Re: tramp 2.0.14: Why does de/encoding fail?

[Kai Großjohann]

"Tim Hermans" <address@hidden> writes:

> Why does tramp try every possible de/encoding method even after it has found
> a valid one?  I even specified a method with 'tramp-encoding-command' and
> 'tramp-decoding-command'.

Did you put a tramp-encoding-command and a tramp-decoding-command
parameter into the tramp-methods entry?

Tramp will ignore this.

In principle, it would be nice if Tramp could make a faster startup
by not testing so many things.  But it is more important to get Tramp
working at all in the first place, and it does not work for
everybody, yet.  Therefore, I haven't got around to writing the
necessary code to shortcut all the auto-detection.

(And then, I'm dreaming of redoing the auto-detection by sending a
shell script to the remote host once at connection setup time which
does all the work, and Tramp would just read the output of that
script.  I'm guessing that this might be so fast that it won't be
necessary to shortcut the auto-detection.)

> (By the way it's a standard Solaris install, so these are in their
> usual places).  Connecting from a Win2K box to Solaris 8.0 box with
> the telnet method.  The tramp session ends with:
>
>       tramp: Checking remote decoding command `tramp_uudecode' for sanity
>       tramp: Checking local encoding command `uuencode xxx' for sanity
>       tramp: Checking remote encoding command `tramp_encode_with_module' for
> sanity
>       tramp: Checking remote encoding command `tramp_encode' for sanity
>       tramp: Checking remote decoding command `tramp_decode' for sanity
>       tramp: Checking to see if encoding/decoding commands work on remote 
> host...
>       if: Remote host cannot execute de/encoding commands.  See buffer
> `*tramp/telnet
>        address@hidden' for details

Whee.  That's bad.  One of the commands should have worked.  That is,
I need to frob Tramp until it contains a method that works in your
case.

As a workaround, you can install Metamail or GNU recode on the Solaris
box and include the appropriate bin directory in tramp-remote-path.

But I would be really happy if you agreed to be my guinea pig to test
various fixes to see if they really work.  I need help from users, as
I don't have a local Windows to test with, and I don't have a great
variety of remote systems, either.  I can't promise to do the fixes
quickly, though :-(

> Of course the tramp/telnet buffer just has:
>
>       Process *tramp/telnet address@hidden interrupt
>
> But the debug buffer shows that not only does tramp try every possible
> de/encoding method, but that most of them work.  Yet still it fails.  Why?

Because it also checks locally, and Windows doesn't have
uuencode/uudecode :-(

> # Checking local encoding command `uuencode xxx' for sanity
> # Checking local encoding command `uuencode xxx' for sanity

Sadly, the results are not shown in the debug buffer :-(  But I bet
the result was that it didn't work.

> Shouldn't you just believe that I set 'tramp-encoding-command'
> correctly?

No, you're lying :-)

I've been working on a Lisp implementation of uuencode, which would
then also work on Windows, but I haven't been able to make it work.
So if anybody out there wants to make a Lisp implementation of
uuencode, please go ahead.

Hm.  Maybe it's a good idea, instead of doing it from scratch like I
tried, to use base64-encode and to transmogrify the result.  That
might work without too much work.  Hm.

kai
-- 
A large number of young women don't trust men with beards.  (BFBS Radio)