Re: Proposed change to `make-readline-port'

[Dirk Herrmann]

On 7 Mar 2001, Neil Jerram wrote:

> >>>>> "Dirk" == Dirk Herrmann <address@hidden> writes:
> 
>     Dirk> On 5 Mar 2001, Neil Jerram wrote:
> 
>     >> (We could later think about removing the line-based assumption,
>     >> but I'd rather not consider that change right now.)
> 
>     Dirk> If I understand things right, you have to add the newline
>     Dirk> because it is not part of the characters that are returned
>     Dirk> by readline.  Why not just let the reader function append
>     Dirk> the newline character before returning the string?
> 
> I think it's a question of usage.  What it would be more general to do
> as you suggest, the generality is not very helpful if it just forces
> all practical users of `make-buffered-line-input-port' to return
> `(string-append str "\n")' rather than `str'.

Well, if I remember right, you actually wanted to provide a general
buffered-input-port rather than just a line-buffered input port, and
that's what my suggestions was targeted at.

But, again:  I am not any more convinced that the current implementation
of line-buffered input port deserves a module of its own, since it is too
specifically designed around the use in the repl.  We should, IMO, rethink
that decision.

Best regards,
Dirk Herrmann