Re: I'd like to rearrange the stream_t API slightly

[Sam Roberts]

Allo,

I committed these changes, including those needed to build the project.

As far as I know, without running the imap and pop servers (and making
a spool for them to run on) everything works ok.

If anybody finds any problems, let me know!

Bon soir,
Sam

Quoting Alain Magloire <address@hidden>, who wrote:
> > How about:
> > 
> >  extern int stream_file_create __P ((stream_t *, char* filename, int flags))
> >  extern int stream_tcp_create __P ((stream_t *, char* host, int port, int 
> > flags))
> >  extern int stream_stdio_create __P ((stream_t *, FILE *))
> >  extern int stream_mmap_create __P ((stream_t *))
> >  extern int stream_memory_create __P ((stream_t *, size_t))
> >  extern int stream_fd_create __P ((stream_t *, int))
> >  extern int stream_buffer_create __P ((stream_t *, stream_t, size_t)) 
> > 
> > And:
> > 
> >  extern int stream_open __P ((stream_t));
> > 
> >    (Open's the stream, how it does this is its business, the information
> >     it needs to open the resource was specified in the stream_X_create()
> >     function.)
> > 
> >  extern int stream_close __P ((stream_t));
> >  extern int stream_read __P ((stream_t, void *, size_t, off_t, size_t *))
> >  extern int stream_write __P ((stream_t, const void *, size_t, off_t, 
> > size_t*))
> >  ... etc ...
> > 
> > 
> > Does this make more sense?
> 
> Yes.  I agree.
> However, if you do the changes, send a note so we can update the other
> utilities in mailutils/*.

-- 
Sam Roberts <address@hidden> (Vivez sans temps mort!)