Multimedia Review and Publishing with OJS

[jamieallen]

Hi all!

I'm writing as I've been developing (with colleagues: Paul Boshears, Nico Jenkins and Bernhard Garnicnig) a quarterly journal called Continent based on all the wonderful provisions of the OJS system. The journal is here: http://www.continentcontinent.cc/

<extra info on who we are and continent>
By way of personal introduction - we're quite an international team, but my main job is currently as Assistant Director of Culture Lab in the UK. Other people working on the project are in the U.S. (Maine and Atlanta, Georgia) and our lead developer (Bernhard Garnicnig) is from Vienna, Austria. We're all associated in various ways with the European Graduate School in Switzerland (http://www.egs.edu/).

The main goal of our quarterly is to collect contemporary philosophical reactions to culture, politics and society, and our specific emphasis is to try and do this in a way that is 'media agnostic'. What this means is we'd like to set up a journal which accepts not only text, but also includes opportunities to showcase photography, media works, film, and audio to be included as an equally valuable 'academic submission' to the publication, and 'academic output' from it.
</extra info on who we are and continent>

We've been doing great so far with OJS, but are at the point where we're looking to develop a kind of multimedia version of OSJ (perhaps a development fork?) that would allow people to submit, and panelist to review and publish, the diverse set of multimedia available elsewhere online. The hope is we can get things working for non-text as well as OJS already works for text based submissions, essentially. There are a number of likeminded journal initiatives out there (and more to come in future) we'd like to help build this with and for. John Willisnky of OJS has already hipped us to Kairos (Kathie Gossett) and the Scalar group that developed Vectors and IJLM. Anyone on this forum part of those initiatives?

Is anyone else out there interested / working in the same direction? Was hoping to put together a small interested user-group to help develop the features of a multimedia fork / plugin / what-ever-it-takes.

All the best - and thanks to everyone here for your work and interest in Open Source publishing!
Jamie

[asmecher]

Hi Jamie,

That journal looks excellent -- I'm sure a lot of hard work went into it.

I'm hoping we can avoid a development fork of OJS if possible -- actually, I suspect a lot of the things you have in mind are somewhere on our own development agenda. Our priorities in terms of encouraging / fostering work with the OJS codebase are roughly as follows:

• Adaptations of the system that make sense for many of our users and are unobtrusive for the rest can be incorporated directly into our own code repository for release in a future version.
• Adaptations that are vendor-specific (i.e. to link with another system) or that aren't broadly useful can, if possible, be written as plugins. These are generally exchanged via the plugin gallery at http://pkp.sfu.ca/support/forum/viewforum.php?f=28 -- though we do also distribute many with OJS itself.
• Adaptations that are too tricky to do via plugins and aren't suitable for distribution with OJS can still be maintained as external modifications. We tend to exchange these when desired as unified .diff files.

We currently maintain both a stable and a development branch ourselves, and I'd be happy to describe this a little further if you'd like. Currently the development branch shouldn't be used for production purposes, but we tend to direct people there if they want to contribute code that's not suitable for the stable branch (i.e. anything that would require a round with our translators, which we don't intend to coordinate until the next development release).

As for multimedia, it sounds like John has already directed you to some of the other journals that have been experimenting. We haven't included many multimedia-specific tools in OJS simply because we haven't had a partner journal to work with and help drive the functionality forward -- but there are a few tools already included that might be useful. For example, have you seen the URL rewriting tools in classes/article/ArticleHTMLGalley.inc.php? Those should allow for URLs within HTML galleys to be rewritten to refer to e.g. supplementary files attached to the articles, or other articles entirely; this should permit interlinking. You've probably seen that you can upload images and CSS as part of an HTML article; you can also include multimedia like videos, and OJS should rewrite the URLs accordingly (e.g. if you use something like FlowPlayer).

I know one of the current limitations when working with multimedia is OJS's "single file" view of what an article should be. I'd be interested in working out some specs for a wrapper to help work around this; our forthcoming OMP software will be much more flexible in this regard, but it's not ready yet. (OMP developments will start to roll back into OJS and OCS, including more modern approaches to the UI.)

Would you mind describing some of the specific requirements and ideas you have in mind? Where is OJS currently limiting you, and what modifications / tweaks have you been making? I might be able to suggest approaches or point out other projects and collaborations that could be useful.

If you're able to make the conference in September in Berlin, we're very interested in kick-starting some projects at the Hackfest. I suspect we'll be able to find some good ideas here.

Regards,
Alec Smecher
Public Knowledge Project Team

[jamieallen]

Hi Alec, all...

Thanks for the response - glad these ideas generate some excitement! Sorry for the delayed response - I actually had a hell of a time logging back into this forum (seems very secure! :)

We would love some advice as to what the right mechanisms for pursuing the goals of a multimedia submission and presentation system within or on top of OJS are. Your adaptations list above is really useful. It may be that a lot of what we’re after can be accomplished via plugins? Good to know that the priorities in any case. We would be very interested in becoming a ‘partner’ journal for driving development, in any case.

