Skip to content
Commits on Sep 18, 2011
  1. @gasi
  2. @gasi

    Add few more examples.

    gasi committed Sep 18, 2011
Commits on Sep 15, 2011
  1. @gasi

    Merge branch 'release/2011-09-15-1'

    gasi committed Sep 15, 2011
Commits on Sep 12, 2011
  1. @aseemk
  2. @aseemk
Commits on Sep 9, 2011
  1. @aseemk
  2. @aseemk
Commits on Sep 5, 2011
  1. @aseemk

    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
  2. @aseemk
Commits on Sep 4, 2011
  1. @aseemk

    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:
    
    http://groups.google.com/group/streamlinejs/browse_thread/thread/58d53eff1dfe02d8
    aseemk committed Sep 4, 2011
Commits on Sep 2, 2011
  1. @gasi
  2. @gasi

    Update CHANGELOG.

    gasi committed Sep 2, 2011
  3. @gasi

    Bump version.

    gasi committed Sep 2, 2011
  4. @gasi
Commits on Aug 30, 2011
  1. @aseemk
  2. @aseemk
Commits on Aug 29, 2011
  1. @aseemk

    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. @gasi

    Merge pull request #3 from aseemk/patch-1

    Update obsolete README parts.
    gasi committed Aug 29, 2011
  3. @aseemk

    Update obsolete README parts.

    aseemk committed Aug 29, 2011
Commits on Aug 28, 2011
  1. @aseemk
  2. @aseemk
  3. @aseemk

    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
  4. @aseemk

    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. @aseemk
  2. @aseemk

    Add new db.queryNodeIndex() method.

    For more complex queries instead of single key-value lookups.
    aseemk committed Aug 26, 2011
  3. @aseemk
Commits on Aug 25, 2011
  1. @aseemk
  2. @aseemk
  3. @aseemk

    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
  4. @aseemk

    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
  5. @aseemk
  6. @aseemk
  7. @aseemk
  8. @aseemk
Commits on Aug 19, 2011
  1. @aseemk
Something went wrong with that request. Please try again.