Adding/Amending Facebook Metadata in a 25 Issue Archive
DOCUMENTATION OF PRACTICE
Blackbird: on online journal of literature and the arts, did not, until volume eleven in 2012, routinely add the necessary meta data to direct the image and content of a shared link in Facebook. (This corresponding with the rise of Facebook and the journal’s own use of it to promote readership and educational outreach.)
On those archival content pages lacking the meta data, the image that appears in the Facebook share (if it appears at all) may not be what the journal prefers, which is either the headshot of the contributor or the journal’s logo. In addition, the text that appears as an auto-abstract will likely be a confusing, collapsed list of contributor names rather than our preferred teaser of the opening lines of content.
For instance, here is a Facebook share of a Larry Levis poem from Blackbird v2n2
vs. the share of another Levis poem from v13n2; this example includes only two additional lines of meta data and reflects a different page-titling standard.
The share above creates a more deliberate and professional look.
Until such time, however, that we can retrofit the entire archive with meta data for each piece of content, we do focused intervention.
If one of our contributors makes the news with a major award or appointment— or dies— we know that archival content is more likely to be shared out by our readers. When we can, we prepare ahead of the curve, if sometimes grimly, as notice comes of news that we think will result in focused shares of archived material. (This, out of our respect for our contributors only, as we seek to maintain or broaden the legacy of those we have published.)
Either I, or another staff member at my direction (most usually the associate editor, lead pagebuilder, or lead developer) will work through a list of previous contributor content, add the metadata, and decide whether to point all pages to the most recent headshot, or to the headshot published adjacent to content in a given issue.
Once the image path and content teaser is added to the code of the archived page, we publish it (overwriting the previous page in the archive) and immediately check it live for any unforseen problems; we then double-check the results of our meta data changes or additions in the Faceook Open Graph Object Debugger (formerly the Facebook linter)
https://developers.facebook.com/tools/debug/
MAINTENANCE & DEVELOPMENT ISSUES
Current practice has us putting a bare minimum of meta code for Facebook, which suits our currently identified needs. The example here is the code for “Elegy Ending in the Sound of a Skipping Rope,” the Levis Facebook share pictured previously above.
<!-- BEGIN FACEBOOK META INFO <meta name="description" content="All I have left of that country is this torn scrap / Of engraved lunacy, worth less now . . ."/> <link href="http://www.blackbird.vcu.edu/v13n2/images/contributors/levis_l.jpg" rel="image_src" title="Larry Levis" id="Larry Levis" /> <!-- END FACEBOOK META INFO -->
Facebook’s debugger warns us, however, that some of the “open graph properties” shown in the debugger for a page with this code are “inferred” from other HTML tags rather than directly given under Facebook’s Open Graph protocol. It is important, then, that we settle whether or not we need to add more comprehensive Open Graph fields to our code for purposes of integrating with Facebook, especially before attempting any major intervention to the archive.
In tandem, we need to investigate library meta-data standards. If we are to include metadata for a preservationist standard, we would want to undertake such an effort simultaneous to the addition of Open Graph code.
M.A. Keller
Categories Archive maintenance, Meta data