Possible bug in qofinstance.cpp: qof_instance_kvp_remove_guid

[John Ralls]

> On Jan 30, 2026, at 09:26, John Ralls <jralls at ceridwen.us> wrote:
> 
> 
> 
>> On Jan 30, 2026, at 08:22, Stefan Koch <stefan.koch.micro at gmail.com> wrote:
>> 
>> I don't know if this is the expected behavior for the higher level
>> code.  It seems a little funky that sometimes the date/time is not
>> stored in the list, or that the keys could be different, but that is
>> likely OK.
>> 
>> I also don't know how this would affect real usage. There is no
>> mentions (that a quick google search could find) of peer splits in the
>> documentation. I did chance this up that the merge is only used in
>> xaccScrubMergeLotSubSplits.  But that is called in more places I am
>> not familiar with. If you find a bug for this I can make a pull
>> request with this.
>> 
>> You mentioned capgains testing.  If you point me in the right
>> direction, I can take a look, but I don't know where to start at this
>> point.
> 
> Stefan,
> 
> I don’t know either, which is why I want better testing of the usage. I called it “capgains” because that’s what drives the lot scrubber. The existing test is libgnucash/engine/test/test-lots.c and as you can see it’s a very simple test that doesn’t really exercise the peer-split code at all. 

Stefan,

One more point: The lot scrubber is used and there aren’t a lot of complaints so we have to start from the assumption that the code does work in spite of the apparent flaws you’ve found.

Regards,
John Ralls

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.gnucash.org/pipermail/gnucash-devel/attachments/20260130/40d0a6a7/attachment.htm>