Problem with magic citations

---

Love the concept of manuscripts. Have been playing around with it a bit, but I notice something that goes wrong when using magic citations from papers. This works well when referencing articles, but when you reference a book, the information about the publisher and the town of publication is 'lost in translation' - it doesn't make it into manuscripts. In the bibliogrpahy, you therefore get books for which the only info that is given are the author and the title of the book. Needless to say, this is not acceptable for most referencing styles (if not all). Is this a bug, or am i doing something wrong?

@Ritske-Rensma this is actually something I think we've now worked out for the next update. The issue is that some citation styles are quite picky in what fields they include for different reference types and what they do not. We have improved some heuristics in the app for picking the appropriate reference type in the upcoming update. If you'd like us to validate this before the next update goes out, please send a document demonstrating this issue to support@manuscriptsapp.com. Thanks!

Thanks for the quick reply! Here is a file I created to demonstrate this behaviour:

https://www.dropbox.com/sh/jlw1hvvi055d4s3/AACLV7ZI4c4qEpaB39wQ4NVYa?dl=0

Manuscripts seems to think in articles only - it doesn't think in books, as there is no way to even add the information for the publisher later. Once a reference is imported, you can't even choose what 'type' of reference it is (journal article, book, online source, etc), which is quite important - all reference managers allow you to assign this metadata to a reference (endnote for example has a really long list with different types, even artworks, etc). This seems like a big omission - not everything people are going to want to cite are going to be articles.

@Ritske-Rensma thank you! There is a type field used internally, issue is really that it's not really that accurately denoted often coming in from reference management tools (hence we make inferences)… indeed allowing it to be edited in Manuscripts is probably a good idea!

Hi - keen to see this too. i would expect the bibliography to reflect the exact same information/format as if you were to right-click on a paper in Papers 3 and select 'copy as reference' (for where the same citation style has been selected in Papers and Manuscripts).

when is next version likely to be available?

many thanks,
stuart

@mallow the citation style choice in the two apps is independent (the citation style is set per manuscript in the manuscript style settings). We are getting a lot of requests for also setting the citation style from Magic Citations – I'll see if I can do something about that (though Manuscripts and Papers are products of two separate companies).

I can't promise firmly a release date but I would hope we make one available by Tuesday.

I'd like to add to this discussion that the inability to 'update' inserted citations based on changing the data in papers is a serious limitation of manuscripts in my view.

To simply say they are two different products is a cop-out when considering how endnote and papers integrate with MS Word. The whole point of having a reference management tool that integrates with a writing tool is that one only needs to manage the data in one place. For example I have a paper that I inserted where some of the authors names were all capitalised when first entered into papers. Having to correct in papers and then re-insert is highly inefficient. The other example I've used previously is the situation of online-early papers being updated later with volume and page number info. Keen to here how you might address this?

Many thanks...

Being able to edit the references once they are in manuscripts is 100% necessary, I think. It's elegant that once the reference is imported, you deal with it internally - you fix errors which sometimes you only discover later, etc. Also, wouldn't it make sense that you should be able to add a reference manually, ie the old-fashioned way, namely by just typing in all the data by hand rather than importing it as a .ris file or from papers? Lastly, being able to set the reference type yourself once it's in manuscripts seems sensible too. Importing reference data is never going to be a spotless process - it isn't like that in papers, either. I always end up having to tidy up a reference once it's imported.

@mz2 Hi - thanks for your response. Yes i realise that the style choice in both apps is independent (and that in Papers it is set globally) but both products surely draw from the same source CSL files? As such I would expect that the output format for each chosen style would be the same in each app.

@mallow by and large, yes. Both apps use CSL to describe citation styles. If you choose the same style in both apps, the formatting should be similar indeed. We are interested to hear of any deviations to this.

Notably Manuscripts uses a different CSL based citation formatting engine than Papers though. Both engines have upsides and downsides (i.e. there may be corner cases where either deviates from the standard) – the one used in Manuscripts is basically the reference implementation that the CSL specification is based on though.

Hi - in Papers i get the exact formatting that i require for my institution by using the 'Harvard Manchester Business School' style. In Manuscripts the content and formatting are different (see below - non italicised publication name, no edition or publisher details and some random page number added):

Papers: Cohen, L., Manion, L. and Morrison, K. (2013). Research Methods in Education. 7 ed. Routledge.
Manuscripts: Cohen, L., Manion, L. and Morrison, K. (2013). Research Methods in Education. p.718

@mallow could you send us the document in question, please? The omission of a page number is actually an issue with Papers, but the rest of the differences are down to that reference not being detected in its correct reference type. Sending us the example document helps us a lot of in confirming this issue gone in an update where we adjust the reference type based formatting.

I have a physical copy of this book. In Papers i just searched for it and added it to my library. Papers has the details correct in terms of it being a book etc.

@mallow if Papers includes a page number in the reference, Manuscripts will present it for certain reference types. Manuscripts has not invented the page number itself. Question is whether showing it is according to the reference style. Sending us the example document helps us confirm the situation and to correct the publication type handling.

happy to send you whatever you need - most of my references though are for physical books that i own. Not documents that i can send. If there is a way to send you the publication details from papers then let me know.

i can see now where the page number is generated from. As you say, it's maybe a flaw in Papers that it is missing from their reference output.

Out of curiosity...why would you not use the same citation formatting engine as Papers? i know the 2 companies are totally separate but there are links between the 2 and the interaction with papers citations is one of the key selling points of manuscripts. Odd that they are not more aligned in this respect.

@mallow ah sorry for the confusion. By "document" I meant the .manuscript project. The Manuscript project itself contains bibliography data in structured form. It is responsible both for parsing the Magic Citations posted data, and for formatting it. What you are asking for is tweaks to how it is formatted (I'm pretty sure the issue is not with the parsing stage).

Out of curiosity...why would you not use the same citation formatting engine as Papers? i know the 2 companies are totally separate but there are links between the 2 and the interaction with papers citations is one of the key selling points of manuscripts. Odd that they are not more aligned in this respect.

We are using the open source citeproc-js CSL formatting engine (https://bitbucket.org/fbennett/citeproc-js/wiki/Home), which is the canonical implementation of CSL formatting used by tools such as Zotero, Mendeley, etc. It is a great component and a good fit for Manuscripts for several technical reasons, portability being the main one (capability to localise its output is another upcoming reason – though really feature wise they are very equivalent). To be clear, the formatting differences are not due to citeproc-js or Papers, but Manuscripts itself (and the CSL implementation in Papers is not open source and available for 3rd parties like Manuscripts).