Re: Incompatible functions + paid development

[Ben Pfaff]

On Fri, Sep 13, 2013 at 10:38:51AM +0200, Fredrik Clementz wrote:
> I'd like to ask some questions to you developers of PSPP about how you feel
> about a few things.
> 
> 1. There is obviously loads of features left before PSPP would get anywhere
> near on par with SPSS in functionality. However, there is a lot of
> statistical functions that isn't supported by SPSS which makes a lot of
> sense to have a in a statistical software. Now my question here is, is PSPP
> open for items such as this yet, or is the ambition to first get close to
> SPSS? The reason I ask is because obviously new features not in SPSS would
> mean that Syntaxwise PSPP would get functions which wouldn't work in SPSS
> for instance. To examplify some of the features I'd personally like to see
> is
> - Rim Weights
> - LISREL functions
> - PLS
> - Chi2 enabled Tables (+-sign).
> - Multiuse variables (Eg: reusing the same variable for 2 things in 1
> function without multiplying it)
> etc etc.
> So is it worth putting these in the Bugzilla?

I don't have any objection to add features to PSPP that are not in SPSS,
but I don't see too much value in adding the feature requests to the bug
tracker.

> 2. The second question is the topic about code bounties and hired
> development. Let's assume for a second that I desperately want a feature
> in. How would it be possible to hire someone to get a feature in? Would the
> current team be restrictive on this and see it more as a problem or as a
> possibility? The reason I ask is because I'm trying to push locally that
> some companies would pick the software up for usage, and if they do, a
> great way to help them adopt is to simply easily push for a way to buy in
> features so to say, so that it gets more usable for them. This is really
> speculative at this point, but I do see a point in investigating it.

Paid development is fine.  One or two features have already come into
PSPP that way.  The code is subject to the same code quality and license
rules as any other code that comes into the tree.