XISF Specification document - Call for contributions

[Juan Conejero]

Hi everybody,

The current draft 9.3 looks like a good candidate to be released as the final XISF 1.0 Specification document:

http://pixinsight.com/doc/docs/XISF-1.0-spec/XISF-1.0-spec.html

(you may need to refresh your browser if you've been reading previous drafts)

Before announcing the official release, I need your help for proofreading the document and detecting possible errors. Of course, if you have any suggestions, or if you think something must be changed or improved, this is the moment. This is a large and complex document, so I understand if you don't want or can't help with it. We all have our own limits of time and work, after all, so I know this is asking for too much.

If you can't review the entire thing, there are just two sections where your contributions can be easy to do and very necessary:

11.4.2 Optional Metadata Properties
11.5.3 Astronomical Image Properties

The idea with 11.5.3 is not to define a comprehensive list of astronomical image properties in a variety of specialties; there could be literally thousands of them. What we need is a dictionary of fundamental properties. More and more specialized properties can be standardized in the future through addendums to the core specification.

Any contribution to improve the XISF specification will be highly appreciated. Thank you in advance.

_________________
N.B.: If you don't like or don't care about the XISF project, this is not the thread to say that.

[Niall Saunders]

Hi Juan,

Excellent work - as usual!

I have given the whole document a 'speed-read', and will try and re-read this over the next few days.

I have to confess that I haven't kept up to speed since your first announced the XISF specification - but the whole document seems well constructed and well-organised.

Keep up the good work!

[mschuster]

Thanks Juan,

Maybe review for properties that could change? For example, my setup's focuser position changes during the exposure.

Maybe change 'average' to 'typical'?

Thanks,
Mike

[kkretzsch]

Hi Juan,
looks very good!

I checked the Observation namespace in more detail for obvious reasons and I have a minor question. There are two properties refering to equatorial coordinates : Observation:Object:RA/Dec and Observation:Center:RA/Dec.  From an image acquisition point of view I am not sure how to ensure that both data are determined reliably if they do not coincide. An image can contain many objects so which Object:name or Object: ccordinates should be set? IMHO, only considerung the Center:RA/Dec properties should be sufficient for the first version.

Klaus

[Juan Conejero]

Maybe review for properties that could change? For example, my setup's focuser position changes during the exposure.
Maybe change 'average' to 'typical'?

Or maybe adding 'measured' to descriptions were a precise value is involved would be even less restrictive. For example:

Instrument:Focuser:Position

    A Float32 property. Measured position of the focuser in millimeters, with respect to a device-dependent origin.

In this way we are not claiming that the value is 'exact', be it an average or not.

[Juan Conejero]

Hi Klaus,

Thanks!

From an image acquisition point of view I am not sure how to ensure that both data are determined reliably if they do not coincide.

The Observation:Object namespace is for informative purposes exclusively. It can be useful for metadata classification and search operations applied to data sets, but should not be used for anything else. Observation:Center is the reference position to be used where positional accuracy is required. For example, Observation:Center:RA/Dec are the appropriate initial coordinates for a plate solving routine like the ImageSolver script. So the answer to your question is, undoubtedly: always try to determine Observation:Center as accurately as possible, as the direction where the telescope is being aimed at the sky. Observation:Object is irrelevant in this regard.

These explanations should be included in the descriptions of these properties to remove any ambiguities, and to state clear the purposes of each of them. Thanks for pointing this out.

An image can contain many objects so which Object:name or Object: ccordinates should be set?

If defined, there must be a unique Object:Name child property of an Image element. See [8.4.1 Property Identifier]:

A property identifier must be unique for the object with which the property is associated.

So Observation:Object:Ra/Dec can only refer to Object:Name. There are many situations where the observed object and the central coordinates of the image can be different. For example, suppose an image where the main subject is Jupiter, but most satellites are placed at one side of the planet's disk. Or deep-sky images where the composition requires displacing the main object...

[Juan Conejero]

Hi everybody,

Last chance!!! The current draft 9.4 of the XISF specification document:

http://pixinsight.com/doc/docs/XISF-1.0-spec/XISF-1.0-spec.html

(refresh your browser) is *the* final version that I'll release officially next week, unless somebody has something to say that stops me.

The reference C++ implementation is also ready for release:

http://pixinsight.com/developer/pcl/doc/html/group__xisf__support.html

and will be publicly available tomorrow on our open-source GitHub repository. This reference implementation allows including XISF support very easily in any application, just by linking it against the PCL library. Of course, no running PixInsight application is required, since the XISF support classes are completely independent and self-contained components. There are no license issues either, since PCL is released under a BSD-like license.

So this is *really* your last opportunity to contribute to the XISF formal specification. As always, all comments, suggestions and error reports will be highly appreciated.

[Niall Saunders]

Hi Juan,

Like everything, there *must* come a time where you launch your creation into the big, wide, world amd see if it will survive.

I believe that the XISF specification has been very carefully thought out - not just by you as its creator, but by the many people who have taken the time to follow its development in detail.

The use of XISF may never replace the use of FITS for the archiving of astronomical (or general) imaging data, but it should give those who maintain the FITS standard 'food for thought'.

I say, "Publish, and be damned!"