Many platforms for disseminating Open Access content online exist, and perform to different degrees of satisfaction. A distinction can be drawn between platforms that are freely downloadable (i.e. open source) and closed (paid for) options. A distinction can also be drawn between fully-downloadable and user-configured platforms, and third-party hosting packages. But foremost, each platform or hosting package is distinguished by how well it allows a journal to perform in an online environment.
That last distinction is more important than it sounds: desktop-bound, offline journal management and print publishing as a practice has existed for ages, and moving those workflows online should be done carefully and only when doing so represents a clear move forward for the journal involved. Choosing which service is right for you is a matter of correctly judging your needs and resources, and matching them against what is available.
Evaluating Online Publication Tools
So what should an online publication platform offer over print, anyway? By far the best argument for online (and especially Open Access) publication is that of access: putting content online makes it available anywhere there is an internet connection, which is now just about anywhere. But not only should your prospective platform be able to provide full-text, openly-accessible content; it should also include mechanisms for actively disseminating that content over the web, and preserving that content as well, preferably in open, standard ways.
Also, the web is no longer a static compendium of docile HTML pages and hyperlinks, but a neworking and research hub where people talk and share information, and where resources grow and change by the second. To account for this, journal publication tools must do much more even than provide content for download: they must provide a context for understanding; tools to support further research; and communication tools to promote access and discussion.
In a related vein, you should determine whether the tools you evaluate are simply mechanisms for the dissemination and preservation of scholarly content, or whether they support and enhance typical offline journal workflows, such as submission solicitation and management; the peer review process (in any incarnation you may need, and especially taking into consideration the sometimes-fickleness of reviewers); and the final editorial and copy-editing workflows necessary to your journal.
Access and Dissemination
Succeeding at online journal publishing takes a lot more than just throwing content online and walking away: this content needs to be publicised, added to search engines and catalogues so that interested parties will find it, easily accessed in standard, broadly-supported file formats and, last but not least, supported and enhanced with the inclusion of research and contextual tools. When evaluating a platform or hosting plan, you may wish to consider whether or how the following points are supported:
• Access options. You should be able to find a platform to fit the access type and business model of your choice, whether that is full Open Access, delayed access, an article processing fee (APC) model, or even subscription access.
• Publication file type options. The two most commonly-supported publication file types are HTML and PDF, but XML is seeing increasing use, especially as it is being used more frequently as part of a larger workflow management system; see the section on XML publishing tools below for a more general discussion regarding XML. Generally speaking, though, you should know what kind of file types you want to present your content in, the advantages and possible disadvantages of each, and whether any given system supports them.
• Support for supplementary files in multiple formats. This may not be of concern to all disciplines, but if supplementary files such as data sets and images are used extensively by your journal, you will want to ensure that any platform you are evaluating somehow allows for them. Since some digital supplementary files can be quite large, server capacity (disk space and file upload capabilities) should also be checked. You will also want to check whether supplementary files are publishable alongside article content if you so desire, and whether they are accessible during any workflow process.
• Email and/or RSS notification of content. Users should be able to receive notification of new content via these mechanisms.
• Search-engine friendliness. Publishing platforms should facilitate good indexing by Google/Google Scholar, Yahoo!, and others; common practices, for example, are to include relevant metadata information in HTML pages, and to provide XML sitemaps for faster search engine crawling.
• Easy access to metadata for harvesting. The most common way to share scholarly metadata is via the Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH). This protocol allows for the efficient distribution and harvesting of article and issue metadata across disparate platforms, namely into institutional repositories, online harvesters and scholarly databases. For example, the Synergies Project uses a central OAI harvester to collect object metadata from regional nodes across the country so that it can be displayed in one convenient, central location. Many scholarly databases rely on the OAI protocol, and you should check whether it is supported.
• Support for necessary usage reports and statistics gathering. This is especially important for satisfying research granting agencies and for authors’ tenure appointments. Understanding how your journal’s contents are being used is also critical in understanding your relationship with the wider world and building upon it. Modern platforms should not only interact well with the more common third-party statistics tools (for example, Google Analytics or the server-side programs AWStats and Webalizer); they should also be able to generate useful and accurate custom usage reports in easy-to-use formats such as XML or CSV. If your institution makes use of the COUNTER protocol for tracking usage data, you may also want to check whether the platform in question supports it.
