I have this problem with the active users function :
Sometime I see me name or some users name many times as an active users,

For Example:

Romeo  09/May/2003  7:36am 09/May/2003  7:36am 0 Min. Microsoft IE 6 Windows XP
Romeo  09/May/2003  7:33am 09/May/2003  7:36am 3 Min Microsoft IE 6 Windows XP

Any one have been in the same prob. ?

As HTTP is a statless protocol it is very hard keeping track of who is online at that time, this means that active user trackers are never 100% accurate.

This user is probally using an IP masking program, or usinmg a proxy server that is changing their IP address. The forum software uses IP addresses to keep track of the user, so if this changes the user is seen as a new user and readded to the active usesr list.

I'm gonna have a go at writing a mod that checks the names and removes etra entries.

Borg is right though, its not a forum fault, its an ISP kinda thing.

I'll post up what I find out this weekend.

Cheers

Kev

waiting you PrimeraRacer ,
Cheers

I had a look at the acitive_users_inc.asp file and it seems that I placed the code to remove second username entires in the wrong part of the file. Place the following at line 107 of acitive_users_inc.asp:-

'Delete older second entries if the user has returned in under 10 minutes with a new IP
 If lngActiveUsersID <> 2 Then
  If strDatabaseType = "SQLServer" Then
   strSQL = "EXECUTE " & strDbProc & "DeleteActiveUserDoubleEntry @lngActiveUsersID = " & lngActiveUsersID & ", @strIPAddress = '" & strIPAddress & "'"
  Else
   strSQL = "DELETE FROM " & strDbTable & "ActiveUser WHERE " & strDbTable & "ActiveUser.Author_ID=" & lngActiveUsersID & " AND " & strDbTable & "ActiveUser.IP <> '" & strIPAddress & "';"
  End If
 
  'Detlete from database
  adoCon.Execute(strSQL)
 End If
 
 'Requery the database to allow access to catch up
 rsCommon.Requery

Active users whether using PHP, ASP, ASP.NET, etc. will never be 100% accurate.

The reason for this is that the protocol (HTTP) that you use to connect to a web site is a statless protocol.

This means that each time a page, file, image, etc. is called by your web browser and connection to the server is opened as soon as the page has finshed downloading the connection is dropped and the server no-longer knows if the person is just reading a page or has left the site.

Also for each image on a page a new sepprate connection is made and then dropped to download that image.

So keeping track of who is on the site is more a limitation with the HTTP protocol and how the Internet works.

So the only way to get around this is to use things like cookies and IP addresses to try and keep track of the user and then leave a 10 minute expiry time to incase the user is reading pages inbetween moving around the site. The problem then is that not everyone has cookies enabled also IP address can change, so there is no 100% way of keeping track of a user.