Advanced Portfolio

Henry M. Benson hmb at
Thu Dec 21 15:49:14 EST 2006

Andrew Sackville-West wrote:

> hi, 

Hi, and thanks for the response!

> i did a bunch of work on the advanced report portfolio, about a year
> ago and as a result, fixed some broken stuff and introduced many
> broken things :(.

LOL - understood :-(

> I don't have a lot of time to devote, but would be happy to help in
> whatever way I can. My biggest issue was I ended up working on things
> I didn't really understand. My original goal was to implement some
> display settings for a user who requested it. I ended up trying to fix
> some other broken behavior in the report (some of which succeeded, I
> think), but left other parts of it undone. The biggest issue I think,
> is that I tried to implement a way of really tracking gains based on
> three different models - FIFO, FILO and average price (i think). I did
> not however, handle all the possible methods that the value of shares
> could change (splits, reinvestments, etc) resulting in bad results for
> any but the simplest cases. 

Yes, that's what I'm seeing.  I have a rather complicated portfolio with
all of these - it'll make a great test case.  If it's any consolation,
Quicken gets it wrong too...

> There were also some difficulties in
> getting reliable data because we couldn't force users to make pricedb
> entries. I think I got that more or less handled by using register
> data when pricedb data wasn't available. 

I think that's working properly.

> The report itself, is really pretty simple (IIRC) and I'm happy to
> pitch in a bit. Have you looked over the code yet? and are you
> familiar with scheme/lisp? 

I looked over the code and can follow the basic flow.  I have some
lisp experience from my MIT days, a disturbingly large number of years
ago.  :-)  I'll give it a try and see how it goes.

> I think if we could get a really good set a
> requirements for how the report handles particular cases, it could be
> shaped up fairly easily. I'll checkout a new tree here in the next
> couple days and take a look myself.
> A

Yes, I agree, requirements would be a great place to start.  I'd be
glad to start on that.  I can also make note of what I've seen, review
the open bugs, and note what works and what doesn't.  I can probably
get to that next week sometime.  How does that sound?

Thanks again and best regards,

More information about the gnucash-devel mailing list