In addition to what Steve said, this particular "feature" is really just the very beginning for us. Our ideas about what query-able version control mean go way beyond the ability to select arbitrary sets of commits based on commit meta data. We want to be able to do higher order queries. For instance:
For each line of code written by Johnny what is the average life span of a line of code.
Basically how lasting on average are lines of code written by a particular developer. You could combine this query to be only on a particular subgraph forest of the repository (for instance a branch, or commits this quarter). These types of queries are aggregation queries, however there are other classes of queries we would like to support as well. See some of the papers we wrote on scribd for more details. We will be writing a another post soon covering our vision for version control query.
In addition to what Steve said, this particular "feature" is really just the very beginning for us. Our ideas about what query-able version control mean go way beyond the ability to select arbitrary sets of commits based on commit meta data. We want to be able to do higher order queries. For instance:
For each line of code written by Johnny what is the average life span of a line of code.
Basically how lasting on average are lines of code written by a particular developer. You could combine this query to be only on a particular subgraph forest of the repository (for instance a branch, or commits this quarter). These types of queries are aggregation queries, however there are other classes of queries we would like to support as well. See some of the papers we wrote on scribd for more details. We will be writing a another post soon covering our vision for version control query.