In the January report, I tested a number of tools. I am glad to report now that some of the feature requests made on the way have already been responded to: BaGLAMa now includes a basic timeline feature, and a comparative visualization of its stats for any given two months is available (Open Access example), replacing the need to compile such materials manually. Furthermore, the code has been optimized on several ends, so that these computations now take less time, which provides the possibility to track more categories to watch out for. Thanks, Magnus!
Having enjoyed considerable popularity in the past, the tool had to be retired in between due to lack of resources. It is currently being redesigned with scalability in mind. I would certainly like to see the possibility of including arbitrary domains, especially DOI prefixes or links to Open Access publishers or repositories. Further useful features would be the extension to other languages or projects and provisions for caching, so as to allow for timelines (but see AmalGLAMate section below). I noticed that WebCite was not able to generate any archival copies of Linkypedia, possibly due to the long response times that are in themselves probably due to the currently strained resources.
One of the issues that come up after uploading materials to Commons is whether and how they are actually being reused. AmalGLAMate (WebCite) provides some answers to such questions by exposing time series of usage data. The usage increases for basically all datasets in there, e.g. from 18% in January 2010 to 23% in February 2012 for images from the Federal Archive of Germany.
It would be nice to be able to sync the tracked categories between AmalGLAMate and BaGLAMa.
Navigation is really cumbersome - perhaps add additional navigation bars or split into different pages, as at BaGLAMa.
The Xmas bug
The X-mas bug
I have had a look at the traffic stats for numerous articles relating to Open Access or scientific topics more generally but could not find any that did not show zero hits for Dec 24 and 25, and even pages like Christmas show the same pattern. Has this already been discussed somewhere? The bug also affected BaGLAMa (scroll down to bottom for pageviews on December 24 and 25).
In the meantime, the grok stats visualization of the traffic data was recently improved by Diederik van Liere through migration from jqplot to dygraphs, with some new features added: users can now toggle between displaying the daily numbers, and besides the stats for a given month, a floating window of 30, 60 or 90 days is available.
From that page: "For December 2011 88 hours of data were missing between 23th and 26th. Counts for December have been recalculated to compensate for this gap. (There is a minor adjustment for November as well)"