Hello, Jernej. 
You wrote in <mid:[EMAIL PROTECTED]>

JS> TB Mem usage: 37 MB
JS> Peak: 68 MB
JS> VM size: 35 MB

  It seems very like my values :). But I also want to note that this value
  depends also from the memory utilization by installed plugins. Plugins also
  utilize the memory and they do it in the "shadow" of TB. If a plugin uses more
  or less complex algorithm, or if it need much memory for some internal data,
  the total TB memory usage can be very high. I find it as a problem when
  writing the BayesIt antispam plugin - because it contains the statistic
  information which can occupy a lot of memory. I think this is the very reason
  why bayesian method is more intended to be used on servers, not on the
  workstations. If you try a BayesIt 0.3a with "Hybrid Graham" or "Robinson"
  method - it will occupy about 50mb - additionally to usual TB memory
  occupation. (now my TB occupies about 32mb with loaded Bayesit 04c, but if I
  unload the plugin, it will occupy only about 16-17mb). By the reason I intend
  to develop it in disk-storage direction in opposite to memory occupation. And
  this is also the reason why bayesit 04c took so much time to be released -
  because such disc procedures were needed to be written.

-- 
Sincerely,
 Alexey.
Using TB 2.0b5 on WinXP Pro SP1 (2600), spelling by ORFO2002 (CSAPI) 
..with Kaspersky Antivirus Plugin (ver 3.5 Gold) & antispam filter BayesIt! 0.4b

   mailto:[EMAIL PROTECTED]


________________________________________________________
 Current beta is 2.0b6 | "Using TBBETA" information:
http://www.silverstones.com/thebat/TBUDLInfo.html

Reply via email to