Re: Matlab-style code in test/classes

[Julien Bect]

Robert T. Short wrote
> On 01/28/2013 01:50 PM, Ben Abbott wrote:
>> On Jan 28, 2013, at 1:51 PM, Jordi Gutiérrez Hermoso wrote:
>>
>>> Does the Matlab-style code in test/classes have to stay? We don't
>>> generally make this concession elsewhere in Octave. Why do we do it
>>> here?
>> I assume the Matlab-style syntax decision was done to make it convenient
>> to run the tests in Matlab and ensure Octave's implementation is
>> compatible?  I don't know if we need to  continue with that approach, but
>> it is very useful in the beginning.
>>
>> For example, I've been playing with the classdef branch.  Which still
>> lacks many features.  I've taken to using Matlab syntax so that I may
>> verify when I'm doing something wrong and when a feature is absent.
>>
>> Ben
>>
>>
>>
> That is exactly why I did it.  I agree that it doesn't need to stay, but 
> I was trying to validate compatibility as well as testing octave - all 
> the while trying to include inheritance into octave.
> 
> Bob

In my humble opinion: it's better if it stays, if you want to maintain
compatibility in the long run, when both Octave and Matlab will keep on
evolving. 

This is why I decided to avoid two- or three-arguments assert() calls when
writing patch #7993 (https://savannah.gnu.org/patch/?7934)... which is what
triggered Jordi's question in the first place ;-)

Julien





--
View this message in context: 
http://octave.1599824.n4.nabble.com/Matlab-style-code-in-test-classes-tp4649207p4649255.html
Sent from the Octave - Maintainers mailing list archive at Nabble.com.