<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
Hello all of you,
<p><b>About the Gravity_guide (attached file):</b>
<br>The 'G' constant should be a protected variable, to avoid conflicts
with other macros:
<br>OUTPUT PARAMETERS (ww,hh,whalf,hhalf,lwhalf,lhhalf)
<br>...
<br>double G;
<br>...
<br>G=9.82;
<p><b>About the Bender:</b>
<br>The Bender doc header is not valid. I here attach a full version.
<p><b>About version control for comp:</b>
<br>I do not think the proposition of Per-Olof will solve the problems
of indentifying what are the component versions and origin. At the end
you would have only 'official updated' and 'unofficial' components, and
this info would not be very useful !
<br>The best way would be to have a 'version' column in the McDoc automatic
page (list of comps).
<br>Also, the column 'Help' is not very informative (except when you click
on 'More...'. perhaps this 'More...' item could be moved to 'More (version
number)'. If McStas is to be opened and shared by users, I think the real
origin of the components should be displayed clearly, even when these comps
are included in the official release. I like to know that the bender comes
from Erlangen.
<br>When you put 'official release', just add also the McStas version it
is associated to (modify headers to have this displayed).
<p><b>Comp update</b>
<br>Also, perhaps it is the time, for version 1.4.1 to come, to update
the POLARISATION PARAMETERS (sx, sy, sz) in every comp ? (I have not done
that yet)
<p><b>Version 1.4.1 to come</b>
<br>I have put in the ILL McStas page (<A HREF="http://www.ill.fr/tas/mcstas/">http://www.ill.fr/tas/mcstas/</A>) an
updated version of McStas 1.4 called mcstas-1.4a. It includes all the updates
that were discussed these two last weeks:
<ul>
<li>
new lib/mcstas-r.h and .c to remove the 2e9 limit (passed run_num counter
from int to double)</li>
<li>
signal handler for Unix/Linux (to see the simulation status with kill -USR1
<pid>, and also to stop it saving results with other signals/interupts)
in lib/mcstas-r.h and .c</li>
<li>
a new internal mcstas function 'double get_run_num()' to get the number
of iterations performed until now in a simulation.</li>
<li>
new Monitor_nD (0.13.9) with 3He gas handling, intermediate results saving
during long simulations (works with the new mcstas-r lib only). a PreMonitor_nD
to make correlations between two positions.</li>
<li>
bug corrected in Source_Optimizer (0.09), and asociated Monitor_Optimizer
(0.08) that caused a SIGFPE error (div per 0 during optimisation of the
spin divergence)</li>
<li>
new Gravity_Guide (1.2) protecting internal variables, and having G as
a protected comp symbol (not a macro)</li>
<li>
new Bender (1.1)</li>
<li>
an update of the 'cogen.c' file for correct handling of string parametrs
in instruments</li>
<li>
an update of lib/mcfrontlib.pl to plot results for simulations using instrument
string parameters</li>
</ul>
This is to wait until 1.4.1 appears !!
<br>Cheers. Emmanuel.
<br>
<p>Per-Olof Åstrand wrote:
<blockquote TYPE=CITE>Hi,
<p>Thanks for pointing this out. I have put the new version in the McDoc
source and both the
<br>Risø and ILL web-pages will be updated automatically by the
McDoc program.
<p>Since Bender.comp is available at the web-pages but is not a part of
the official McStas
<br>release, it gives me the opportunity to elaborate on this. The web-pages
at Risø and ILL
<br>generated by the McDoc program contain components from the official
release, unofficial
<br>components provided by users, and updated versions of official components.
Just by inspection
<br>on the web-page, it may be very difficult to be able to decide which
of these categories a
<br>component belongs to. For the official components developed at Risø,
the "Origin" entry
<br>always say "McStas release", even if it is an updated component different
from the version in
<br>the official release. For components kindly provided by users, the
"Origin" entry is mostly
<br>used for giving a geographic origin regardless of if it is an official
component or not.
<p>For me (and I guess also for others), some time is spent for each component
(e.g. the Bender
<br>component) to figure out where its "origin" is. I therefore suggest
that the "Origin"
<br>definition in a component can only contain three different entries:
"McStas release", "McStas
<br>release, updated version" and "Unofficial", and that the "Authors"
definition is used for
<br>specifying the geographic origin. I would like to have some response
on this, otherwise I
<br>will just implement it in a week or so.
<p>Finally, the McDoc program is a very efficient tool maintained by ILL
and Risø for sharing
<br>components and keep them up-to-date. Users are therefore encouraged
to make their components
<br>public either by sending them to us or to establish an additional web-server
handling McDoc
<br>(which is not too difficult) in addition to ILL and Risø. If
a component is of general
<br>interest, it may eventually be, with the permission of the author,
included in the official
<br>McStas release and documented in the manual.
<p>Best regards,
<p>Per-Olof Åstrand
<p>Stuart Rycroft wrote:
<p>> Hello
<br>>
<br>> I just wanted to mention a problem with the bender component, which
can be
<br>> downloaded from the Riso or ILL website. They both have an
early version
<br>> of the component (Feb. 7th 1999) but from the email archive, a later
<br>> version (Feb. 22nd 1999) was sent by the author (Philipp Bernhardt)
which
<br>> corrects a small bug in the calculation of the reflectivity for the
top
<br>> and bottom mirrors. Without this correction the output from
my simulated
<br>> guides is obviously not correct when looking at the divergence, but
it is
<br>> very hard to see a problem when just looking at an intensity since
the
<br>> only effect I have seen is a small loss, maybe around five percent.
<br>>
<br>> The updated component and the problem are given by the author in
the
<br>> mailing archive:
<br>>
<br>> <a href="http://neutron.risoe.dk/cgi-bin/wilma_hiliter/neutron-mc/9902/msg00023.html?line=24#hilite">http://neutron.risoe.dk/cgi-bin/wilma_hiliter/neutron-mc/9902/msg00023.html?line=24#hilite</a>
<br>>
<br>> Maybe the component list on the Mcstas webpages should have this
version
<br>> instead of the old one.
<br>>
<br>> Regards,
<br>> Stuart Rycroft
<br>> IRI, Delft
<br>>
<br>></blockquote>
<pre>--
What's up Doc ?
--------------------------------------------
Emmanuel FARHI, <A HREF="http://www.ill.fr/tas/people/Farhi.html">http://www.ill.fr/tas/people/Farhi.html</A> \|/ ____ \|/
CS-Group ILL4/156, Institut Laue-Langevin (ILL) Grenoble ~@-/ oO \-@~
6 rue J. Horowitz, BP 156, 38042 Grenoble Cedex 9,France /_( \__/ )_\
Work :Tel (33/0) 4 76 20 71 35. Fax (33/0) 4 76 48 39 06 \__U_/</pre>
</html>