Sven,
thank you for your honest opinion, and I know that you are not alone with it - but I also heard a lot of people express excitement and joy about the deployment, and based on the activity it seems that a lot of people like it. We consider ourselves happy to be part of an intelligent and critical, and at the same time sympathetic community.
I fully agree that the first deployment of Phase II functionality was early. A lot of features are missing, as we have repeatedly communicated.
Was the deployment too early? That I disagree with. It is widely accepted wisdom for software projects to "release early, release often". I think that is very valid advice, and I decided to follow it with the project. This allows us to see if some of our basic assumptions "work". This allows us to test a few things before we fully commit to them and spend big amounts of effort without any reality check.
The project plan as a whole was planned like this. Basically language links are just a test-run for some of the technology we will need in order to implement Phase II. Many back-end features -- the propagation of data from a central repository to the Wikipedias, the way recent changes deals with this, the scalability of some of our assumptions -- are equivalent in Phase I and II. Phase I was always implemented with Phase II in mind. We are doing the same thing now. We implement some features -- namely statements per se, references for statements -- but with a limited set of data sets and with some major limitations. But these will get expanded over time.
Compare it with another project like the Visual Editor. It is deployed to the English Wikipedia. Plenty of features are still missing. But only with their current deployment schedule can the VE team gather crucial data for their further development. The main difference is that VE is an opt-in feature -- statements in Wikidata are not, they are there, in your face.
I regard a project like Wikidata not as a software development project. It is a growing, living socio-technical system, and in this case actually it is one embedded in an even bigger such system, the Wikimedia movement as a whole. We are developing technical features that we think will lead all of us towards our common goals, and then we watch how the communities adapt to them, which social rules they build on them, which technical developments of their own are built on top of ours. We (as the development team) are part of this ecosystem, and we (as all of us Wikimedians) are growing together. Technical possibilities shape the rules the Wikidata editor community agrees on, and the actual usage of the system and your feedback shapes and prioritizes the future technical development that we plan and undergo as the development team.
I also see that some decisions of the community are based on the currently available features, but i do not think that this is problematic -- because I am very confident that future new features will continue to shape new rules and that the existing ones will be revisited and updated accordingly.
The timing of the deployment of phase II to
wikidata.org has nothing to do with the deployment of phase I to the English Wikipedia, which is currently scheduled for Monday. We simply deployed features when they are deemed ready. We do not plan features ahead with the intention to keep interest high, or in order to win editors from other communities, etc. Also, we regard phase II only as sufficiently finished when it is actually deployed to the Wikipedias. And this, obviously, still requires a much better support for references and a bigger number of data types. Also, so far there is no reason to believe that there will be any major problems revealed once we deploy to the English Wikipedia.
I regard your feedback as very important, and I am thankful for it. I understand that everyone would like to have all the features immediately. But I disagree with you on the point that we should not have deployed last Monday. We were working very hard in order to be able to deploy on Monday, and not wait even longer, and we are very proud with how smoothly it went - fully conscious of the limitations of the current state. The situation will soon improve, and we would like to stay a project for now that deploys new features in a comparably quick succession.
After this explanation I hope that I have the support of the Wikidata community to continue in this spirit.
Cheers,
Denny