Statistics Builder Bugs

[Manfred]

When using the Statisics builder I do the following:

Highlight a group, click the drop down menu item.
The builder immidiately starts sucking in all the data from all the groups even yhough I have only highlighted one group. This is great waste of time.

The data should be brought in only as each group I'm interested in is checked off in the builder itself.

Next I do a complete profile over the entire message range fro the group I have selected. There is no new member data and no lurker data. Why not?

If I use a date range I get data, There should be no difference in the calculation just by changing ranges.

Manfred

[Wilson Logan]

>The data should be brought in only as each group I'm interested in is checked off in the builder itself.

I agree.

>Next I do a complete profile over the entire message range fro the group I have selected. There is no new member data and no lurker data. Why not?

>>>> Do you have the member list downloaded? This feature depends on an up to date member list.

Cheers,

Wilson.

[Manfred]

Using  the date range produces lurker idata. Using the message range (1-last) produces no lurker idata.
Same member list, loaded on the 18th.

Refreshed the member list just 10 minutes ago. Ran the date range (1-last) and still no lurker data.

[Manfred]

oops, that last part should have been 'message range'

[Wilson Logan]

Hi Manfred,

I think lurker data can only be time dependant. If it were to be implemented via message number we would be looking still at the date the message was posted to determine lurker status so lurkers by message range has no meaning.

You see what I mean?

Cheers,

Wilson.

[Manfred]

No, I don't follow your logic. A lurker (to me) is a group member that has authored no messages.

The range variable has no effect on the count other than to establish the boundries.

So, for message X to message X+300 any member who has no associated messages is a lurker.

How is that any different than counting lurkers from date X to date X+300 ?

[Wilson Logan]

>How is that any different than counting lurkers from date X to date X+300 ?

My point is that lurking is a time dependant phenomenon, but I take your point also. I shall add it to the development queue.

Cheers,

Wilson.

[Manfred]

Thanks, now that's great service !

By the way this topic/discussion also applies to the new member data also. It's also missing with message number boundries.

[Wilson Logan]

Sure, I'll look at that too.

Cheers,

Wilson.