- fixed a little bug with hkl: phi limits were not properly tested

- updated powder file writing to properly work with mean and stddev

doc/user/commandlog.htm

@@ -7,7 +7,8 @@
 <P>
 The commandlog is a file where all communication with clients
 having user or manager privilege is logged. This log allows to retrace each
-step of an experiment. This log is usually switched off and must be
+step of an experiment. This log is normally configured in the startup
+file or can  be
 configured by the instrument manager. There exists a special command,
 commandlog, which allows to control this log file.  
 <DL>
@@ -27,6 +28,10 @@ Log files are written to the log directory of the instrument account. There
 <DT>commandlog tail n
 <DD>prints the last n entries made into the command log. n is optional and defaults to 20. Up to 1000 lines are held in an internal buffer for this command. 
 </DL>
+It is now possible to have a script executed whenever a new log file is 
+started. In order to make this work a ServerOption with the name logstartfile 
+must exist in the instrument configuration file. The value of this option 
+ must be the full path name of the file to execute.   
 </P>
 </BODY>
 </HTML>

doc/user/samenv.htm

@@ -207,39 +207,43 @@ current value.
 <DT> name val 
 <DD> will drive the device to the new value val. Please note that the same
 can be achieved by using the drive command. 
-<DT> name log on
-<DD> Switches logging on. If logging is on, at each cycle in the
-<b>emon</b> 
-the 
-current value of the environment variable will be recorded together with a 
-time stamp. Be careful about this, for each log point a bit of memory is
-allocated. At some time the memory is exhausted! <b>Log clear</b>
-frees 
-it again
 and <b>log frequency</b> (both below) 
-allows to set the logging time intervall.  
-<DT> name log off
-<DD>Switches logging off.
+</DL>
+<h3>Logging </h2>
+The values of any sample environement device can be logged. There are two 
+features:
+<ul>
+<li>Logging to a file wih a configurable time intervall between log
+file entries.
+<li>Sums are kept internally which allow the calculation of the mean
+value and the standard deviation at all times.
+</ul>
+The last system is automatically switched on after the first drive or
+run command on the environment device completed.
+This system is run through the following commands.
+<DL>
 <DT>name log clear
-<DD> Clears all recorded  time stamps and values.
-<DT>name log gettime
-<DD>This command retrieves a list of all recorded time stamps.
-<DT>name log getval
-<DD> This command retrieves all recorded values.
+<DD> Resets all sums for the calculation of the mean value and the
+standard deviation. 
 <DT>name log getmean
 <DD>Calculates the mean value and the standard deviation for all logged
 values and prints them. 
 <DT>name log frequency val
 <DD> With a parameter sets, without a parameter requests the logging intervall 
-for the log created. This parameter specifies the time intervall in seconds
+for the log file. 
+This parameter specifies the time intervall in seconds
  between log records. The default is 300 seconds. 
 <DT>name log file filename
-<DD> Starts logging of value data to the file filename. All normal logging to 
- memory  will be
- disabled. Logging will happen any 5 minutes initially. The logging frequency
+<DD> Starts logging of value data to the file filename. 
+Logging will happen any 5 minutes initially. The logging frequency
  can be changed with the name log frequency command. Each entry in the file is
  of the form date time value. The name of the file must be specified relative
  to the SICS server.
+<DT>name log flush
+<DD>DigitalUnix buffers output heavily. With this command an update of
+the file can be enforced.
+<DT>name log status
+<DD>Queries if logging to file is currently happening or not.
 <DT>name log close  
 <DD> Stops logging data to the file.
 </DL>