The web scripting page refers to samples/estron.cml
The applet chokes when in tries to load this file. On IE it dies badly,
requiring a reboot.
I believe you may have said once before that this file is incorrect. Can
you please confirm?
If it is a bad file then we should get rid of it.
From: Peter Murray-Rust <pm286@ca...> - 2003-09-14 09:02:03
At 22:27 13/09/2003 +0200, Miguel wrote:
>The web scripting page refers to samples/estron.cml
>The applet chokes when in tries to load this file. On IE it dies badly,
>requiring a reboot.
>I believe you may have said once before that this file is incorrect. Can
>you please confirm?
As far as CML goes this was correct CML1 but is not correct CML2. The
coordinate3 element has been dropped from CML2, although a coordinate3
*type* is supported in STMML. This is a messy area - we tried to keep
compatibility and have kept attributes to support this concept. Thus CML2
atoms support xyz3 attributes in which 3 components can be added.
The advantage of having coordinate3 as an element is that units,
uncertainties, etc can be added. But the overhead is probably not worth it
for normal work.
I have copied in the CDK list as there is a cmltest directory with "CML"
files, many of which are not compliant (some are not even XML) and I
suggest the whole directory is cleaned out.
There are examples of exhaustively roundtripped CML1/2 array/nonarray in
openbabel and it would be worth including come of these in both
distributions. There is a variety of 2D and 3D coordinates so you would not
use all for a given program. If you can't find them I will mail the developers.
I have recently revisited crystallographic CIF files and now have CML files
which contain xFract, yFract, zFract, x3, y3, z3 and x2, y2.
>If it is a bad file then we should get rid of it.
>This sf.net email is sponsored by:ThinkGeek
>Welcome to geek heaven.
>Jmol-developers mailing list