Commits on Sep 18, 2011
  1. Add few more examples.

    committed Sep 18, 2011
Commits on Sep 15, 2011
  1. Merge branch 'release/2011-09-15-1'

    committed Sep 15, 2011
Commits on Sep 12, 2011
Commits on Sep 9, 2011
Commits on Sep 5, 2011
  1. Minor: Upgrade to Streamline 0.1.41.

    Fixes minor sync stack trace filename bug for CoffeeScript.
    Streamline seems to be fairly stable now, and hopefully Bruno will stop
    introducing regressions ;), so loosening dependency requirements too.
    aseemk committed Sep 5, 2011
Commits on Sep 4, 2011
  1. Upgrade to Streamline 0.1.40 for sync traces!

    Requires requiring the flows module, which sets the appropriate hook. We don't
    currently use the flows module, but this is no big deal.
    Details and discussion:
    aseemk committed Sep 4, 2011
Commits on Sep 2, 2011
  1. Update CHANGELOG.

    committed Sep 2, 2011
  2. Bump version.

    committed Sep 2, 2011
Commits on Aug 30, 2011
Commits on Aug 29, 2011
  1. Error handling tweaks.

    - For Cypher queries, Neo4j incorrectly returns HTML in cases of 500 Internal
      Server Errors instead of JSON as it should, so protect against that.
    - Don't show the server-side stack trace on the client side; that's useless,
      especially now that Streamline shows sexy synchronous stack traces! =)
    aseemk committed Aug 29, 2011
  2. Merge pull request #3 from aseemk/patch-1

    Update obsolete README parts.
    committed Aug 29, 2011
  3. Update obsolete README parts.

    aseemk committed Aug 29, 2011
Commits on Aug 28, 2011
  1. Hotfix: freeze Streamline dependency.

    We were using Streamline 0.1.35 locally, and 0.1.37 introduced a regression
    with futures. It's been quickly fixed in 0.1.39, but I'm not sure yet if
    0.1.37 is a safe upgrade yet: it awesomely transforms error stack traces to
    synchronous stacks instead of async ones, but it might not work as expected if
    the code doesn't explicitly require the flows library, which we don't.
    Investigating that, but for now, freezing the dependency on 0.1.35 to protect
    against the regression.
    aseemk committed Aug 28, 2011
  2. Fix error handling for database errors.

    Instead of throwing just a number, throw the actual response, so that the
    adjustError() function can use it to (nicely) set the server stack trace.
    And update for newish request: don't parse JSON if it's already parsed.
    aseemk committed Aug 28, 2011
Commits on Aug 26, 2011
  1. Add new db.queryNodeIndex() method.

    For more complex queries instead of single key-value lookups.
    aseemk committed Aug 26, 2011
Commits on Aug 25, 2011
  1. Internal API change: Relationship constructor.

    Instead of `db, start, end, type, data`, making the first two args `db, data`
    required and in the same order as the Node and PropertyContainer constructors,
    removing `type` since it's already part of `data`, and making `start, end`
    optional args since most of the code fakes/mocks the Node instances anyway, so
    the Relationship constructor now does that automatically.
    aseemk committed Aug 25, 2011
  2. Make query() API actually usable.

    Still return an array of rows (matches), but have each row (match) return a
    map from variable name (from the query) to the value, instead of returning
    just the values with a separate array of variable names.
    aseemk committed Aug 25, 2011
Commits on Aug 19, 2011