Dear Cultural Partners, GLAMtools, and GWT v.1 steering group,
Following feedback from WMF executive[1] on the GLAMwikiToolset v.2 grant application, it has become clear that the grant’s request for the WMF to reserve code review time cannot be provided as non-monetary support. Nor is the GLAMwikiToolset considered “core site functionality”[1], which would warrant product ownership by the WMF in the long term. Both of these were key points listed in the “non-financial requirements” section of the grant request.[2]
To include these tasks as an enumerated costs in the grant would push the budget of the grant above the WMF's maximum allowable grant amount (an extra 20 hours a week was suggested for code review [1]; no amount of time was suggested for the other items mentioned in the non-financial requirements). Conversely, to reduce the scope of the grant in order to include these costs would reduce the expected benefit below what Europeana sees as helpful to its own non-profit mission. Therefore, I have confirmation from Europeana executive to say that we will be formally withdrawing the grant request and will go back to the beginning.
Specifically, the suggestion was made to “start over with a new external tool“ because the service that the GWT seeks to provide does “...not benefit dramatically from deep integration” in the WMF’s view.[1] We agree that there are significant advantages, both technical and organisational, to this “going independent” approach which make us quite enthusiastic about it.
Re-writing the GWT from scratch as an independent software project (rather than a mediawiki extension) is made possible by new functions (especially OAuth) which didn’t exist when we first started this project (as also mentioned in [1]). This should increase the speed and flexibility of development and hopefully lower the threshold of development knowledge so that developers with little MediaWiki framework knowledge can participate in the project. This should also allow us to attain feature-parity with the current system relatively quickly.
Therefore, soon we will be submitting a new grant proposal to the WMF for a “proof of concept” of an independently operating replacement to the GWT. This will be our way of properly testing if all of the necessary infrastructure (e.g. the API) is in place to do what the GWT needs. If that proof of concept is successful, we will return with a full grant request.
For the moment we suspect that the proof of concept would be deployed on Wikimedia Labs but would probably need to find a more stable home for its finished state. However that, and many other issues (like how to adjust the GWT user-right system to accommodate an external uploader bot; and the question of long term support), are operational questions that would be thought through as part of the proof of concept stage.
Sincerely, Liam Wyatt, on behalf of Europeana (in particular Dan Entous, lead GWT developer, and David Haskiya, Product Manager).
[1] This email, and subsequent emails in this thread: https://lists.wikimedia.org/pipermail/glamtools/2015-February/000343.html
[2] Points number 1 (“timely code review”) and 4 (“product ownership long term”): https://meta.wikimedia.org/wiki/Grants:PEG/Europeana/GLAMwiki_Toolset#Non-fi...