From: Nick Holford n.holford@auckland.ac.nz
Subject: [NMusers] Suggestions for the Official Bug List
Date: Mon, April 19, 2004 4:41 pm  

Bill,

Given that at least two users have been bitten by the GENCOM bug in the last
2 weeks (99apr082004.html and 98apr192004.html)
can you suggest why the Official Bug List (http://www.globomax.com/products/nonmem.html)
for NM-TRAN does not include the GENCOM fixes?

I see that the bug PDFs have different dates since they were first released
(03/01/04). I assume some changes have been made. It would be helpful if changes in
the Bug List were individually dated to make it easier to identify modifications. 

I tried copying the PDF text from two versions and doing a 'diff' on them but
because of the inconsistent formatting in the PDF of identical text this was not
very helpful. Perhaps you could consider offering a plain text version, or even a
'diff' output, with each release of the Bug List?

Would you also consider sending a note to nmusers each time any change is made to
the Offical Bug Lists? Please accept my apologies if you have been doing this and I
have overlooked it.

For completeness it should be noted that there is an additional fix suggested for
GENCOM which allows the user to set LNP4 to a value greater than 1000. Tom Ludden
sent me a copy of an email from Stuart Beal when I was bitten by this in 2002:

> From: S.Beal [mailto:stuart@c255.ucsf.edu]
> Sent: Wednesday, July 10, 2002 1:55 PM
> To: luddent@globomax.com
> Subject:
>
> Tom:
[stuff deleted]
[The > > material was originally written by Tom Ludden]
> > 2.  Try the following code changes to NMTRAN subroutine GENCOM.for after
> > copying the original file to GENCOM.save.
> >
> > After the line
> >       CHARACTER*12 WORK
> > add the line
> >       CHARACTER*13 WORKX
> >
> > Edit the line
> >        WRITE (WORK,'(A,I4.4,A)') 'BBBBBB(',I,')'
> > to be
> >        WRITE (WORKX,'(A,I5.5,A)') 'BBBBBB(',I,')'
> >
> > In the IF THEN ELSE block that follows
> > edit the lines
> >        LINE(LL+1:)=WORK
> >        LL=LL+12
> >        ELSE
> >        LINE(LL+1:)=','//WORK
> >          LL=LL+13
> > to be
> >        LINE(LL+1:)=WORKX
> >          LL=LL+13
> >        ELSE
> >        LINE(LL+1:)=','//WORKX
> >        LL=LL+14
>
> This fix is fine.   Alison was also impressed with you ability to create
> the fix.  We shall also fix Ver. VI accordingly.
> Obviously, Nick is pushing the boundaries of NONMEM, but should this
> problem arise again with Ver. V, note that the fix is not complete.
> In GENCOM one must also
>
> Locate:
>        WRITE(LINE,'(A,I4.4,A)') 'DIMENSION COM(',COMMAX,')'
>        LL=19
>
> Change to:
>
>        WRITE(LINE,'(A,I6.6,A)') 'DIMENSION COM(',COMMAX,')'
>        LL=21
>
> ( Explanation: When "$ABBR COMRES=n" (n>0) or verbatim code is present,
> FSUBS includes the following lines:
>
>       DIMENSION COM(1000)
>       EQUIVALENCE(BBBBB0(1),COM(1))
>
> When LNP4>1000, there will be a compile error with the DIMENSION statement.)
>


--
Nick Holford, Dept Pharmacology & Clinical Pharmacology
University of Auckland, 85 Park Rd, Private Bag 92019, Auckland, New Zealand
email:n.holford@auckland.ac.nz tel:+64(9)373-7599x86730 fax:373-7556
http://www.health.auckland.ac.nz/pharmacology/staff/nholford/

_______________________________________________________