New from the University of Wolverhampton in the UK, “Open Access Books in the Humanities and Social Sciences: an Open Access Altmetric Advantage”…

“examines the altmetrics of a set of 32,222 books (of which 5% are OA) and a set of 220,527 chapters (of which 7% are OA) indexed by the scholarly database Dimensions” in Humanities and Social Sciences.

Open Access Week: Events listing for 22nd – 28th October 2018.

My thanks to Jan Szczepanski for alerting me that his list has a new home. Szczepanski’s List of Open Access Journals is now available via the EBSCO.com website, as a freely downloadable .DOCX Word file. The page describes this 35,000-title / 9Mb / 672-page subject-ordered list as… “probably the world’s largest list of open access journals in the humanities and social sciences” plus Law, Statistics and Geology. No “probably” about it, this fine work is surely the largest OA-only list which is also freely available to the public. The EZB is at a similar scale, and does have 37,600 OA titles since 2010, but the EZB covers all disciplines including the sciences.

I had assumed the list had long since stopped updating circa 2015 and was now a legacy item, but the new header now notes “2,932 titles added 2017”. The file’s internal datestamp says the last update in Word was made March 2018. Unfortunately there’s no “First included in (date)…” component to the list entries, by which one might extract just the new 2016-18 finds. Nor does EBSCO appear to offer a rolling “What’s New”.

Having found the list again I’ve now added a link to it on the JURN Guide / Directory / FAQ pages. Despite its inevitable 15% linkrot (see below for details on that) Jan’s list has a big advantage over the JURN Directory. In that it can serve as a discovery tool for OA titles published in languages other than English, and which are unlikely to be on the DOAJ for various legitimate reasons. In contrast, the JURN Directory only lists titles which publish in English or are genuinely multi-language with English.

The home-page URLs given in Jan’s list appear to be hyperlinked with blue-underlining, but my copy of Word won’t allow me to launch these. That’s probably just a security thing on my PC. A simple “Save as…” to a .PDF does give me launchable URLs.

One can also save the list from Word to a filtered .HTML page, which gives a 23Mb Web page with launchable URLs. Despite being so large, my Linkbot software didn’t have problems with my saved Web page version of the list. 60 minutes later the Linkbot results reported that the list has the disadvantage of link-rot:

* Contains 45,536 individual home-page URLs across approx. 35,000 titles (some records have more than one URL). Of broken…

* After discounting the useful redirects (e.g.: http:/ to https:/ or the ubiquitous OJS redirects from the home-page to the current issue page), an Excel tally told me there were 6,808 ‘fatal’ URLs such as outright 404s, ‘Unauthorized’ or ‘Host not found’. That’s a 15% rot rate.

Perhaps someone could now look at helping Jan with a crowd-funder, to pay some Web-elves (on Fivver.com or similar) or unemployed recent graduates to fix the broken 15% of URLs? At $20 per batch of 25 URLs, I figure it should cost around $6,000.

itty.bitty, new from the design leader at Dropbox. Itty.bitty uses the URL to contain the text of a Web page. The page can have 2,000 bytes, or about 170-200 words, if you’re going to support legacy Web browsers such as Internet Explorer.

No hosting server is required, and as the data sits after the # symbol. What comes after the # is meant to be page-position related, and as such it never gets sent to the server.

The base64 link code is not pretty…

But the same link/page displays as…

How it works

This link is the page.

Scripting and hyper-linking is enabled in such pages, so long as it all fits in the URL length. The code can’t do images, but you can do old-school ASCII-art.

The main drawback seems to be that you’re going to have to be 1000% sure that your text is exactly as you want it before you make the link… because there’s no after-post editing for errors or updating of dead hyperlinks in the page.

In which case you’d ideally consistently version and date-stamp the

">How it works

bit, as…

">How it works (v.0.1 | 14/07/2018)

…so that people and search-tools can discover later updated versions of the same content. Otherwise the itty.bitty system risks becoming an intertwingled mess of half-baked and old/broken stuff that you (and probably Google) won’t want in search results.

I’m guessing that advanced Web browsers such as Brave will soon ‘add a feature’ in relation to this, by enabling much longer data-carrier code to be read from URLs. Perhaps also some simple automatic “…and can we find a later version of this itty.bitty.site?” query, done inside the browser. There would, however, also have to be some sort of dynamic ownership hash embedded in the page, to protect against impersonation of the page-author. Perhaps the system of authoring an ownership-hash and datestamp could be combined into a simple ‘one-click operation’ in a desktop authoring tool.

Anyway, it’s one example of the coming uncensorable Decentralized Web.