[plt-scheme] shared state across multiple REPLs

From: Carl Eastlund (cce at ccs.neu.edu)
Date: Thu Jan 29 13:33:17 EST 2004

Well, I have personally come to like what bash does with multiple
histories; others may or may not agree.  Each process (each shell) reads
the global history when it starts, and retains a local history as it runs.
When it terminates, it appends its local history to the global history.
Thus when you open a new shell, it sees the history of each shell that has
terminated before it, but not of those that are still running.

--Carl

On Sat, 24 Jan 2004, Robby Findler wrote:

>   For list-related administrative tasks:
>   http://list.cs.brown.edu/mailman/listinfo/plt-scheme
>
> This is (somewhat) by design. It is the same basic problem as the
> history in your shell when you run multiple xterms in X11. Eventually,
> there must only be one history that is saved when DrScheme exits. I
> chose to make only one history ever and all interactions get put at the
> end of that single history. You'll find that if you hit meta-p one more
> time at the end of the sequence you describe below, you'll get (foo).
>
> If you'd rather something else, I'm all ears, do be sure to explain
> what happens after restarting drscheme.
>
> Robby
>
> At Sat, 24 Jan 2004 11:10:03 -0500, David Herman wrote:
> >   For list-related administrative tasks:
> >   http://list.cs.brown.edu/mailman/listinfo/plt-scheme
> >
> > What's the rationale behind the following DrScheme behavior?
> >
> > Open two DrScheme windows. In the first window, define a `foo' function
> > in the definitions window and execute. In the second window, define a
> > `bar' function in the definitions window and execute. Go back to the
> > first window and type (foo) at the REPL. Then go to the second window
> > and type (bar) at the REPL. Now return to the first window and hit
> > Meta-P. You get (bar) as the last command, not (foo).
> >
> > This means there's a global state representing the recent-command list,
> > shared across multiple windows. Is this by design, an oversight, or
> > just not important enough to fix?
> >
> > Thanks,
> > Dave
> >
> >
>


Posted on the users mailing list.