NXguide and NXmirror are almost identical

[stuartcampbell]

Original reporter: prjemian

This relates to tickets #41 and #45. These two classes are almost identical.

Posted this message to the full NIAC:

> NXguide and NXmirror look a lot alike.  Almost identical.  I'm editing
> NXmirror now based on feedback from the DESY workshop last month.
> 
> There are two TRAC tickets that apply as well:
> http://trac.nexusformat.org/definitions/ticket/41  NXguide
> http://trac.nexusformat.org/definitions/ticket/45  NXmirror
> 
>   From my limited point of view, NXguide is redundant. If you agree, we
> should consider deprecating or removing it.  However if you object to
> that proposition, please take on the job of providing much better
> documentation, especially the initial documentation string that now
> reads "Template of NXguide".
> 
> Now that parts of the manual are generated from the NXDL files, we
> should add documentation into the NXDL files appropriate to each defined
> item.  NXguide is a case in point.

[stuartcampbell]

Original reporter: prjemian

On 11/18/2010 4:00 AM, stephen.king@stfc.ac.uk wrote:

> I think these two originated from consideration of different types of
> instrument (in which I may have had a hand...), but I take the point
> that the physics is the same.
> 
> So NXguide was to describe a neutron guide (2 or 4 coated walls), or
> supermirror bender (multiple coated vanes), for example.  And NXmirror
> probably came more from the X-ray side though of course instruments like
> KWS3 at Munich use a focusing mirror.
> 
> On LOQ we have a device called a frame overlap mirror but in
> construction terms it's very similar to a bender (ie, multiple coated
> vanes), just smaller with its vanes oriented horizontally rather than
> vertically.
> 
> There is also the issue of the polarizing supermirrors used in neutron
> reflection.  Are these covered by NXpolariser (I assume such exists) or
> NXmirror?
> 
> In answer to your question I personally have no great leanings in either
> direction, just so long as the things above are captured and can be
> described!

[stuartcampbell]

Original reporter: prjemian

> I agree that NXguide is in bad shape. But basically a guide consists of 
> several mirrors building a shape through
> which neutrons can be guided. The simplest such form is just box shaped. 
> But these days people are building
> elliptical guides too. The individual parts of a guide usually have the 
> same characteristics but there are cases
> where they are different. I now start wondering if we should make 
> NXguide a collection of NXmirrors each having
> their own NXgeometries describing where they are. But I disagree that 
> guide are superfluous.
> 
> ```
> ```

[stuartcampbell]

Original reporter: prjemian

(In [764]) fixes #41,#152: note NXdata spec in NXGuide -- looks like a good example to follow for similar cases