To get discussion brewing (hopefully!) I’ll list a few plain language ideas we’ve bandied about for continent. If we could get peoples thoughts and impressions on these, especially if there are already ways of accomplishing them, that would be excellent.

Our current push is to try and get a bit of funding in to develop in OJS for continent and the community a set of tools for interlinked and direct uploading of any type of digital media. For our journal, a main tenant of this direction is that there should be a ‘media agnosticism’ in the way that both submissions and presentation/publication of works/articles. We reject the prioritization of text over image, book over video, research over creativity. This is an ever-changing concept for our publication, one which by definition must remain open, but one that we think is an interesting and exciting part of the future of academic and online publishing. We are looking to tap our board of advisors to help articulate these ideas further (this amazing group currently includes Clay Shirky, author of 'Here Comes Everybody'), as well as help advise development.

So here’s a very forward looking, long list of what we’d like develop in OJS:

1. Submission of multimedia. This would allow submitters to select a type of submission (photo, video, sound, other, etc) and submit this via the usual review system offered by OJS. Because uploading, transcoding and hosting of any type of media files is something thats hard to “get right” we encourage the use of existing leading media hosting services like Vimeo for video, Flickr for photosets, Soundcloud for audio etc. There could also be account linking between OJS logins and these sites, much as social media sites are doing (via Facebook/Twitter login, OpenID etc)

2. Review of multimedia. Paired with the above, it would be great if the review process now associated with text and article submissions could also preview, embed and allow comments on the same multimedia pieces referred to above (either uploaded or hosted). This would require admin interfaces for playing, perhaps timeline-tagging and rating multimedia for acceptance or rejection by the journal.

3. Presentation of multimedia. The final step to this chain is, of course, being able to embedded hosted and uploaded journal media, automatically as part of the acceptance procedure of compiling an issue full of articles. There would be a good amount of work to do here providing formatting constraints and dealing with players, file types, online conversions, and interlinking. Again - if this could work like social media currently does with Wordpress-like sites, automatically embedding and applying CSS styles, etc., all the better (but this is no small task, I realise).

4. Direct links to downloadable versions of each article. Currently, we format and generate PDFs of each article and include these as links, so that interested readers can grab a copy to read later in a presentable format (instead of printing from the browser). For a publication addressing certain humanities communities concerned with digital aesthetics, this has proven quite important as people still very much ‘pass around’ PDFs, and it’s also (kind of oddly) a kind of validation of the publication to have PDFs available. What would be great for text is a PDF generator for the journal with a layout manager that the Editors could control/author. Likewise, something of this kind for multimedia (specifying a maximal resolution, download type, etc.) would be great.

5. Auto generation of ‘packaged’ versions of entire journal issues. This is less defined in our current design discussions, as there are difficulties imagining what the ‘downloadable’ version of a multimedia journal should be. An application? A set of folders / files? This latter idea seems straightforward if OJS could, given a fully configured journal issue with all articles, generate a ‘.zip’ of the whole issue, with text converted to PDF (as per point 4.), video to an appropriate format, audio to MP3, etc., in a formal structure of folders that made sense to the user when unzipped. Static versions of a packaged journal issue would also be great - but difficult here is the idea of a ‘rendering’ engine that might include a large PDF edition, which automatically uses the poster frame from a video to format a ‘page’ representing its position and article description in the document.

All of this, in our particularly context, would in a perfect world need to be implement in a way that people who don’t code CSS or fully understand “embed tags.”

So that’s a lot of stuff, and quite general in scope, but hopefully gives a sense of our end goals. Thoughts? Again - if anyone out there is interested in such things, we’re really looking for collaborators to move forward with funding/resourcing a development push in this direction.

Best,
Jamie

[asmecher]

Hi Jamie,

Perhaps the trickiest part of this is going to be figuring out what stable of services you want to integrate. The more you can use external services (Soundcloud, Vimeo, Flickr) the less tricky stuff you'll have to write yourself, but at the same time, the more you'll limit yourself in terms of how you can interact with these services.

Getting OJS to link out to them instead of using file uploads for submissions is a relatively small task, but you'll have to consider how each service permits rich interaction with the content -- reviewing, downloading/uploading for archive, revision, etc. -- and how much of the review / archiving / etc. you'll be able to entrust to the external service and how much you'll keep locally. I'd suggest starting with a good review of the APIs available to one of the more mature and understood services, like Flickr. You could choose any of a number of approaches for each -- for example, would the review process be entirely kept within OJS, with only a link out to external content for presentation? Would the review process make use of the Flickr commenting system somehow? And how would you deal with submissions that crossed media boundaries, mixing several services together?

It might be worth mocking up a few "best-case" examples and seeing how close you can get with the technologies you've got in mind. And of course I'm happy to make suggestions and clarify OJS's own limits.

Regards,
Alec Smecher
Public Knowledge Project Team