User talk:Attipaci/sandbox

Need input on fixing submission

@WikiDan61:)

Thanks for reviewing the submission. I'm not sure I understand what you mean by "None of the sources in this draft represent independent coverage of the topic."

The subject is a software library that implements reading and writing FITS files. Accordingly, the references mainly address the FITS standard that the software implements, as well as specific documentation relating to the implementation itself. The file format is the primary 'coverage of the topic', since that is what the library allows to access. That format specification is entirely independent of the library.

Lots of astronomy software libraries have Wikipedia pages, such as SOFA, NOVAS. I do not see them provide references in a fundamentally different way (If anything the mentioned ones contain references that are much less 'independent'). Yet they are legit articles.

nom.tam.fits, being widely used and venerable, should also have a Wikipedia entry. I hope you can provide me with some suggestions on how this page can be realized.

Thanks in advance for your input.

-- Attila Attipaci (talk) 21:13, 13 May 2024 (UTC)

Attila, your sources need to show that this software library is notable; i.e. that others have taken note of it and written about it. WikiDan61^ChatMe!_ReadMe!! 21:23, 13 May 2024 (UTC)

Thanks for clarifying. I added some references for software that rely on the nom.tam.fits library. Hopefully that takes care of it.

Still, I'm a bit surprised, because neither the SOFA and NOVAS wikipedia pages contain any reference (not a signle one!) that demonstrates that they are notable ("i.e. that others have taken note of it and written about it"), yet they exist as Wikipedia pages...

cheers Attipaci (talk) 23:24, 13 May 2024 (UTC)

Regarding Standards of Fundamental Astronomy and Naval Observatory Vector Astrometry Subroutines, or any other page that has no references showing notability, there are two remedies: improve them (i.e. add references that show that these libraries are in wide use, such as journal articles that have been written based on results of these libraries, etc.), or nominate them for deletion. But the existence of other poorly referenced articles is not a valid argument for the creation or retention of this article. WikiDan61^ChatMe!_ReadMe!! 11:29, 14 May 2024 (UTC)

Hi again @WikiDan61

I decided that adding second (or third) references, or additional statements, to the dependent application individually to indicate the source of how their dependence on nom.tam.fits can be verified was making a mess of things, and made things more confusing rather than less so. So, instead I opted up front to clarify how dependent , applications can be identified / verified in general -- which applies to all the software that is listed.

When you initially requested to demonstrate 'notability', I have spent many hours this way to identify dependents (there are many), and pick a selection of those that I found as interesting examples, either because they seemed widely used (e.g. based on their GitHub stars), or because the were well documented (esp. if they had publications).

I hope we can move on now.

cheers,

-- Attila. Attipaci (talk) 05:44, 15 May 2024 (UTC)

Regarding links for other software that relies on nom.tam.fits, you have provided links to the github repositories for the software packages, but no verification that they rely on nom.tam.fits. WikiDan61^ChatMe!_ReadMe!! 11:34, 14 May 2024 (UTC)

@WikiDan61,

I'm not sure what verification you need. Most of the software listed appear as 'Dependent' projects at: https://github.com/nom-tam-fits/nom-tam-fits/network/dependents (meaning they themselves declare nom.tam.fits as a dependence in their build systems).

A few come from google searches for "uses nom.tam.fits", and they are software that in their documentation specifically mention that they use nom.tam.fits.

Finally, there are two pieces of software (IRC and CDH), where I worked closely with the developer, and know with 100% certainty that they were using nom.tam.fits (there really was no other Java library for FITS files in those days). It is a pity that these software do not have publications to reference. I can easily obtain confirmation (e.g. by email) from their developer on their use of nom.tam.fits. Otherwise, I'd put this into the category of 'private communications' as the source. Alternatively I can remove them from the list, which w=ould be a pity, because they really were some of the more impactful adopters...

I hope this offers enough clarification...

-- Attila Attipaci (talk) 18:25, 14 May 2024 (UTC)

I forgot to mention, that another 4 come from the nom.tam.fits page from Maven Central: https://mvnrepository.com/artifact/gov.nasa.gsfc.heasarc/nom-tam-fits. These are software on Maven Central that declare nom.tam.fits as a dependence in their POM. You can click on the number under "Usages" which takes you to the dependent software on Maven Central. I then proceeded to look up an appropriate reference (a publication if there was one, or else a web site, such as their github repo). Attipaci (talk) 18:36, 14 May 2024 (UTC)

Actually, for IRC, the source code is available through the reference web page provided, which provides proof... I've added a couple of sentences describing how the dependent software were identified so that anyone can repeat the process and verify independently that these are in fact users of the library. I hope this helps.

-- Attila. Attipaci (talk) 19:03, 14 May 2024 (UTC)

The dependence of other software on this library is not really what is needed to establish notability. We need evidence that the library has been written about in independent literature. I'm not going to decline the submission again, but neither will I accept it. I'll leave a comment for other reviewers with my opinion and allow them to make their own assessments. WikiDan61^ChatMe!_ReadMe!! 11:53, 15 May 2024 (UTC)

Hi @WikiDan61,

I accept your decision. However, I also disagree with it.

What you ask for is an impossibly high bar for a software library. People simply do not publish about other libraries in 'literature'. At most they mention it (and mentions of nom.tam.fits abound in journal articles, conference presentation, and software documentation -- including in the references I provided). Open-source software is also a form of publication that is equivalent to 'literature' in and by itself. Referencing a library in source code is equivalent to writing about in literature.

If you applied your standard to all software libraries on Wikipedia, you'd have to remove all but perhaps a very few of them. Typically people do not write about other software libraries in literature independently.

Even Astropy, which is probably the most widely used astronomy library, by pretty much every astronomer all around the world. There is nothing on their wiki page that indicates that "[it] has been written about in independent literature". In fact, I don't think you will find much to that effect in the classical sense of scientific literature.

I feel like nom.tam.fits is held to a much higher standard than pretty much any library that has a Wikipedia page. That is unfortunate.

cheers,

-- Attila. Attipaci (talk) 12:19, 15 May 2024 (UTC)

This is why I've left it to other editors to evaluate. They may come to the same conclusions that you have. WikiDan61^ChatMe!_ReadMe!! 13:16, 15 May 2024 (UTC)

Fair enough.

And I do appreciate the time and effort you put into evaluating the article, and in improving the citations (I assume it was you -- Thank you if it was!).

As an additional remark on the topic of notability, I have been browsing around Wikipedia since you passed on the review. In about 2 dozen articles I checked on the most widely used astronomy software and/or libraries, not a single one meets your requirement of notability. Most don't even show independent references, and only a few list referenced examples of use at all.

Even pages of very significant large projects like Dataverse or International_Virtual_Observatory_Alliance or the Starlink_Project do not provide any references to establish their notability in an independent way.

So my conclusion is either the vast majority of Wikipedia pages fail to meet that standard, or else it has been seldom applied in the way you described... Attipaci (talk) 13:40, 15 May 2024 (UTC)

I have now added a second reference to each dependent application, linking either the GitHub dependents page or the Maven Central repo as the source for verifying their dependence on nom.tam.fits. For all others, the dependence is clearly stated in the source that was already referenced. In case of IRC, I have added comments that the available source code can be used to verify dependence (it is there for all to see). And finally, for CDH I have stated that it is based on private communication with the developer.

I hope this answers your concerns about verification sufficiently. Attipaci (talk) 19:33, 14 May 2024 (UTC)