• Easy deposit into discipline-specific catalogues or databases. Some databases, such as Pubmed Central, have specific guidelines on how content is added. If you think you may want your journal content to be added to such databases you will want to check that the platform or service you use can support said deposits.
• The inclusion of value-added tools. Most systems now provide some sort enhanced reading environment for visitors: links to author information; word lookup tools; access to relevant databases; citation information; and so on. Examples include Open Journal Systems’ Research Support Tools and Hyperjournal’s Dynamic Contextualization tool.
See Kevin Stranack’s article, Getting Found, Staying Found, Increasing Impact: Enhancing Readership and Preserving Content for OJS Journals, for other dissemination considerations ( http://pkp.sfu.ca/files/GettingFoundStayingFound.pdf).
Equally as important as dissemination is the matter of preservation. Whether you are hosting your journal with your own service provider; relying on your home institution for hosting support; or paying a commercial provider for the service, you should ensure that there are acceptable preservation methods in place. This goes beyond simple file and database backups, although those are important too: you will want to ensure that institutional change, rapid advances in technology, migration from platform to platform, and the simple effect of time have as minimal an effect as possible on your body of work. There are a number of common practices to keep in mind as you evaluate platforms and hosting options:
• Keep offsite backups. This can be as simple as automatically making daily/weekly backups of all files and database content via the Internet to a machine hosted elsewhere, or could involve backing up all data to another medium such as CD/DVD or tape. The key points to remember are that these backups themselves are corruptible, and need to be checked; and that the backup plan and procedure itself should also be checked periodically.
• Use your publishing platform in conjunction with an institutional repository. Institutional repositories are being adopted across the academic world, and typically include some sort of robust backup/preservation plan. DSpace http://dspace.org and Fedora Commons http://www.fedora-commons.org/ are the two most widely-used IR platforms, and some journal platforms support direct deposit to them via various protocols.
• Use a comprehensive preservation system like LOCKSS http://lockss.org. The LOCKSS (“Lots of Copies Keep Stuff Safe”) program is a distributed storage/backup system where multiple institutions store copies of any given item; it also includes a number of checks against stored content to ensure that nothing has been tampered with, compromised, or otherwise damaged.
• Evaluate the platforms’ internal preservation methods, if any. Does the platform manage and archive a paper trail for submissions, or for actions taken on a submission? Is there some form of due process and acknowledgement for editorial decisions? A lot of the traditional practices of keeping paper around an office are lost when moving to a digital environment, and any reasonably capable workflow-management software should account for this somehow.
• Evaluate your own, or your service provider’s preservation of offline workflow material. It may be the case that you are still accomplishing a fair bit of workflow work outside the system proper, for example when it comes to galley creation. Whether using Quark or Pagemaker to create a final issue or using an XML workflow process, these files should also be strongly considered for preservation.
There are a number of online resources specific to digital preservation, some of them national in scope (for example, The Library of Congress’ National Digital Information Infrastructure & Preservation Program http://www.digitalpreservation.gov/ in the United States, or the Digital Preservation Coalition http://www.dpconline.org/graphics/index.html in the UK. Wikipedia http://en.wikipedia.org/wiki/Digital_preservation also covers digital preservation and lists a number of additional resources.
Workflow patterns differ from journal to journal and from platform to platform, and quite often the only certainty you can count on is that a move to online publishing will necessitate some degree of change in how you would otherwise handle affairs. The good news is that typically, you can choose whether to jump in wholeheartedly and use any given platform’s workflow patterns, or to stick with your own for the time being and use only a subset of the system (ditching online submission management, review and editing workflows but using online publication and preservation components, for example). Also, changes might be less disruptive than you might think, and may even result in more streamlined and cleaner practices overall. Common workflow components include:
• Submission solicitation and management. This includes the ability for authors to submit content to your journal quickly and efficiently, and for them to include important metadata and other supplementary files as need be; for editors to be able to efficiently manage incoming submissions, separating chaff and moving worthy submissions onward; and for authors to receive ongoing feedback on their submission.
