error in advanced portfolio report (svn r16781)

[David Reiser]

On Jan 2, 2008, at 8:43 PM, Andrew Sackville-West wrote:

>> a few lines above the scm that appeared in the last report, I get:
>>
>> * 17:59:57 DEBUG <gnc.scm> b-list is ((#<<gnc-numeric> num:  
>> 4189799520000
>> denom: 204700000000> . #<<gnc-numeric> num: 72682214900000000 denom:
>> 418979952000000>)) b-units is #<<gnc-numeric> num: 0 denom: 10000>  
>> b-value
>> is #<<gnc-numeric> num: 0 denom: 100> b-method is average-basis
>
> You have a txn with no shares and no value in the split that touches
> this account. What is the other side of that transaction? What are you
> trying to do with that txn?
>
> It's a real bug, no doubt, but I want to make sure it gets handled
> properly -- there is no code to handle that case, duh, so its
> <#unspecified>. So if you can find that txn, and tell what it looks
> like and what its trying to do, then I can fix it straight-away.


OK. My error. But the register wasn't complaining, either. What  
happened was: 3 years ago, I couldn't figure out why my tracking of my  
wife's TIAA-CREF contributions left me 0.002 shares low in one of her  
three funds. So I added 0.002 shares at zero cost. Probably would have  
worked out right if I had used the stock split wizard, but I didn't  
think of that. It has been tracking fine since (and I now import all  
the transactions via ofx...).

It turns out I had both the 0.002 shares and the 0 value balancing  
split coming out of the stock fund account. (Remember, the register  
wasn't complaining, and no imbalance account.). Once I put the $0  
split into the parent 'brokerage' account, the report works.

Can you tell me why the advanced portfolio report gives me transaction  
based calcs even though the preference for price list use is selected?  
With transaction based calculations, the retirement accounts all show  
0% total return (which isn't what I was expecting).

Thanks for all your work on the report. It is now fast enough to be  
usable. Before I never bothered to figure out the options because any  
change took so long to test.

Dave
--
David Reiser
dbreiser at earthlink.net