The sprint notes from the PKP Hannover Sprint, hosted by the Leibniz Information Centre for Science and Technology in September 2023 are now available.
Sprints involve PKP community members joining diverse groups to work on PKP software and support. The Leibniz Information Centre for Science and Technology (TIB) hosted six working groups at the PKP Hannover Sprint in September. This is a summary of one such group’s work.
Group members
- Mariya Maistrovskaya, PKP
- Gerrit Fröhlich, ZPID – Leibniz Institute for Psychology
- Philip Munch
- Oliver Krüger, Staats- und Universitätsbibliothek Hamburg
- Radek Gomola, Munipress
- Alec Smecher, PKP
- Touhidur Rahman, PKP
- Sebastian Schmidt, SLUB Dresden
- Patricia Mangahis, PKP
- Mark Huskisson, PKP
Background
More journals on OJS are looking to publish using a non-traditional publishing model
Use Cases
Continuous Publishing
Continuous Publishing is used in place of or parallel with traditional articles. Journal editors want the option of a continuous AND traditional volume-based issue to publish journals as they are ready, mimicking other publishing models like F1000 and WordPress. This would help reduce the time it takes them to assemble a complete issue, which may take up to a year for smaller journals.
Following this publishing model, the workload is spread out evenly, rather than having a high-pressure period when wrapping up an issue.
Even if some editors know about the Forthcoming plugin, they don’t use it often, as they are unsure how metadata is handled downstream (one-time collection, i.e., GS vs. continuous.
Online First
Online First refers to the process in which accepted articles that still need to be copyedited and typeset are made available upon acceptance and are later replaced with the final version (either Online or Print). It is important to note that Online first publishing is separate from preprints.
In some cases, journals publish special issues, but not all submitted articles fit those subjects, and authors don’t want to wait several years for the new topic so that those articles can be published immediately without an issue assignment.
At the same time, journal editors usually run into articles that go through a short peer-review process and are assigned to the next issue, and they want to be able to display it once it is ready.
Dependencies – what depends on the issue-based model?:
- Preservation pipelines (e.g. PKP|PN, CLOCKSS, LOCKSS)
- PKP|PN – relies on issues, and when an article is updated, it can be repackaged
- Citation
- The citation we present changes.
- Different citations affect the citation count [scientific gamification]
- Metric-based (Altmetrics)
- How would authors go about citing different versions? How do disciplines handle issue-less articles? Bibliographic software?
- APA – use DOI instead, “e.g., Advance online publication. doi:10.1037/a0028240”.
- CiteProc-PHP (our library for CSL-based formatting of citations) supports issue-free publications.
- Crossref (CrossMark)/DOI/DataCite
- DOIs (see Crossref- Publication stage + DOIs)
- “[Each version of] a preprint should have its DOI (DOI A).
- Accepted versions (including PP, AOP, AAM, and VoR) should have a separate DOI (DOI B).
- In the case of a significant change to the published version, a notice explaining the correction/update/retraction should be published. The updated version should have a new DOI (DOI C)”.
- Crossref: “Issue elements are only required if a DOI is being deposited at the issue level.”
- From PKP staff on DOIs and Crossref/registration:
- In OJS 3.4 – Editors would need to select the option to assign DOIs at the Copyediting.
- Minting- Default format DOI does not care about issues.
- Registering/ Depositing – would depend on the Registration agency and whether they accept something that doesn’t have an issue.
- OJS – need to differentiate missing info (issue because someone forgot).
- Later, an article with issue info metadata with Crossref will need to be updated.
- DOIs are not minted for versions in OJS (just in OPS), but it is possible to do so in OJS.
- Issue publication triggers a deposit, but the Crossref plugin checks whether an article is already published and only registers published articles.
- XML validation on whether the issue is required or not should be based on whether it is intended to be there or not (editors would need to actively indicate whether an article is meant for an issue or not).
- Pending Question: What happens with the DOI + Crossref status when unpublishing an issue/removing issue metadata from an article?
- DOIs (see Crossref- Publication stage + DOIs)
- Datacite (see Datacite – Versioning)
- “When content underlying a DOI is updated, we recommend updating the DOI metadata and, for major changes, assigning a new DOI.
- For minor content changes, a new DOI is not required, as the same DOI may be used with updated metadata.
- For major content changes, we recommend assigning a new DOI and linking the new DOI to the previous DOI with related identifiers.”
- Based on the required metadata fields, Datacite does not require issue metadata is not required. Source
- Printing
- Journals that are published in print still depend on issue packaging.
- How are page numbers handled? Are they assigned online first? How would it be cited?
- Indexes (DRIVER, Google Scholar, Coalition Publica, DOAJ/DOAB, …)
- Google Scholar – preserves at the article level.
