User:Sj/Platform/Suggestions

What would you like to see from the Foundation or the Board of Trustees? What would you like to see the projects accomplish? What should I personally be doing better? Please leave suggestions, comments, or criticisms here. +sj +

Technical roadmaps
The WMF takes on major projects and editors are left in the dark about when they will be finished. This seems to be better with core MW features than with monolithing projects like sul and flagged revs. Somewhere between meta, mw.org and private project plans, clarity is lost. And when there's an implmentation pushback as with 'create a book', that should show up on a roadmap somewhere as well. I don't know how to fix this, but it's important.

Local discussions
Not all conversations are meta-related; having some planning or other chats here on en would be useful.

Projects
The Board should provide better support for the needs and requests of current and new projects. There is no lack of suggestions and ideas for new Wikimedia domains and initiatives, or for improvements to existing projects that would make them more useful or more friendly to contributors to similar projects elsewhere online.

New project proposals
We need to define a clear process for handling requests for new projects, and we must let these friends and potential partners know how to proceed. In the case of useful free knowledge projects that the Foundation cannot or does not yet wish to include under its umbrella, we should be able to offer other types of advice and support -- they are also pursuing the Foundation's mission in their own way.

Of the major new projects, Rodovid and Wikikids are the most significant projects with active sites up already in multiple languages. They have both asked for Wikimedia support in the past, and received no real reply. The Board is seen as the bottleneck here.

Wikipedia & Commons
Wikipedia gets the bulk of the attention from developers and planners, and has the largest community to advocate for changes. It still lacks a regular review of interfaces and major feature updates (see features below). It is the primary target of the recent Usability initiative, and gets regular specific attention from the Board.

Commons, like Wikipedia, is getting an interface improvement through a specific dedicated grant.

Sister projects
Most of the other sister projects need more usability attention, promotion, and general discussion.

Wiktionary has both a large community and some of the lowest-hanging improvements. Its technical and effort-multiplying needs are well known (starting with sharing one array of translations of a definition across languages). To the extent that the community understands what improvements are needed, this should be a priority. There is much to be learned from OmegaWiki (started and run by Wikimedians), Wordnik (run by an Advisory Board member), and other large online dictionaries.

Wikisource has people and groups interested in using it for large-scale inclusion of OCR and original scans of page-images, for correcting and translating OCR transcriptions. If the community is interested in accepting that sort of use, there are databases of hundreds of thousands of works which could be mined for notable works to add to it.

Wikibooks represents the hottest area of free knowledge over the past six months, with the state of California and even the US President stating their financial and moral support for using free texts to make education more affordable and accessible to all. It does not contain wiki versions of most freely-licensed texts, and has not attracted authors who are deeply committed to free knowledge to edit or actively upload to it. Despite this, there is a growing interest in using wikis -- or similar public, versioned tools -- as a medium for book creation, revision, and translation. We have a number of Advisory Board members who are working on related projects - we should ask them to work explicitly on making import/export to wikibooks function on their own projects (CK-12 is an obvious candidate - it uses MediaWiki as part of its backend, but has yet to export any work to Wikibooks).

Wikiversity is similarly of very broad interest, and the body of freely-licensed digital learning materials has been growing extremely rapidly, but has no other obvious place for public discussion, review, and editing/merging/splitting of them. To the extent that WV can identify where it needs help and technical or outreach support, other projects and the foundation as a whole should provide it.

Wikispecies has strong natural affinities with the Encyclopedia of Life project, and should investigate ways to collaborate. There have been some private discussions between EOL representatives and members of the Foundation or Board; but since this is a content partnership, it is precisely the sort of discussion that should be taking place among and with community members, in public where possible.

Wikinews has great potential both as an outlet for interest among Wikipedia readers in investigating breaking events, and as a way of inspiring a more collaborative sphere of community journalism (which has grown steadily for years but remained a largely solo activity). It has managed to get a few software changes implemented, but could use easier ways to exchange stories and collaborate with other community news bodies.

Wikiquote is occasionally maligned as being a legal liability; something which should be publicly recanted (assuming this is not true) or remedied (in cases where it is). This is an assessment that the Board is in a position to make.

Feature requests and review

 * New page creation by anonymous users : this right was revoked on en:wp by decree after the Siegenthaler affair, theoretically as a test, but its impact was never reviewed and the 'test' has remained in place without further discussion. This sort of change should be reviewed for its effect, so that other projects can decide whether they wish to implement the same, and so that the en:wp community can decide whether to reverse it.
 * The Special:Book extension : this caused a bit of a fuss because it was implemented without sufficient transparency to a group of users who were distracted by its side-effects. This could have been avoided with more active discussion and testing leading up to the roll-out.
 * Flagged Revs : Wikipedians who contribute actively to code discussions are still uncertain of when this is coming, and what is involved... it would be great to have a well-developed list of projects and their timelines and links out to implementations on different test wikis and smaller projects.
 * Side-by-side translation : There are crude versions of this available and regularly used on wikisource to match two language versions of a source document. This could be a useful tool for all projects, and deserves more active discussion.  To the extent that this addresses an element of systemic bias in the projects, it needs more attention than its representation in the community of active Wikimedians would suggest.
 * Translation memory : Google, Microsoft and others have automatic translation tools that can be used to break a page into chunks and translate each chunk into a target language, capturing any modifications made by a human operator who edits and tweaks the translation. We should be capturing this information ourselves into a free public translation memory, which our projects and translation efforts worldwide can use.  The Wikimedia corpus is the only body of general-interest text of its size that is available in many languages -- and even though they are not at all translations of one another, this is a holy grail of sorts of people who study natural language processing.