Path: news.mathworks.com!not-for-mail
From: "Alan B" <monguin61REM@OVETHIS.yahoo.com>
Newsgroups: comp.soft-sys.matlab
Subject: Re: cell mode execution
Date: Tue, 2 Apr 2013 16:00:13 +0000 (UTC)
Organization: The MathWorks, Inc.
Lines: 73
Message-ID: <kjevad$guu$1@newscl01ah.mathworks.com>
References: <gr5c91$4km$1@fred.mathworks.com> <514dfcd0$0$13514$c3e8da3$dd9697d2@news.astraweb.com> <kjcp9o$pal$1@newscl01ah.mathworks.com> <kjes7l$6eu$1@newscl01ah.mathworks.com>
Reply-To: "Alan B" <monguin61REM@OVETHIS.yahoo.com>
NNTP-Posting-Host: www-06-blr.mathworks.com
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
X-Trace: newscl01ah.mathworks.com 1364918413 17374 172.30.248.38 (2 Apr 2013 16:00:13 GMT)
X-Complaints-To: news@mathworks.com
NNTP-Posting-Date: Tue, 2 Apr 2013 16:00:13 +0000 (UTC)
X-Newsreader: MATLAB Central Newsreader 1916026
Xref: news.mathworks.com comp.soft-sys.matlab:792491

"Steven_Lord" <slord@mathworks.com> wrote in message <kjes7l$6eu$1@newscl01ah.mathworks.com>...
> 
> 
> "Alan B" <monguin61REM@OVETHIS.yahoo.com> wrote in message 
> news:kjcp9o$pal$1@newscl01ah.mathworks.com...
> > anyone <jaup@h/WSWLG{nR1qHe> wrote in message 
> > <514dfcd0$0$13514$c3e8da3$dd9697d2@news.astraweb.com>...
> >> On Fri, 22 Mar 2013 20:50:18 +0000, Matt Raum wrote:
> >>
> >> > The poor error reporting in cell mode execution is still a major
> >> > drawback to developing this way.
> >>
> >> so... you'd prefer it was silent, or were hoping Matlab would simply 
> >> re-code itself in some way?
> >>
> >> > I'm running MATLAB Version 7.13.0.564
> >> > (R2011b).  Is there still nothing that can be done to get more
> >> > informative error messages??
> 
> The way MATLAB executes code when you evaluate on a cell-by-cell basis is 
> slightly different from the regular way MATLAB functions or scripts run and 
> that difference makes doing what you request a little trickier than you 
> might expect.
> 
> >> See: Help -> breakpoint
> 
> Good suggestion anyone, but it won't work in this particular case. Running 
> individual code sections ignores breakpoints.
> 
> http://www.mathworks.com/help/matlab/matlab_prog/run-sections-of-programs.html
> 
> >> > For instance, I might execute a 200 line cell and get the following
> >> > message at the command line:
> >> >
> >> > Index exceeds matrix dimensions.
> >> >
> >> > Then I have to figure out which of the 200 lines caused this error --
> >> > exhausting.....it would be really great if a machine could do that for
> >> > me.
> 
> It would be great. But the cell execution mechanism is currently not able to 
> do so.
> 
> Some workarounds you can use right now:
> 
> 1) Select the first half of the code in the cell and right-click. Evaluate 
> that selection. If the error occurs, the problem is in the first half of 
> your cell's code; if not, it's in the second half. Repeat this binary search 
> until you've located the error.
> 2) Break your 200 line cell into smaller cells. The less code you have to 
> look through to isolate the issue, the easier it will be to locate.
> 3) Run your code normally (not cell-by-cell) with breakpoints as anyone 
> suggested, locate the issue, fix it, then return to your workflow of 
> executing cell-by-cell.
> 
> >> damned computers, eh? ...always trying to do what you say, not what you 
> >> mean.
> >>
> >
> > It seems a reasonable request, considering that A) the machine CAN do it 
> > for us, in the context of a function, and B) it was "on the radar" as of 
> > four years ago.
> 
> It is still on the radar. It has not come in for a landing yet as of the 
> current release R2013a.
> 
> -- 
> Steve Lord
> slord@mathworks.com
> To contact Technical Support use the Contact Us link on 
> http://www.mathworks.com 

Thank you, Steve. I appreciate the explanation, and I understand that implementing it is not trivial. I do not appreciate being treated like an ignorant child by 'anyone', for making a reasonable feature request for software that I pay for.