- Coalition Publica – preserves at the article level – confirm whether they require issues.
- DOAJ has a very loose XSD, in terms of what is required. If an article has no volume or issue, we do not require one. This List of DOAJ XML elements gives an idea of how lenient DOAJ is – they do not mind what the publishing model is as long as each article has its own DOI and a unique full-text URL.
- OpenAire – relies on OAI PMH.
- Driver – relies on OAI PMH.
- Scopus – This would depend on the agreement the journal makes with the indexes.
- EBSCO –
- Application based indexes
- Are x number of issues a requirement for acceptance
- Tenure and promotion concerns
- Will citation counts decrease if a citation contents (issue) changes?
- Will indexes require a minimum issue count before inclusion? What if we don’t use issues?
- OAI-PMH (DC, JATS, etc)
- JATS does not require issues. OAI PMH – Does it get a new date? Yes, there is a timestamp of the update.
- The only mention of an issue in an OAI field is a mention of it within article metadata (source field), and it’s ok for it to not be present
- German National Library requires specific issue data.
- Copyright/rights transfer/licensing
- This dependency might be more relevant for open peer review
- Would be covered under the author’s agreement?
- Would you change the license between preview and final? Image that there would be a different licence.
- Suppose an author publishes in a journal that does online first using open peer review. Most journals won’t accept articles that are published elsewhere. This may be more of an ope-peer review discussion question.
- On OJS, stamping a different license on different versions is possible. Technically, it shouldn’t be possible to backtrack to a more restrictive license.
- Notification on “item” publication
- As of OJS 3.4, does not support article-level publication.
- Consider allowing users to subscribe to thematic topics and weekly digest.
- Action item: Need to capture more info on this
- Using OJS for pre-print publishing
- Is there a risk that OJS will be used for pre-prints/non-final version publishing
- Interoperability between OJS and OPS via SWORD
- There should be a way to flag when submission belongs to OJS vs. OPS
- Issue assignment/suggestion at submission
- An ability for the author to indicate whether the submission is meant for a packaged issue or a continuous stream
- Article versioning
- At which point does an editorial workflow version become a published/galley version
- Identifying if something will eventually be assigned to an issue or if it’s perpetually continuous
- There may need to be a distinction. Assigned to an issue could only be visible to the editors on the backend.
- Native XML import/export
- Currently supports issue as a unit of export.
- How will issueless publication affect XML exports?
Possible models:
- Just continuous, no issues
- Preview of online first content, then packaged as an issue
- Continuous in parallel with issues
- Issues only
Goals
- Collect & describe use cases for the online first and continuous publishing models
- Narrow down which model to prioritize for the pending OJS release
- Determine whether there is an existing terminology or a need to develop terminology – see if NISO already has something
- To do: Mark H. reached out to NISO, pending a response.
- Need to prioritize model, with a manageable scope to be included in a release
Results
Goal:
- Articles that are ready to go should be prioritized without the requirement of issue assignment
Prioritized model:
- An issue is not required but can be optionally assigned (identified as a high-priority, low-resource/self-contained project/task)
- We could use the existing workflow and have a checkbox to specify that an article can be published online (and assigned to an issue if desired). A DOI could be minted earlier in the workflow. According to Eric, OJS should already support this with not a lot of modification needed, Crossref & Datacite don’t require issue metadata for articles either.
- An editor would need to actively specify whether an article is meant for an issue (could be a back-end field) or for a continuous flow, to enable Crossref validation on the presence/absence of issue/vol number.
- Data model issues needing clarification:
- Article publication status (is the article visible?) – needs to be a characteristic of the version
- Issue publication status: is it publicly visible that any articles in this issue disclose that relationship?
Final use case for Github:
- As an Editor, I would like to be able to publish the work of authors as early as possible rather than waiting for a complete issue to be ready for simultaneous publication.
- I may want to schedule an article for an issue but still have it available in advance of the issue’s formal publication. (I may want to use issue scheduling as an organizational method for when I have a lot of articles awaiting posting.)
- I may or may not want to designate an issue at all, for example, my journal may organize content through thematic areas (categories) and have no use at all for issues.
- I may decide to publish e.g. quarterly issues based on some of the content that has been posted early; this may mean going through already-posted content and publishing a formal issue based on things that have already been made available online. (For articles, this will mean that previously issueless content will have an issue.)
- I would like major functions e.g. DOIs to function predictably regardless of the mix of the above.
FILED! https://github.com/pkp/pkp-lib/issues/9295
Next Steps
To be determined:
- Implications for integrations – see list above
- Upgrade path
- Publication dates may be a problem.
Of interest but out of scope for this issue:
- Open reviews
- Making non-final versions of articles available early
- Reconciling preprint functionality in OJS with OPS