• Peer review management. This includes the assigning of reviewers; completion of a full review of a submission, including the use of review forms if necessary; support for multiple review rounds and a fully blind review process.
• Copy- and layout editing, and final proofing. This includes moving submissions between copyeditors, proofreaders, layout editors, editors and the authors themselves; support for layout templates, if used; and support for any final checks that you may find necessary, as well as scheduling and publishing.
Finally, the notion of general flexibility is something you will want to keep in mind as you explore your options. The questions you should be asking yourself are, how easily can I adapt this system to my own needs? Can I change the code, and if so how will that affect me later? Are there custom support services available, and at what price? What happens if I eventually want to move to another system? You may never find yourself in a position where these questions really need to be answered; but if you do, and the answer is negative, you might wish you had asked them in the first place.
There are a number of hosting options you will have to consider when starting to publish your content online. Quite often, your institution may be able or willing to provide hosting, and may in fact already do so; you may want to try hosting on your own, with your own Internet service provider; or you may want to go with an established commercial practice that provides academic tools online. The most common hosting options are as follows:
• Self-hosting. Typically, this means you download, install and maintain the necessary software packages yourself, in a web environment of your choosing. Web hosting options are now cheap and abundant enough that individuals can purchase a domain and web space for as little as a few hundred dollars a year; after which it is simply a matter of installing software packages that meet your needs. A word of caution, however: "simply" is relative, and although many of these software packages' system requirements are low, they take time and a certain level of expertise to set up and maintain. You will also want to ensure that your web hosting package properly meets all system requirements for any software you wish to run. The biggest advantage of self-hosting is that you control the software and so can make any modifications that are supported by the software and within your means.
• Institutional Hosting: Many academic institutions now provide system installation and support for the more popular online academic tools, or may be willing to if there is enough interest from their community. You may lose some control over the configuration of the software, but you will generally be guaranteed reasonable maintenance and support options. One major advantage of this option is that you have the weight of the institution behind you, and can likely fit in with their backup and preservation practices. Your institution's library or IT support staff should know whether there are institutionally-operated services available to you, and should additionally be able to help in configuration and ongoing maintenance issues.
• Commercial Hosting: As the prevalence of online academic publishing increases, so do commercially-provided hosting and support options. Service levels and business models differ widely, but at a minimum you should ensure that you have full editorial control over the review, editing, publishing and dissemination of content. Please note, also, that some academic institutions may offer commercial support to non-affiliated users; examples include the Electronic Text Centre at the University of New Brunswick http://journals.hil.unb.ca, and Simon Fraser University Libraries http://software.lib.sfu.ca/support.html. As with any contract you sign, the one between the service provider and yourself should be carefully read and understood, especially with regards to access to scholarly content: some service providers mandate varying levels of access control as a cost-recovery measure. Examples of commercial providers include Co-Action Publishing http://www.co-action.net/, Scholarly Exchange http://www.scholarlyexchange.org/, BioMed Central http://www.biomedcentral.com/, Multimed Inc. http://www.multi-med.com/index.php, and the Hindawi Publishing Corporation http://www.hindawi.com/.
What follows is a brief survey of the most commonly-used open access publishing tools. Please note that the author works for the Public Knowledge Project (PKP) http://pkp.sfu.ca and so while the information on PKP products below is based on actual usage, all information on non-PKP software applications comes from their respective websites.
For an excellent and in-depth overview of the state of the most commonly-used publishing tools circa 2007, see Mark Cyzyk and Sayeed Choudhury’s research article A Survey and Evaluation of Open-Source Electronic Publishing Systems https://jscholarship.library.jhu.edu/handle/1774.2/32737,
Open vs. closed-source?
