Re: Package Contribution Procedure and Best Way to Interpret Name, Value Arguments?

[Juan Pablo Carbajal]

On Sun, Jan 1, 2017 at 3:34 PM, Chris <address@hidden> wrote:
>
>
> On 01/01/17 11:09, Juan Pablo Carbajal wrote:
>>
>> On Sat, Dec 31, 2016 at 5:43 PM, PhilipNienhuis <address@hidden>
>> wrote:
>>>>
>>>> Question 2)
>>>> Now that statelevels exists, I can write dutycycle. This function will
>>>> have
>>>> a few *Name,Value* variable arguments. Is there a standard way to
>>>> interpret
>>>> these?
>>>
>>> There's the inputparser class in core Octave. I've never used it but I
>>> think
>>> it can do the job.
>>
>> It depends on the maximum number of expected arguments. If you plan to
>> have a lot of configurable options, then I strongly suggest
>> InputParser to keep the code clean (it also handles default values and
>> input validation). If you are thinking of just a few options (one,
>> two?) then do the switch or if block.
>>
> Thank you. I'll check it out.
>
> In regards to development of high level functions. I think it would be good
> to have a list of instructions for people who want to submit a function.
> Including how to test, how to make a changeset etc.
>
> This would achieve two things
> 1) People who are very capable of writing high level functions might have
> never used version management. - I'm thinking about the scientists I work
> with.
> 2) This might streamline the process for including new functions.
>
> What do you think? I'm happy to write a wiki article if appropriate.
>
> I'm new here but I am really keen to help the octave project.
>
> B/W,
>
> Chris
>

There is something in the manual already
https://www.gnu.org/software/octave/doc/v4.0.3/Basics-of-Generating-a-Changeset.html
You can always improve the wiki
http://wiki.octave.org/Mercurial