Re: Treo 300: pilot-link works, gnome-pilot doesn't

[Adam C Powell IV <hazelsct mit edu>]

Hello again,

Adam C Powell IV wrote:

> David A. Desrosiers wrote:
>
> > > Funny thing is, if gpilotd is running and I push sync, the device hangs
> > > and never times out.  If I then run pilot-xfer, it sees /dev/ttyUSB1 
> > > and
> > > tries to start syncing, but *then* the device times out (and 
> > > disconnects).
> >
> >     What version of gnome-pilot are you using?
> >
> >     What version of pilot-link was it built against?
>
> 0.1.71-5 Debian woody backport, 0.11.7-2 backport.

So is only 2.0.1 supported now?  I'd be happy to do some maintenance on 
the old versions, since I'm not planning to upgrade to GNOME 2.* for 
some time.

> > > So I think something is wrong in gpilotd's assumptions about when 
> > > things
> > > will happen.  Perhaps it isn't monitoring /dev/ttyUSB1 like it says 
> > > it is,
> > > because there's no such device before sync?
> >
> >     gnome-pilot doesn't read the Treo device info from /proc, because as
> > you probably figured out, /proc/bus/usb/devices for these newer 
> > devices (OS5
> > devices, Treo, etc.) do not have the string "Palm Handheld" in them 
> > as they
> > did previously (gpilotd/gpilotd.c in struct product_names). This has 
> > to be
> > patched to handle the lookups for the actual product_id and the 
> > vendor_id
> > (in this case, 082d and 200 respectively). I think someone did this 
> > for the
> > Tungsten awhile back on this list. You might want to search the 
> > archives.
>
> Actually, the treo 300 is showing up as 082d and 0100, like a plain 
> Visor.  Here's what shows up in /proc/bus/usb/devices:
>
> T:  Bus=01 Lev=01 Prnt=01 Port=00 Cnt=01 Dev#= 12 Spd=12  MxCh= 0
> D:  Ver= 1.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS= 8 #Cfgs=  1
> P:  Vendor=082d ProdID=0100 Rev= 1.00
> C:* #Ifs= 1 Cfg#= 1 Atr=a0 MxPwr=  2mA
> I:  If#= 0 Alt= 0 #EPs= 4 Cls=ff(vend.) Sub=00 Prot=00 Driver=serial
> E:  Ad=01(O) Atr=02(Bulk) MxPS=   8 Ivl=0ms
> E:  Ad=82(I) Atr=03(Int.) MxPS=   8 Ivl=1ms
> E:  Ad=03(O) Atr=02(Bulk) MxPS=  64 Ivl=0ms
> E:  Ad=84(I) Atr=02(Bulk) MxPS=  64 Ivl=0ms
>
> There's no "Palm" or "Handspring" anywhere, so no match to anything in 
> product_names.  Perhaps a more involved patch is required?  I'd be 
> glad to help... 

So we have a device which doesn't give its name.  What to do?  It seems 
we can either:

   * Make an "exception" loop, after the main loop which checks the
     names for a match; this exception loop would trigger if the main
     loop didn't see anything and check for the one (or more?) Vendor
     and ProdIDs which correspond to devices without names.
   * Switch entirely to using Vendor and ProdID instead of name,
     copying in some of the values from kernel visor.h.

I'd be happy to code either of these, test, and send a patch to the 
list.  If nobody replies, I'll do the first of the two, probably early 
next week..

Cheers,
--

-Adam P.

GPG fingerprint: D54D 1AEE B11C CE9B A02B  C5DD 526F 01E8 564E E4B6

Welcome to the best software in the world today cafe! 
<http://lyre.mit.edu/%7Epowell/The_Best_Stuff_In_The_World_Today_Cafe.ogg>