Gnucash 2.4.0 (Windows)

David T. sunfish62 at yahoo.com
Thu Dec 30 22:48:13 EST 2010


Colin--

--- On Thu, 12/30/10, Colin Scott <gnucash at double-bars.net> wrote:

> From: Colin Scott <gnucash at double-bars.net>
> Subject: Re: Gnucash 2.4.0 (Windows)
> To: sunfish62 at yahoo.com, warlord at MIT.EDU, mortimer.hobart at gmail.com, gnucash at double-bars.net
> Cc: gnucash-user at gnucash.org, gnucash at double-bars.net
> Date: Thursday, December 30, 2010, 4:02 PM
> 
> > The problem is in gconf, not Gnucash.
> 
> I understand that, and yes, it is a significant
> point.  Nonetheless, if the basic assumption (that in
> Windows you must have an internet connection in order to
> start gnucash 2.4.0) is true then this is a serious bug in
> gnucash, even if it is not in the gnucash code.  (This
> latter point is IMHO too fine a distinction to be meaningful
> or relevant to the user, however relevant it may be - and
> undoubtedly is - to the developers!)
> 
> > This has been true since 2008 at least 
> 
> I note from the Bugzilla entry you reference in a later
> message that the bug actually dates back to 2006.  I do
> not see this as in any way excusing the inclusion of so
> serious a bug - very much the opposite, in fact, and in my
> view calls into question the soundness of subsequent design
> and/or implementation decisions.  I accept that to some
> extent developers are in the hands of their library
> providers, and I also note from the same Bugzilla entry that
> the problem only seems to affect certain versions of
> Windows, which I accept does provide some (limited!)
> mitigation.  However, my presumption would still be
> that the developer should be in a position to get problems
> fixed, or fix them themselves, or find a way of by-passing
> them, before product release, or at least that there should
> be some warning of significant malperformance.  That
> apparently such a bug remains in place some 4 years after
> being reported does not imply that such matters are taken
> sufficiently seriously.
> 

It is not clear from this whether you read all the comments in that bug; the developers in question (both for ORBIT and Gnucash) put quite a bit of time and effort into the problem, and believed that they had created a remedy for the problem back in 2007. But because of the quirkiness of the problem and the lack of understanding of what precisely was going on, none of the developers were certain that they had come up with an appropriate solution. This was complicated by the fact that none of the developers were able to view the problem firsthand, and the fix they implemented never received any further feedback from the Windows user base. Until another person added a new comment in 2008 saying the problem had cropped up again; this last comment was never replied to.

As I later said, it might be appropriate to take this up with the gconf/ORBIT folks.

David

> I continue to liberally sprinkle "apparently" and similar
> words throughout my text, as there seems to remain some
> doubt as to the actual extent of the problem, and whether
> (for example) if it didn't bite in 2.2 it will or will not
> bite in 2.4.  Therefore my complaint is, I fully admit,
> somewhat hypothetical!  Some authoritative input on
> this would be welcome, preferably *before* before this
> thread develops into too much of a witch-hunt.  :-)
> 
> Colin
> 


      


More information about the gnucash-user mailing list