[Date Index] [Thread Index] [Author Index]

Re: "Abort Evaluation" does not work

• To: mathgroup at smc.vnet.net
• Subject: [mg111949] Re: "Abort Evaluation" does not work
• From: AES <siegman at stanford.edu>
• Date: Sat, 21 Aug 2010 04:17:37 -0400 (EDT)
• References: <i4lnr2$7gs$1@smc.vnet.net>

In article <i4lnr2$7gs$1 at smc.vnet.net>,
 Themis Matsoukas <tmatsoukas at me.com> wrote:

> As Simon and Seeziwa explained, the loop 
> does respond to 'Abort Evaluation' (but in a way that leaves the user 
> wondering): it breaks the loop but it does not interrupt the printing of the 
> i's that have been queued to the output. In this particular example, the loop 
> is executed very fast (too fast to Abort during its evaluation)  while 
> printing, which takes much longer, does not respond to Abort and so it goes 
> on till completion, leaving the impression that nothing was aborted. 
> 
> The question then is, would it be too difficult for Mathematica to implement 
> an abort that kills the Print? When I debug I make liberal use of Print 
> statements and sometimes I do run into this problem of runaway printing that 
> I cannot Abort.

Many, many of us have of course run into the same problem.

So, another question would be:  if you look up "Abort" in Mathematica's 
documentation, will it give you a little warning about this, so you'll 
understand what's happening?

Wanna bet?

[Scroll down a screen or so to see what the "Additional Information" 
section at the top of the Abort Help page says.]













You can call Abort anywhere within a computation. It has the same effect 
as an interactive interrupt in which you select the abort option. 

You can use Abort as an "emergency stop" in a computation. 

Once Abort has been called, Mathematica functions currently being 
evaluated return as quickly as possible. 

In an interactive session, the final result from an aborted computation 
is $Aborted. 

You can use CheckAbort to "catch" returns from an abort.