McStas installation and questions

[Kristian Nielsen]

Hi Emmanuel,

> Hello Kristian,
> 
> At last, I could compile McStas on mica (SGI machine). parsing instrument (mcstas) and compiling is ok. running simulations by hand seems ok (instrument asks for config parameters).

Good! Did you find out what the problem was (so that I could improve the
installation or clarify in the manual)? Did you give up on the HPs?

> # mica~farhi 104% perl gscan.pl 21 ./linup-1 sim/linup_1_50.sim -39,-35 -73 0 1e5
> Invalid parameter specification '-39,-35' at gscan.pl line 35.

Ouch! This is using the old syntax from a very old version of McStas,
before the first release. Where did you find that example, I will erase
it immediately! Hm I guess you found it on the web page, parts of that
badly needs an update.

The correct new syntax would be something like

    perl gscan 21 1e5 ./linup-1 sim/linup_1_50.sim PHM=-39,-35 TTM=-73 C1=0

> Then I try something else :
> 
> # gscan 1 21 ./linup-1 sim/linup_1_50.sim -39,-35 -73 0 1e5
> Warning: translation table syntax error: Unknown keysym name: ClearLine
> Warning: ... found while parsing '<Key>ClearLine:      delete-to-end-of-line()'
> Warning: Cannot convert string "-*-screen-medium-r-normal--15-*-*-*-*-*-iso8859-1" to type FontStruct
> dgl error (protocol): remote machine not DGL capable - macfarhi.ill.fr:0.0
> dgl error (default init): default dglopen(macfarhi.ill.fr:0.0,4) returned -13

This is just plain wierd ... I think DGL is SGIs own extension to the X
Window System (for 3D graphics I guess), I cannot think of any way in
which the gscan program would cause this. Maybe you hit a bad key on the
keyboard ("ClearLine" ...) by mistake :-)

> OK, it's ok. I'm now running tas1-examples batch ... suspens !

Great!

> The McStas computation on mica is going on, slowly. ok.

Good, do not hesitate to ask again if you need help, or want to tell
about your results.

> Arnaud heiss is coming to Risoe, on TAS6 with Des Mc Morrow. I've

Thanks for the tip; Des has promised to introduce me to Arnaud.

> talked about TAS1 with him. It could be great to compare McStas
> results with our 'home' program ResTrax (which is perhaps faster, by
> really more specific for TAS instruments, and doesn't enable to
> describe anything as McStas). For this, we need a 'mecanical'
> description of TAS1 instrument (distances, components, etc) Of course
> it's in the lineup-1 file for instance, but a drawing is perhaps
> better (with the exact significance of parameter used). I guess you
> had to do so when writting lineup files for TAS1...

Yes, I am sure we can find something suitable. I actually started
something similar, but in the opposite direction, for the RESCAL program
with Alan Tennant before Christmas. It is a McStas simulation for a
RESCAL-type generic instrument that simulates an instrument defined by
RESCAL instrument definition parameters.

 - Kristian.