For a full discussion on what 'open source' means, visit the Open Source Initiative's website http://www.opensource.org/. It is worth noting that while publishing open access content does not necessitate the use of open source tools, the two do pair up very nicely philosophically: both advocate opening traditionally closed or restricted development models for the benefit of researchers, academics, developers and the greater good in general. The tools discussed below are all open source (although specific licenses between them may differ), and as such are freely downloadable and modifiable.
Open Journal Systems
Open Journal Systems (OJS) http://pkp.sfu.ca/ojs is easily the most popular application developed by the PKP. It is a journal management and publishing system currently being used by more than 2000 online journals, in over 57 countries. OJS is a full-scale, multi-journal publishing platform that supports a wide range of access options and business models, including full open access; varying levels of article processing charges; author-initiated institutional deposit of materials; and even delayed-access and full subscription options. OJS offers the following features and more:
• OJS is installed locally and locally controlled.
• Editors configure requirements, sections, review process, etc.
• Online submission and management of all content.
• Optional subscription module with delayed or full open access options.
• Comprehensive indexing of content part of global system.
• Reading Tools for content, based on field and editors' choice.
• Email notification and commenting ability for readers.
• Easy to translate, with 17+ localizations http://pkp.sfu.ca/ojs-languages already fully or partially complete and available for use.
• Complete context-sensitive online Help support.
OJS has to be downloaded http://pkp.sfu.ca/ojs_download and installed on a server to run; the server system requirements (listed on the download page) are minimal, with older versions of PHP and MySQL or PostGreSQL supported. Significant development effort has been spent to ensure that OJS is easily maintainable and upgradable, not to mention extensible: it includes an extensive plugin system to make it easier for developers to extend its core functionality. More information on OJS can be found on its documentation page http://pkp.sfu.ca/ojs_documentation; demonstration and test drive examples are also available http://pkp.sfu.ca/ojs_demo.
Current Version: OJS 2.2.3 http://pkp.sfu.ca/ojs_download
Minimum System Requirements: PHP >= 4.2.x; MySQL >= 3.23.23 or PostGreSQL >= 7.1; Apache (preferred), Lighttpd or IIS web server environment. More details are available on the download page.
HyperJournal http://www.hjournal.org/ is an open source software package specifically tailored for open access publishing. Features include OAI compliance; dynamic contexualization (whereby cross-references in articles are identified and hyperlinked); access to the HyperJournal Network (connecting all journals using the HyperJournal software); a customizable interface; anonymous article submission and blind peer review processes; and support for a plurality of permitted file formats.
Current Version: 0.5 (beta) or 0.4.1.2 http://www.hjournal.org/download
Minimum System Requirements: Linux (Debian suggested), mac OS X10.2 or higher, *BSD, any flavour of Unix; PHP >= 5.0.1, with specific PEAR and PHP packages installed. For more detailed information, see the system administrator documentation included with the download package.
The ePublishing Toolkit (ePubTk) https://dev.livingreviews.org/projects/epubtk/ is an open source software package, written in Python, to help with the running of scientific journals. The package is developed by the Living Reviews http://livingreviews.org organization, a publisher of scientific open access journals.
Current Version: n/a. ePubTk is not currently available as a package for download, although you can download the source code from their repository https://dev.livingreviews.org/projects/epubtk/browser. There are plans to eventually release a version of the software as a Python module.
Minimum System Requirements: Subversion client (for software download); Python >= 2.3, with a number of specific packages also installed; PostGreSQL; Apache 2.x. For more detailed information, see the ePubTk installation pages https://dev.livingreviews.org/projects/epubtk/wiki/Installation.
GAPworks http://gapworks.berlios.de/ is an online publication system developed by the German Academic Publishers (GAP) Project http://www.sub.uni-hamburg.de/gap/. It is not explicitly a journal-publishing platform, and can handle different types of publications. Features include a peer review process; user management functions; OAI-PMH compliance; and a configurable template system.
Note that, although GAPworks is still available online for download, it appears that development has been nonexistent since 2006.
Current Version: GAPworks-core 0.9.63 http://developer.berlios.de/project/showfiles.php?group_id=4921
Minimum System Requirements: Recent versions of PHP and PostGreSQL; its own (virtual) www server, tested with Apache 2.x.
