[LON-CAPA-cvs] cvs: loncom /lonnet/perl lonnet.pm

[raeburn]

raeburn		Tue Nov 30 07:01:30 2021 EDT

  Modified files:              
    /loncom/lonnet/perl	lonnet.pm 
  Log:
  - Add a comment about stuff flushcourselogs() can end up storing (and where).
  
  
Index: loncom/lonnet/perl/lonnet.pm
diff -u loncom/lonnet/perl/lonnet.pm:1.1471 loncom/lonnet/perl/lonnet.pm:1.1472
--- loncom/lonnet/perl/lonnet.pm:1.1471	Tue Nov  9 20:14:04 2021
+++ loncom/lonnet/perl/lonnet.pm	Tue Nov 30 07:01:30 2021
@@ -1,7 +1,7 @@
 # The LearningOnline Network
 # TCP networking package
 #
-# $Id: lonnet.pm,v 1.1471 2021/11/09 20:14:04 raeburn Exp $
+# $Id: lonnet.pm,v 1.1472 2021/11/30 07:01:30 raeburn Exp $
 #
 # Copyright Michigan State University Board of Trustees
 #
@@ -4963,6 +4963,29 @@
             if (! defined($dom) || $dom eq '' || 
                 ! defined($name) || $name eq '') {
                 my $cid = $env{'request.course.id'};
+#
+# FIXME 11/29/2021
+# Typo in rev. 1.458 (2003/12/09)??
+# These should likely by $env{'course.'.$cid.'.domain'} and $env{'course.'.$cid.'.num'}
+#
+# While these ramain as  $env{'request.'.$cid.'.domain'} and $env{'request.'.$cid.'.num'}
+# $dom and $name will always be null, so the &inc() call will default to storing this data
+# in a nohist_accesscount.db file for the user rather than the course.
+#
+# That said there is a lot of noise in the data being stored.
+# So counts for prtspool/  and adm/ etc. are recorded.
+#
+# A review of which items ending '___count' are written to %accesshash should likely be 
+# made before deciding whether to set these to 'course.' instead of 'request.'
+#
+# Under the current scheme each user receives a nohist_accesscount.db file listing 
+# accesses for things which are not published resources, regardless of course, and
+# there is not a nohist_accesscount.db file in a course, which might log accesses from
+# anyone in the course for things which are not published resources.
+#
+# For an author, nohist_accesscount.db ends up having records for other items
+# mixed up with the legitimate access counts for the author's published resources.
+#
                 $dom  = $env{'request.'.$cid.'.domain'};
                 $name = $env{'request.'.$cid.'.num'};
             }