Discussion List Archives

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Getting started

  • To: Multiple recipients of list <pddmg@iucr.org>
  • Subject: Re: Getting started
  • From: "Brian H. Toby" <brian.toby@nist.gov>
  • Date: Mon, 22 Nov 1999 16:19:38 GMT
I am forwarding a message from Lachlan that did not go through when he
posted it, with a few comments from me integrated. 

I think that it is now fixed so that all pdDMG members can now post
directly to the list. It would be great if everyone in the pdDMG would
either post a short message to pddmg@iucr.org to test this, or at least
send me a message at brian.toby@nist.gov so that I know that e-mail is
reaching you.

Brian

L. Cranswick wrote:
> >         http://www.iucr.org/~commit/itg/draft/23.ps  (PostScript)
> >         http://www.iucr.org/~commit/itg/draft/23.pdf (PDF)
> 
> Main comment on the paper - for 4.2 - is it correct CIF usage
> to have the standard uncertainties in brackets after the counts?
> 
> Shouldn't there be another Standard_Uncertainty_keyword
> defined in the loop?

There has been talk of abolishing the use of crystallographic S.U.
notation in CIF and perhaps this is now gone from the macromolecular
DDL, but the use is correct in pdCIF (DDL1). [For those who want to know
details, the flag "_type_conditions esd" that shows up in the ASCII
version of the dictionary for data_pd_meas_intensity_ signifies that use
of S.U. values is allowed. N.B. it is not present for
data_pd_meas_counts_ where they are not allowed.] (BHT)

> The main point I would like to bring up on the list
> is how to get CIF out being used for archiving data
> and data/results transfer between programs.
> 
> Robin Shirley on the SDPD mailing list made the point
> that it would be near impossible for him to be able to
> import CIF at the moment.  And private communications
> from some program authors pretty much back that up.
> Programming a robust CIF interpreter in is too much
> effort for too little perceived gain.
> 
> Is there some way to get the IUCr to support/fund a
> programmer to create some Fortran and C libraries that
> would make it easier to software authors to link their
> programs into CIF? As we live in the age of ultra
> scarce free time, it is unreasonable to expect someone
> to volunteer their time to implement.  Is this something
> the Commission on Powder Diffraction could be asked to
> take up with the CIF commission?    In theory, CIF could
> enable some very nifty things if it was common used for
> import/export in powder diffraction software.  Not only
> in data/structure transfer but also reviewing/proofing
> papers as is routinely done with Single Crystal CIF.
> 
> Is this something the powers that be the CIF commission
> care about (i.e., people making good use of CIF) or are
> they too high in the sky to consider such things?
> 
> Lachlan.
> 
> --
> Lachlan M. D. Cranswick
> 
> Collaborative Computational Project No 14 (CCP14)
>     for Single Crystal and Powder Diffraction
> Daresbury Laboratory, Warrington, WA4 4AD U.K
> Tel: +44-1925-603703  Fax: +44-1925-603124
> E-mail: l.cranswick@dl.ac.uk  Ext: 3703  Room C14
>                            http://www.ccp14.ac.uk

Two comments. 

1) There does exist a free FORTRAN subroutine library, CIFtbx
(http://www.iucr.org/iucr-top/cif/software/ciftbx/README.html). I do not
have much experience with it, but it looks pretty good to me. If CIFtbx
is flawed or insufficient, this does need to be addressed.

2) I agree that a pdCIF file is not always that easy to read, since
there are many ways to record data and software will need to be
flexible. In the case of an indexing program, one will need to look for
peak positions as _pd_peak_2theta_maximum or _pd_peak_2theta_centroid or
_pd_peak_d_spacing. Of course it is OK to write a program that requires
that peak positions be specified as d-spaces. Obviously an indexing
program can't run if there is no peak table present. If one uses CIFtbx,
I don't think it requires much more than a few dozen lines of code to
pull in the peak positions and convert them, if needed. (BHT)

[Send comment to list owner]
[Reply to list (subscribers only)]