On Saturday, September 1, 2012 9:02:33 PM UTC+5:30, Ramchandra Apte wrote:
> On Friday, 31 August 2012 03:27:54 UTC+5:30, vasudevram  wrote:
> 
> > I wrote PipeController recently to experiment with doing UNIX-style pipes 
> > in Python.
> 
> 
> Doesn't the pipes module already do this?

Yes. As Ian Kelly points out here:

http://www.mail-archive.com/python-list@python.org/msg335943.html

the pipes module of Python does deal with actual UNIX pipes.

But I had mentioned this - Python's pipes module - in my first post, some 
months ago, about doing UNIX-pipes in Python ( on my blog, here: 
http://jugad2.blogspot.in/2011/09/some-ways-of-doing-unix-style-pipes-in.html 
), and that first post was in turn linked to, in the current post under 
discussion ( i.e. this one: 
http://jugad2.blogspot.in/2012/08/pipecontroller-v01-released-simulating.html ).

To Ian Kelly:

> No, that deals with actual Unix pipes. This appears to be about pipelined
> processing within a single program and not IPC; the description "Unix-like"
> is a bit misleading, IMO.

I guess it can be interpreted as a bit misleading, but it was not intentionally 
so. The way I meant it was that PipeController tries to achieve _roughly_ 
similar functionality, of composing a program out of components, as UNIX pipes 
do. (That too, only a small subset, as I mention in my blog post). In the case 
of UNIX the components are commands, in the case of my Python approach the 
components are functions. Also, I used the term "UNIX-style pipes", not "UNIX 
pipes", to stress that it was the overall concept of a pipeline that was being 
simulated, not the exact design / implementation details, and you can also 
notice that my first post has links to many different third-party ways of 
simulating UNIX-pipes in Python, some of which are not like actual UNIX pipes 
at all, in their implementation. But I agree that your description of 
PipeController as "pipelined processing within a single program" is more 
accurate.

I could have used a better choice of words but there was no intention to 
mislead.

- Vasudev Ram
www.dancingbison.com
jugad2.blogspot.com








-- 
http://mail.python.org/mailman/listinfo/python-list

Reply via email to