[sdpd] DIfferent CIF powder file formats found by Armel at IUCr

[Robin Shirley <R.Shirley... @surrey.ac.uk>]

I read with interest Armel's survey of the widely differing [yet apparently
all CIF-compatible (?)] formats that he found among deposited powder data
files at the IUCr site.

This emphasises my point that CIF may be fine as a write-format, but because
of its flexibility (desirable for its primary role in publishing), as a
read-format it presents considerable problems to diffraction-program authors.

One way in which this could be overcome (while we are waiting for
sophisticated CIF-data-conversion utilities to be written), would be for a
deliberately limited subset of CIF to be adopted as an interim solution for
use where the purpose of the CIF is re-input to powder-diffraction programs.

The simple CIF output format used by CRYS might serve for re-input of peak 
lists:

data_y2o3
_pd_block_ID
'y2o3 (1999/11/15 at 15:18:26) = y2o3 XFIT test data'
_computing_data_reduction
'CRYS v9.33k (QBasic), 8Oct99, (R. Shirley:  r.shirley... @surrey.ac.uk)'
_diffrn_radiation_wavelength  1.5406
_loop
_pd_proc_d_spacing
  4.33143017
  3.05789233
  2.64946348
etc...

Perhaps the profile-input situation might be helped if users adopted some
similar simple, CIF-compatible format as a practical, interim standard for
profile data.

P.S. Should that be _loop or loop_?  In 1996, when I added this facility to
CRYS, the CIF documentation used the former, but I notice that Armel's
examples contain the latter.  Or can it be either (speaking as a program
author who may have to produce code to read CIFs, hopefully not).

Robin Shirley
School of Human Sciences
University of Surrey
Guildford
Surrey
UK

--------------------------------------------------------------------

> Reply-to:      sdpd... @egroups.com
> Date:          Sat, 13 Nov 1999 19:35:26 +0100
> To:            sdpd... @egroups.com
> From:          Armel Le Bail <armel... @fluo.univ-lemans.fr>
> Subject:       [sdpd] Powder files .rtv at IUCr

> Powder data at the IUCr Journals Web site are available as .rtv
> files (not CIF). 
> 
> Below are examples of deposited data. I have found a Babel Tower
> there too.
> 
> Some (very few) are CIF files as built by GSAS :

(then follow examples of 11 different actual formats found by Armel) 

> So, if all that above is CIF-compatible (??), I understand
> that programmers of converting programs have difficulties.
> In fact, I did not find two identical format in IUCr
> powder deposited data...
> 
> So that it seems that PowBase is now the biggest available
> database of powder patterns (90) that may all be read and
> displayed by the same software (WinPlotr).
> 
> Best
> 
> Armel
> 
> PS - Of course, I ask authors permission to take their
> data at the IUCr site.