Opening balance reconciling investment account incorrect.

[Mike or Penny Novack]

>  Since I am looking at Gnucash as a potential user, I am alarmed to see
>that there are over 1000 open bugs in Bugzilla.  I wonder if this is a
>warning about the reliability of the software.
>
>Regards.
>  
>
As a person who before retirement made my living designing and writing 
(and debugging) software let me make an important point that might not 
be obvious to the average user.

The simple answer is NO.

That's the number of REPORTED bugs, not the number that turn out to be 
actual bugs. All but a very few of these, if and when analyzed, will 
turn out to be user error of some sort. Now that CAN be flawed 
documentation (which is different from a program bug) but again, most 
user error will be things like:

1) Expecting the program to do something other that it does or that the 
documentation implies that it does. Oddly many users imagine that a 
computer can read their minds.
2) User didn't actually do what the user thought was done (thought 
entered one thing and that's what is on the bug report but actually 
entered something else. The fickle finger problem. Very common.

The FIRST step a developer has to do when tackling a bug report is to 
reproduce the error. If they can't do that (or know from their own 
experience using gnucash that this feature works for them) unlikely to 
take it much farther so the "bug report" remains unsolved. In my working 
days I was paid to resolve these to the bitter end; either find and fix 
the bug or prove that it was user error. But I didn't need user 
cooperation to do this as I could set up the exact situation by 
extracting the master record and transaction from backup and set these 
up in a test database. So, for example, simple user error easily 
detected (hey, the actual transaction had a number transposed from what 
the user reported to have been entered -- another case of fickle finger. 
Once a case of gross incompetence (out of 12 items on the transaction, 7 
were completely different numbers!).

If YOU, as a user, have a particular vested interest in a bug you have 
reported, it is going to take more than "just submit a report". You 
would need to find a developer to work with and commit some of YOUR time 
(because without the backups YOU have from immediately before this error 
the developer cannot be certain of reproducing the EXACT situation). If 
you don't have backups from immediately before there really isn't much 
that can be done. As a developer, I would take MUCH more seriously a bug 
report submitted where the user submitting the report had gotten their 
files into the situation immediately before, done a save (created 
backup), tried again whatever caused the error and submitted the report 
attaching the backup. Or of course a user that would work with me to do 
that after the fact.

Michael D Novack, FLMI