running "./autogen", which is in macros/, but it's not
executable in my distribution.  I used 'sh' to execute it, but it refers
to "$srcdir" all over the place, which wasn't set.  I set it manually,
and m4 complained about an error in the GOB detection stuff.

I backed off at this point, thought about it, decided I wasn't trying to
submit a patch on this (yet), and just modified configure by hand so
that LIB_PISOCK referred to lpisync as well every place it referred to
libpisock. This time it compiled clean, and I installed it in /usr, as
well.

Next, I grabbed gnome-pilot-conduits-0.8, and applied JPR's patch to
it.  When I tried to compile it, I got an error in
email/email_conduit.c, where O_RDONLY was an undefined symbol.  I threw
a "#include <fcntl.h>" at the top, and got a clean compile, there, too. 
I also installed it in /usr.

So, after all that:

IT ACTS EXACTLY THE WAY IT DID UNDER pilot-link 0.95 and gnome-pilot
0.1.63 I HAD BEFORE!

Which is to say, when I run gpilotd with --activate-goad-server (or, to
be more precise, when gpilot-applet runs it) it says:

gpilotd-Message: gnome-pilot 0.1.65 starting...
gpilotd-Message: compiled for pilot-link version 0.11.1
gpilotd-Message: compiled with [VFS] [GOAD] [USB] [IrDA] [Network] 
gpilotd-Message: Activating CORBA server
Message: state = CONNECTING_TO_DAEMON
gpilotd-Message: Watching Cradle (/dev/pilot)

And it then says nothing else.  Ever.  If I try to hotsync, the syslog
says:

Jul 22 10:34:39 localhost kernel: hub.c: USB new device connect on
bus1/1, assigned device number 29
Jul 22 10:34:39 localhost kernel: usbserial.c: Handspring Visor
converter detected
Jul 22 10:34:39 localhost kernel: visor.c: Handspring Visor: Number of
ports: 2
Jul 22 10:34:39 localhost kernel: visor.c: Handspring Visor: port 1, is
for Generic use and is bound to ttyUSB0
Jul 22 10:34:39 localhost kernel: visor.c: Handspring Visor: port 2, is
for HotSync use and is bound to ttyUSB1
Jul 22 10:34:39 localhost kernel: usbserial.c: Handspring Visor
converter now attached to ttyUSB0 (or usb/tts/0 for devfs)
Jul 22 10:34:39 localhost kernel: usbserial.c: Handspring Visor
converter now attached to ttyUSB1 (or usb/tts/1 for devfs)
Jul 22 10:34:42 localhost /etc/hotplug/usb.agent: Setup visor for USB
product 82d/100/100
Jul 22 10:34:45 localhost kernel: usb.c: USB disconnect on device 29
Jul 22 10:34:45 localhost kernel: usbserial.c: Handspring Visor
converter now disconnected from ttyUSB0
Jul 22 10:34:45 localhost kernel: usbserial.c: Handspring Visor
converter now disconnected from ttyUSB1

So, I can't sync my Treo, after all this.

Any brilliant ideas?  Is it my autoconf hacking?  If so, any advice on
how to successfully run autogen on gnome-pilot would be appreciated...

Also, has anyone else gotten pilot-link 0.11.1 to to work with
gnome-pilot?  Am I just too bleeding edge?  I'm willing to try to put
together a patch if it is in fact really broken and not just my
configuration...

Thanks...

Brett