Oslo, último acto

Por primera vez, Charles Young, un usual sostenedor de las iniciativas de  Microsoft, ha decidido hablar con crudeza acerca del anteúltimo proyecto de desarrollo de la empresa, Oslo, y sus productos asociados, Quadrant y M. En semanas pasadas, varios especialistas hablaron de la desaparición final del proyecto, desde hace ya algún tiempo reducido a producto de nicho, al pasar del ámbito general de Visual Studio al particular de SQL Server. Hasta ahora, sin embargo, había preferido considerar estas afirmaciones como conjeturas, no existiendo una palabra oficial. Pero ahora, Young, decepcionado, recoje palabras oficiales o casi, y esto sí significa un certificado de defunción. Defunción en segundas nupcias, ya que además de haber caído Oslo, ahora cae lo que queda en el proyecto SQL Server Modelling.
Así lo ve Young:

RIP Quadrant. Adios, SSM Repository
So it's official. Don Box has blogged an update on SQL Server Modelling. As widely rumoured, he has confirmed that Quadrant is dead.   The shock is that the model repository has expired as well.
M is still hanging on in there, but we don't know anything more about what the future holds.  For some time, the official word has been that it would arrive in conjunction with a future SQL Server release. I detect a possible change here, in that Don simply says 'We are continuing our investment in this technology and will share our plans for productization once they are concrete'.   For my part, I maintain that tying 'M' to the SQL Server licence would be a really bad move. I know that every time Doug Purdy (who has recently left Microsoft to join Facebook) was challenged on this, he kept saying...no, SQL Server is a platform, not a product...watch this space.   However, in the next breath it certainly sounded as if 'M' was in danger of being subsumed under a product license. No, no, no. Set it free to be used across Microsoft's platform.   That platform is not SQL Server.  It remains Windows, and for developers, it is Windows (on-premises and cloud) mediated through .NET.

Atrás quedan las justificaciones (recordadas por Young) acerca  del paso del proyecto Oslo al SQL Server Modelling, y no me queda duda que las palabras de Don Box corren el mismo camino que las de Douglas Purdy. Cómo cierra Box la historia de Oslo, Quadrant y M:

We created the “Oslo” repository to make the model of a system or application easily accessible without relying on application-specific machinery to consume or query those models. The “Oslo” repository achieves this by storing the models for applications and systems in a shared SQL Server relational database.

Over the past year, we’ve gotten strong and consistent feedback from partners and customers indicating they prefer a more loosely-coupled approach; specifically, an approach based on a common protocol and data model rather than a common store. The momentum behind the Open Data Protocol (OData) and its underlying data model, Entity Data Model (EDM), shows that customers are acting on this preference.

Given the increasing adoption of both OData and EDM, we have decided to focus our investments in those technologies to make our modeling vision a reality. One important aspect of that focus is that we will not bring “Oslo” repository to market. We believe that taking a loosely coupled, federated approach using OData and EDM will ultimately get more models exposed sooner than an approach based on building a common repository database.

We created a visual tool codenamed “Quadrant” to help people query, update and visualize information that is stored in SQL Server databases.  As with the “Oslo” repository, customers told us that they wanted to work with information from a variety of sources, not just data stored in an RDBMS. Customers also told us that they wanted the experience to be native to the tools they are already using; specifically either Visual Studio or Microsoft Office. Given this feedback, we are not bringing “Quadrant” to market. Instead, we will work to make the experience with OData and EDM in Visual Studio and Microsoft Office even better.

Finally, we created a language codenamed “M” for defining schema, constraints, queries, and transformations. While we used “M” to build the “Oslo” repository and “Quadrant,” there has been significant interest both inside and outside of Microsoft in using “M” for other applications. We are continuing our investment in this technology and will share our plans for productization once they are concrete.

Dada la errática línea de investigación (si se puede hablar de ello), quizá haya que hacer apuestas sobre el futuro de estas nuevas, terceras afirmaciones sobre un mismo asunto.

Quisiera aclarar que es lo particularmente incómodo de esta historia:
1) La dificultad en confiar en este tipo de proyectos: un largo período de tiempo empujando el mercado en un sentido, creando expectativas inútiles.
2) Relacionado con lo anterior, la ola de afirmaciones de entusiastas seguidores y partners acerca de la brillantez de un producto que nunca pasó de CTP (Community Technology Preview). Y esto, dicho sin medir el impacto que implica vender promesas sobre otros productos que tratan de hacerse duramente un lugar en el mercado.