DPubS (Digital Publishing System) http://dpubs.org/ is an online academic publishing system that supports publishing scholarly journals, conference proceedings, monographs, and more. It is developed by Cornell University Library, and is what runs Project Euclid http://projecteuclid.org/, among other sites. It has been developed with flexibility in mind; is extensively customizable, right down to the user display; supports different access types from full open access to subscription/pay-per-view; and can interoperate with institutional repositories (DSpace and Fedora).
Current version: 2.3 http://sourceforge.net/projects/dpubs Minimum System Requirements: Perl 5.8+; Apache/mod_perl; Java 1.4.2+.
Topaz http://www.topazproject.org/trac/ is a nonprofit organization related to the Public Library of Science (PLoS) http://plos.org. It manages the development of the Topaz application framework and the Ambra publishing system that overlies the framework. From their website: “Ambra is a high-volume, efficient and economical system for the publication of quality-assured research in all areas of science. Ambra is a unique publishing forum that exploits the full potential of the web to make the most of every piece of research. Written mostly in Java, Ambra is a webapp that can be run compliant servlet containers.”
Current Version: 0.92 http://www.topazproject.org/trac/wiki/Download
Minimum System Requirements: Cygwin (if running on Windows); Subversion (1.3 or later); JDK (1.5 or later); Maven (2.0.8 or later).
The Drupal E-Journal http://drupal.org/project/ejournal module allows you to create and control your journals through the Drupal http://drupal.org CMS. Inspired by OJS, the module provides access and user management options, and allows the user to use any other Drupal components that might be useful. Although currently considered a proof-of-concept system by the developer, it has been in use for a couple of years and is still seeing active development.
Current version: there exists a stable and development version for both the 5.x and 6.x Drupal lines.
Minimum System Requirements: A working Drupal install.
XML Publishing Tools
Finally, a word (or several) about the use of XML in publishing. While word processing documents are still the norm for desktop review and editing, and while some journals may still be using larger-scale desktop publishing software such as QuarkXPress or Pagemaker (superseded by InDesign), the use of XML in publishing is becoming increasingly popular. There are a number of good XML schemas or DTDs out there to choose from, but the most popular for academic publishing are Text Encoding Initiative (TEI) http://www.tei-c.org/index.xml; Docbook http://docbook.org; and the NLM Journal Publishing DTD http://dtd.nlm.nih.gov/publishing/.
Encoding journal content in XML has several advantages: XML is an easily-readable, standards-based and easily preserved markup language that is also easily mutable into other filetypes, including HTML, other types of XML, and even PDF. The academic community is very used to dealing with XML, and there are a number of XML authoring tools out there that make working with it easier, including Oxygen XML http://www.oxygenxml.com/; XML-Spy http://www.altova.com/; and the XMLmind XML Editor http://www.xmlmind.com/xmleditor/. (Most popular XML editors should include support for the above XML formats.)
Of the XML formats mentioned above, NLM deserves particular consideration, as it has been written specifically with journal publishing in mind, and has been widely adopted for use outside the medical academic world (see “A Standard XML Document Format: The case for the adoption of NLM DTD?” http://www.alpsp.org.uk/ngen_public/article.asp?id=335&did=47&aid=1244&st=&oaid=-1). There are a number of XML authoring tools that handle NLM creation specifically, including the PKP’s Lemon8-XML http://pkp.sfu.ca/lemon8, the Microsoft Article Authoring Add-in http://www.microsoft.com/downloads/details.aspx?FamilyID=09C55527-0759-4D6D-AE02-51E90131997E&displaylang=en, and eXtyles NLM http://www.inera.com/extylesinfo.shtml#NLM.
It is worth considering whether XML might a useful tool in your publishing toolbox, and whether the platforms you evaluate support it. XML isn’t a panacea, but with a bit of extra work you may find yourself with a very flexible, powerful and long-standing publishing backend for your journal.
A Survey and Evaluation of Open-Source Electronic Publishing Systems
See also the OA journal launch services list at Open Access Directory: