Re: 23.0.60; Crash after emacsclient -n foo.txt

[Dan Nicolaescu]

Reiner Steib <address@hidden> writes:

  > On Sat, Feb 02 2008, Dan Nicolaescu wrote:
  > 
  > > Reiner Steib <address@hidden> writes:
  > >   > M-x report-emacs-bug RET wrote:
  > >   > 
  > >   > > Please describe exactly what actions triggered the bug
  > >   > > and the precise symptoms of the bug:
  > >   > 
  > >   > After `M-x server-start RET' and running `emacsclient' from an xterm,
  > >   > Emacs crashed:
  > [...]
  > >   > The emacsclient binary belongs to Emacs 22.1, not to the CVS trunk as
  > >   > the running emacs server instance.
  > >
  > > This works fine for me. But instead of:
  > >
  > >   > > In GNU Emacs 23.0.60.2 (i686-pc-linux-gnu, GTK+ Version 2.10.6)
  > >
  > > I have:
  > >  GNU Emacs 23.0.60.1 (i686-pc-linux-gnu, GTK+ Version 2.12.3))
  > 
  > I can't reproduce the crash (I still have in a screen session,
  > though), but I get an error from emacsclient:
  > 
  > | [...]/cvs-HEAD/emacs$ [...]/cvs-HEAD/i686/src/emacs -Q
  > | 
  > | [1]+  Stopped                 [...]/cvs-HEAD/i686/src/emacs -Q
  > | [...]/cvs-HEAD/emacs$ bg
  > | [1]+ [...]/cvs-HEAD/i686/src/emacs -Q &
  > | 
  > | [`M-x server-start RET' in Emacs]
  > | 
  > | [...]/cvs-HEAD/emacs$ /usr/bin/emacsclient vpath.sed 
  > | Waiting for Emacs...
  > | -error Unknown&_command:&_[...]/cvs&-HEAD/emacs/vpath.sed
  > 
  > When adding `-n', I get no error, but the file isn't opened in the
  > emacs server session.  (I don't think that the client tries to talk to
  > the crashed server, since the socket's timestamp is new.  IIRC, `M-x
  > server-start RET' overrides previous sessions.)

That's better given that you don't see any crash. After the multi-tty
merge emacsclient used to refuse to connect to a different server
version. It doesn't seem to do it anymore. But IMHO there are more
interesting issues to solve than this.