I really need a way to edit references in the bibliography ASAP. My document is essentially done yet all my references (Vancouver - numbered) have an unnecessary URL as well as labeling all the references [intranet] (also gratuitous). References were inserted with Papers 3. Please fix this as soon as you can!!
I agree this is a major problem. I would like the ability to revise either the citation style or the original reference in papers, and then to re-sync, or update the bibliography and references to incorporate those changes. There should be no need to manually edit references within manuscripts itself.
@nrb sync or updating from reference manager end is not very easy with Papers because Papers has not really been built for that (it does not really send the uniquely identifiable information that would allow us to send back metadata changes from Manuscripts). If you'd like to see that kind of a feature, you should definitely request for this from the Papers support because Papers needs adaptation for the purpose.
The reference manager support in Manuscripts is in no way intended to be Papers only of course (indeed it isn't even now), and we may be able to implement this feature independently of it.
I should also clarify, the refinement of the citation style with settings as overrides on top of the citation styles provided in the app, and certain other adjustments that further improve the detection of the reference type (which affects the presence or absence of the citation style in the formatted reference) are coming up sooner.
Thanks for the reply. To clarify, I would hope that Manuscripts can work with Papers in the exact same way that Papers, Endnote, and other bibliographic software programs have worked with programs like word for many years. That is, citation markers are entered into the document that tells manuscripts to copy across the relevant metadata into the text and into a bibliography, with conditional formatting based on journal style etc. Most importantly, the use of a unique citation key allows an easy way to reimport the metadata from the bibliographic software in the future, should that metadata change (e.g. page numbers, author initials etc). This is the whole point of a database of papers being linked by a primary key to citations in a document. You can freely edit the database and update the dependencies (in this case citations in papers).
Is this not how you envisage manuscripts working with papers? If not, what is superior about your proposed workflow?
@nrb inserting citations with the bibliography metadata inserted into the document, as well as a formatted reference, is course precisely what you can do right now: you can update the record simply by replacing the citation at a certain spot. Manuscripts does duplicate detection and merges the bibliography data so that all references matching the one you replaced get updated. More broadly, our functionality goes quite a lot further than done before in the sense that we have a full bibliography database inside the app and Manuscripts (not the external citation tool) handles the bibliography formatting.
This is important for many reasons:
- We use the internal reference database to create different representations of the bibliography data: e.g. biblatex representation of the bibliography data for LaTeX based exports (or indeed we allow also formatting references using CSL into a LaTeX document – a first as far as we know).
- Not everyone working on the manuscript needs to use the same reference manager: you can import reference files and collaborate amongst a group of users with a choice of reference management tools.
- The document is a clean, self contained record: it does not need an external tool to format itself, and you cannot end up in a state where the "field codes" containing the references are lost from the document (which can mean that you get left with a mixture of formatted and un-formatted references).
What I was referring to is bulk updating to and from specifically Papers: it is not necessarily exact science as Papers does not include the necessary information in the data it sends over: a stable, globally unique identifier for the record in the Papers database. For instance the citation key it posts can in some instances change (when the reference becomes semantically different to what it was before, after an edit). I guess I've gone a bit into an implementation detail here: it certainly can be approached in ways that do not require this absolutely certain global identification but the feature certainly becomes better for all parties if Papers were adapted to include the identifier. So, if you would like to see this kind of functionality, it's best to request for the global identifier for the record to be posted by Papers (Citations) and that is done best via firstname.lastname@example.org.
Thanks for the clarification. I appreciate the strengths of the approach being used by by manuscripts, especially in terms of multiple collaborators sharing references etc. This is however a very different workflow to that most people are familiar with (e.g. using Papers or Endnote with Word), in which any edits to the metadata are done within the bibliographic database before re-syncing rather than re-inserting a reference. In the former case it would only occasionally be necessary to reinsert a reference due to the sorts of changes you mention (such as a change in the citation key).
Given the familiarity many people have with existing software and workflows I think it's worth your while clarifying somewhere on the website why this alternative behaviour has been adopted, how it differs, and why it is superior overall (noting it has some weaknesses, or at least requires some change in thinking). Maybe such info is already available I'm not sure.
I wanted to ask something related to collaboration on manuscripts. I guess for most authors it will for the foreseeable future remain required to export a document made with Manuscripts to a word file and then (after co-authors edit it) to re-import this file into Manuscripts. To this end, it will be essential that references will be stable (i.e., maintain the associated metadata) and editable after export and re-import - is this already the case and, if not, when will this feature likely become available? Similarly, will markup created in Word be imported once Manuscript's tracking features become available?
Thanks for taking the time to explain these issues.
@hubob reference metadata is not yet embedded in the Word documents (it is already exported in BibLaTeX form when you complete LaTeX exports).
We are presently reviewing options for how we will tackle this: there is not one standard way of doing this (different versions of EndNote use different encodings, and Papers, Zotero and Mendeley all have their different ways of doign this). We would like to support the maximum amount of interoperability as we can and in this sense the Zotero form of embedding the metadata (which is very close to what Manuscripts also does internally) is the likely route we take.
We believe though the most important improvement we can introduce in this regard is taking out the need to import and export as a means of collaboration, or indeed for all collaborators to install an application at all.
@ mz2 Thanks for the explanation! Your last statement sounds very appealing indeed - are there any ideas already how you would envision this to be possible? Some web-based solution? Document formats that contain their own reader (and annotation function)?