Ok,
After undoing the sticky tag (doh) from when I originally rolled back my CVS copy
of the avg-bal report, most of the issues disappeared. The original problem
(which was why I rolled things back) with early transactions not being tallied and
causing the balance to be off no longer presents any issue (running balances are
again tracked in the report-cool,thanks). Its great to see everything add up
correctly + stock values plotted again!
Last night I had an issue where amounts earlier than 1/1/2000 were not showing
(thought it might have been related to the orig. problem), however I cannot
reproduce it today so maybe it can be chalked up to general wierdness.
Dave Peticolas wrote:
>
> > Anyway, somebody familiar with the query code will probably have to do some
> > snooping around to solve this issue !!!
>
> > Sitting back and considering the repercussions of this in a broader scope: it
> > might be wise at some time to have a set of "check" operations with known
> > "correct" results such that correctness of the calculations (reports, account
> > balances, etc) can be demonstrated, preferably with some kind of script. This
> > is often done with computational packages by having a "make check" section of
> > the makefile.
> > Something to consider !
>
> That is a very good idea.
Was thinking about this while recompiling yesterday and decided that
this would be a perfect example of value which a commercial entity
could add to this project: validation service. A "boxed" version could included
"verified" copies of all scripts/code/binaries (new set of scripts and encrypted
checksums periodically).
The prog could accept all scripts/plug-ins/etc but warn if you try to run one which
has not been
verified. Many people are nervous about trusting something developed on the internet
handling their money,
but having some kind of security blanket would certainly promote acceptance. It also
would be
entirely complementary to develpment efforts without radically controlling them (other
than pointing out calc. errors).
Of course, OSS fairness would dictate that any user would be allowed to generate
his/her own
validation keys for personal use. You know, "good enough for me"...
Just some random thoughts.
--
Matt Martin
[EMAIL PROTECTED]
600 West Grove Parkway #1042
Tempe, AZ, 85283
(480) 775 2660
--
Gnucash Developer's List
To unsubscribe send empty email to: [EMAIL PROTECTED]
