[antispam-f] Re: Next version of AntiSpam coming up

  • From: Jeremy Nicoll - freelists <jn.flists.73@xxxxxxxxxxxxxxxxxxxx>
  • To: antispam@xxxxxxxxxxxxx
  • Date: Wed, 29 Oct 2008 23:34:46 +0000

Richard Porter <ricp@xxxxxxxxxxxxxxxx> wrote:

> On 25 Oct 2008 Jeremy Nicoll - freelists wrote:
> 
> > Richard Porter <ricp@xxxxxxxxxxxxxxxx> wrote:
> 
> >> I normally open the progress window when I start up the computer as
> >> there are likely to be a large number of messages to process.
> 
> >> I accept that if I were fetching from several mailboxes it would still
be
> >> difficult to see the final stats for any except the last.
> 
> > Frank's Progress window might be different from the one I was using when
I
> > first invented the Summary window, but when I added the Summary window,
the
> > whole point of that was that it would display summary info for all boxes
> > processed in the most recent scan (using nudge arrows to move from one
box's
> > summary to another).  Does Frank's version not offer the same facility?
> 
> The Progress window is different from the Summary window. It displays 
> messages as they are progressed, and accumulates counts for each 
> result (accepted, diverted, etc). The problem is that all the values 
> are cleared immediately after the last message has been read,

No doubt because they need to be for the next box?

> so it's almost impossible to read the totals.

That's why I invented the Summary window (and because as I was usually
scanning at least 3 or 4 boxes I wanted to see the detail counts for all of
them not just the most recently scanned one).  But as I say, Frank's version
has doubtless moved on a bit.

It's also possible that the generic code in place for handling scans of
multiple mailboxes isn't making too much sense if you're only scanning one
box.

> The window itself remains on the screen for several more seconds while the
> messages are being processed. All I am asking is for the icons to be
> cleared before starting a mailbox and not at the end.

I'm pretty sure that my version cleared stuff at the start of each mailbox,
but the action that took place in my code (and Dave Higton's code which it
was based on) at the end of a box meant that logic for starting a next box
would occur (and clear the fields) even if there wasn't actually a next box
to scan (or various other conditions applied to stop things).  It's possible
that you're seeing start-of-box (n+1) processing when you think the code
should be at end-of-box(n).   

-- 
Jeremy Nicoll - my opinions are my own


Other related posts: