[Wien] QTL

[Stefaan Cottenier]

Hello Jan,

We found the problem: in SRC_qtl/readc.f there is a line with free write 
format :

        do 150 j=1,iso*n
        read(ifile,5)aline
        write(90,*)aline(2:200)   <=========
 5      format(200A)
        close(90)

In some environments, this produces a fort.90 file where the string that is 
written is split in blocks of 80 characters. Subsequent reading of fort.90 
later in the program then reads only the the first 80 characters, putting 
all remaining numbers to zero. This produced our '****' and a failed 
unitarity test. Most likely, there are linux environments who have different 
behaviour, and write the 200 characters on one line, even with free format.

If that line is changed to write(90,5) [this format line nr. 5 was already 
there, probably it's only accidentally not used], then our problems seem to 
have gone.

I suggest you verify this, and send Peter updated version?

Another small bug: in qtl.def, the scratch directory is omitted in front of 
the vector files. That should probably be corrected as well in the script 
that produces this def.

Thanks for your help,
Stefaan and Doru