Logging from 2010-11-01 Monday Week 44

#=================================================================
LOG: 2010-11-01 15:08:33 Monday  Week 44 <nanjiang@illergard>
    kalign and kalignP are different for the sequence "test/1havA_ref1.fa"
    The difference of kalign and kalignP is caused by the _ps_ procedure. 
	in the function  backward_hirsch_ps_dyn_new2()
    The last part of the code:
	the code
				s[j].gb = MAX(s[j].gb-gpeArray2[j]*wf2_i, ca-gpeArray2[j]*wf2_i);
    should be 
				s[j].gb = MAX(s[j].gb-gpeArray2[j]*wf2_i, ca-gpoArray2[j]*wf2_i);
    It is a typo of gpoArray2 to gpeArray2

		if(hm->startb){ /*the difference between the kalingP and kaign is caused by this code as of 2010-11-03 14:45:34 Wednesday Week 44*/
				/*s[j].gb = MAX(s[j].gb+prof1[28], ca+prof1[27]);*/
				/*s[j].gb = MAX(s[j].gb-gpeArray2[j]*sip, ca-gpoArray2[j]*sip);*/
				s[j].gb = MAX(s[j].gb-gpeArray2[j]*wf2_i, ca-gpoArray2[j]*wf2_i);
		}else{
				/*s[j].gb = MAX(s[j].gb,ca)+prof1[29];*/
				/*s[j].gb = MAX(s[j].gb,ca)-tgpeArray2[j]*sip;*/
				s[j].gb = MAX(s[j].gb,ca)-tgpeArray2[j]*wf2_i;
		}
    This bug is fixed as of 2010-11-03 14:50:50 Wednesday Week 44
#=================================================================
LOG: 2010-12-22 12:27:28 Wednesday  Week 51 <nanjiang@illergard>
  Using 'icc' instead of 'gcc', kalignP is 15% faster, but the result is
  slightly different, check it
  
