[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Getting Tablicious on Octave Forge

From: Andrew Janke
Subject: Re: Getting Tablicious on Octave Forge
Date: Tue, 21 Jan 2020 05:53:25 -0500
User-agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.14; rv:60.0) Gecko/20100101 Thunderbird/60.9.1

On 1/21/20 4:52 AM, Kai Torben Ohlhus wrote:
> On 1/21/20 3:01 PM, Julien Bect wrote:
>> Le 18/01/2020 à 20:18, Andrew Janke a écrit :
>>> Hi, Octave folks,
>>> I think my Tablicious package is now stable enough (with the version
>>> 0.3.0 release) for it to be exposed to the general Octave user
>>> community. I would like to get it on Octave Forge as an External
>>> Package, so people can make use of it, and I can get some user feedback
>>> and bug reports.
>>> https://github.com/apjanke/octave-tablicious
>>> If you're okay with this, could someone help me through the process of
>>> submitting the package to Octave Forge?
>> Hi Andrew,
>> Why not propose your code for inclusion in Octave directly ?
>> Table and related classes (categorical, etc.) have been available in
>> Matlab since R2013b...
>> Just my two cents (I think Carne suggested the same a while ago).
>> @++
>> Julien
> +1 for Julien's statement.  These are features I like to see in Octave.
>  The only thing that prevented me from pushing further is your statement
> from last May [1]:
>> "I'd say it's still only about 50% complete"
> But are those 50 % ready to be included?
> Kai

I'd say it's up to 80% now, actually! And the missing parts are stuff
like table I/O and timetable. Which is why I think it's ready for
general users.

It's encouraging to hear this would be wanted in core Octave! But I'd
still rather keep it as a separate package for the time being.

1) I'd like to maintain direct commit control over it for now, and keep
it in Git (which I know much better than Mercurial). There's going to be
a lot more changes coming in as I get that last 20% down. I also have
ambitions to add a port of Python's xarray[1] to it, which will be
significantly more work. That's going to result in a flurry of commits.

2) I'm not yet confident in the quality of the code to put it in core
Octave. I really need to get some actual users besides myself testing
and banging on it before I'm going to feel good enough about it to
submit it to core Octave.

3) Adding table I/O (readtable, writetable, xlsread, etc) is probably
going to take a dependency on the IO package and its current ongoing
development, per https://savannah.gnu.org/task/?15419. My understanding
is that core Octave cannot take dependencies on any packages.

4) Maybe more important, I'd like to support it for Octave 4.4.1, both
for myself and other users. Since the Mac GUI in Octave 5.x and 6.x
isn't working on macOS, and I'm a GUI user, I'm stuck on 4.4.1 for the
time being, and I think other Mac users may be too. And maybe even
support Octave 4.0.x for Linux users still using their distro-provided


[1] http://xarray.pydata.org/en/stable/

reply via email to

[Prev in Thread] Current Thread [Next in Thread]