Re: Moving C++ in the direction of m-files

[Rik]

On 06/10/2017 07:08 AM, John W. Eaton wrote:
> On 06/09/2017 04:25 PM, Rik wrote:
>
>> Coders who are moving from Octave's m-file language to writing .oct files,
>> or contributing to core Octave, face a huge learning curve.  It's not just
>> that the language, C++, is different, but without reading all of the header
>> files it is hard to know what functions are available.  Often there is an
>> analog in the Octave language that is available in C++, but because we use
>> different naming schemes it is again difficult to find these functions
>> short of reading the header files.  Some examples are is_empty and
>> is_cellstr which have analogs of isempty and iscellstr.  How would you feel
>> about renaming the obvious analogs to match the Octave language names?
>
> I'd be OK with that.
>
> While I've been looking at the symbol table code I've had similar
> questions about whether I should rename "scope" to "workspace",
> "top_level" to "base", etc.  I haven't made those changes yet, but maybe
> I should?  OTOH, the symbol table is not something that people writing
> .oct files should normally touch directly.

I think this is an area where we don't need to follow Matlab.  For anyone
trained in computer science, "workspace" doesn't make as much sense as
"scope".  And, as you say, the symbol table is not a piece of Octave that
faces ordinary users.

--Rik