[Spambayes] Spambayes 1.1a1 feedback - Outlook add-in, Notifications T ab

[Tony Meyer]

> Have been trying out the 1.1a1 alpha release of the
> Outlook add-in.

Great - thanks for both trying it out and for the feedback; it's much
appreciated.  Please let us know if there are way that you think this could
be improved.

> Good Sound - Functions as expected.  There is some confusion between 
> Window's settings to play a sound on receipt of mail and the setting 
> in Spambayes.  Which takes precedence?

The Windows one operates completely independently, but I would suspect that
ours would come after that, especially if the background filtering is on,
since Outlook then 'knows' about the mail some time before SpamBayes does.
We play the sound asynchronously, so I guess it could actually overlap with
the Windows one; I'm not sure.

Since the Windows one is out of our control (I don't think we can even
detect whether it is playing), I'm not sure there's anything we can do about
this.  I imagine that most people would simply replace using the Windows one
with the SpamBayes ones.  (Although the Windows one would play for any mail
received that SpamBayes doesn't filter).

> Unsure Sound - Can't get this to work at all.  Have sent myself mail 
> previously classified as spam so that Spambayes puts this in my Junk 
> Suspects folder, but sound does not play...

Odd.  Someone else has reported problems with the unsure sound as well, but
I just tried it here (I hadn't tried the sounds myself before; Kenny did
those) and it works for me.  When the unsure message was received, did you
get any other mail at the same time?  If you got ham, then the ham sound
would play instead of the unsure.

If you set the log verbosity to 3 (SpamBayes->SpamBayes
Manager->Advanced->Diagnostics) then there will be a note in the log each
time any of the sounds are played (as well as a *lot* of extra stuff).
Could you try that and then try your test again?  It would be good to know
whether the note is there in the log, or if there is an error, or what.
 
> But... Some of the flags/options/check-boxes on the Advanced tab seem 
> to interact with the setting in the Notifications tab.  Am unsure of 
> the relationship between the settings and the effects. eg disabling 
> background filtering disables sounds for junk mail, but Spambayes 
> still processes the spam into my Junk Mail folder - strange!
 
There shouldn't be any relationship, other than if background filtering is
enabled, then there will be an additional delay before the notification is
triggered (the notification is triggered when the message is processed,
which is after the background filtering delay, if it is enabled).

Disabling background filtering didn't disable sounds for spam for me here.
It would be great if you could try this with the log verbosity at 3, as
well.

> I've been waiting/hoping for an improvement (for the Outlook add-in) 
> that plays different sounds when good mail (ham) is received, but 
> ignores Spam.  In my opinion, getting the function that plays a
> (different) sound for new suspect mail is far more important than 
> playing a sound for certain spam.

I believe that if you set sounds for ham (good) and unsure, but leave spam
unset, then you'll get this behaviour (although if both ham and unsure
arrive at the same time, then you'll only hear the ham sound).

> Overall - an EXCELLENT addition to an already invaluable utility!!!  
> Everyone should use this.
 
Except those that have to work in quiet offices <wink>.  It's possible that
(in time) there will be another addition to the notification tab, which
would control a tray icon, which could be a replacement for the Outlook "new
mail" icon that appears (one of the most requested changes), which would
suit those people.  Sounds are more straightforward, though, so it's best if
we get those figured out first.

=Tony.Meyer

-- 
Please always include the list (spambayes at python.org) in your replies
(reply-all), and please don't send me personal mail about SpamBayes.
http://www.massey.ac.nz/~tameyer/writing/reply_all.html explains this.