From: Armel Le Bail
Newsgroups: sci.techniques.xtallography
Subject: Bug in Fullprof 3.3 (97/11/14)
Date: Sat, 20 Dec 1997 12:06:31 +0100

Hi Fullprof users,

Care with the last fp90.exe dated 97/11/14. Apparently you will have more difficulty to extract any "|Fobs|" with this late 3.3 version. If you have to perform an ab initio structure determination, prefer the previous version (either fpold or the 3.3 version downloaded before 97/11/14). The test is simple, download :
http://www.cristal.org/iniref/data/naoxa0.zip
run the latest version, it diverges immediately,
run the previous version, it converges.

The bug (or call it a change if you prefer) has for consequence unstability in cases which were stable previously. Nevertheless, the new version works if you first allow some refinement cycles with few parameters (the zeropoint only, for instance) then let IRF=2 and then add more cycles with cell and profile parameters refined. Don't reset IRF to 0 or the refinement will diverge (at least, in the case of the naoxa0.pcr and naoxa0.dat files). Hard to say what has changed without the source.

Thanks to Zhu Jianghai who had first this surprise.

All the best,

Armel Le Bail - Universite du Maine - Laboratoire des Fluorures
CNRS UPRESA 6010 - Avenue O. Messiaen - 72085 Le Mans cedex 9 - France
Web : http://www.cristal.org/

Another problem due to the new bugged FP90.EXE
 

Subject: SDPD tutorial
Date: Wed, 4 Feb 1998 18:39:10
Dear Prof. Le Bail,

I have been reading for the last days your Web-tutorial about Strategies for Structure Determination from Powder Data. It is really very kind of you to make such a tutorial available for the scientific community, which I am finding really very interesting.

I am writing to you now because I am having troubles trying the data for Cimetidine and I don't know where the problem is.

I have downloaded the FULLPROF version for a pentium machine from the saclay ftp site (fp90.exe) and when I run the cimetidine files (from section 2.2.2 Automatic Indexing in the tutorial, file cim0.zip, including cim0.dat and cim0.pcr) the refinement/extraction diverges after 28 cycles. I was expecting to obtain the results shown in cim2a.html (from section 3.2 Extracting Structure Factors), is this right?

I look forward to your reply and thank you very much in advance.

Yours sincerely,