import/export bounty

[Christian Stimming]

Am Dienstag, 23. September 2008 22:37 schrieb Ian Smith-Heisters:
> >> Now I find that hiring a GnuCash specialist is impossible, so I'm
> >> willing to work with someone more familiar than myself, but not a
> >> specialist per se.
> >
> > Not to discourage you _at_all_, but just so you're aware what kind of
> > investment would be required to make a significant impact on GnuCash,
> > you should probably be thinking in the $15k-$50k+ range before this
> > starts looking viable.
>
> How did you arrive at those figures? It'd be useful to know what needs
> to get done before any features can be added, and how much time is
> involved since the $ sum will differ depending on who's doing the
> work. Nonetheless, those figures aren't at all surprising.

I agree with the range of the figures. My estimate is that a developer 
familiar with gnucash can implement a significant part of your 
proposed/requested feature as a 1 month full-time project. Which would end up 
in the price range Chris mentioned.

> Has GnuCash considered doing a Google Summer of Code? I would be
> unqualified to act as a mentor, but I'd be willing to do the tedious
> administravia. Again, I'd need the support of the core team.

The "tedious administrativa" was the easiest part. The difficult part is the 
actual mentoring - talking to the student on a weekly basis. If the GSoC 
project really arrives at results, 10% of the student's time would have to be 
invested by the mentor as overhead. As the student invests 3 months 
full-time, you can see how much person-days are required by the mentor.

But as Derek already said, we have tried it last year and the effect wasn't as 
significant as we had hoped. We might give it a try next time, but the 
student work definitely does not come "for free" but instead the mentors have 
to invest significant time as well.

In any case, the succeeding projects showed that a developer *not familiar* 
with the gnucash codebase could implement a useful feature within 3 months 
work (instead of the 1). If you can find someone who gets paid for or donates 
3 months working on gnucash, I would probably volunteer to donate some of the 
required overhead time to mentor them into the gnucash codebase.

Regards,

Christian