Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with
or
.
Download ZIP
Commits on Jul 1, 2015
  1. @bdarnell

    Update outdated comment.

    bdarnell authored
Commits on Jun 11, 2015
  1. @bdarnell

    Support Python 3.5 async/await native coroutines.

    bdarnell authored
    Requires changes to be included in 3.5b3.
Commits on May 6, 2015
  1. @bdarnell
Commits on Apr 18, 2015
  1. @bdarnell
Commits on Mar 30, 2015
  1. @bdarnell
Commits on Feb 9, 2015
  1. @bdarnell
Commits on Feb 8, 2015
  1. @bdarnell

    Run autopep8

    bdarnell authored
Commits on Jan 17, 2015
  1. @bdarnell

    Log more uncaught exceptions instead of suppressing them.

    bdarnell authored
    Add more docs about Future exception logging.
Commits on Dec 12, 2014
  1. @bdarnell

    Add support for Future.exc_info when logging abandoned futures.

    bdarnell authored
    Simplify some Future internals.
    
    Add enough magic calls to Future.exception that the test suite
    passes without warning (at least on py34)
Commits on Dec 9, 2014
  1. @bdarnell
Commits on Jun 7, 2014
  1. @bdarnell
Commits on May 25, 2014
  1. @bdarnell

    Autopep8 formatting updates.

    bdarnell authored
Commits on Apr 27, 2014
  1. @bdarnell

    Merge branch 'master' into http_refactor

    bdarnell authored
    Conflicts:
    	tornado/iostream.py
  2. @bdarnell
Commits on Apr 6, 2014
  1. @bdarnell

    Add timeouts for idle keepalive connections in HTTPServer.

    bdarnell authored
    Add gen.with_timeout wrapper.
Commits on Feb 17, 2014
  1. @bdarnell

    Fold TracebackFuture into Future.

    bdarnell authored
    Since it is no longer possible for TracebackFuture to modify
    concurrent.futures.Future instead, there is no need for separate
    classes.  Removing the subclass speeds things up a bit by removing
    a level of indirection.
    
    Also microoptimized a few methods of Future.
Commits on Jan 26, 2014
  1. @bdarnell

    tornado.concurrent.Future is now always thread-unsafe.

    bdarnell authored
    This improves performance and provides more consistent semantics
    (independent of the presence or absence of the concurrent.futures
    package).
    
    The parts of Tornado that accept Futures will accept both Tornado's
    thread-unsafe Futures and the thread-safe concurrent.futures.Future.
Commits on Dec 1, 2013
  1. @bdarnell
Commits on May 16, 2013
  1. @bdarnell

    Add a close method to Resolver and use it where necessary.

    bdarnell authored
    This fixes a thread leak when running the test suite with
    ThreadedResolver.
Commits on Mar 17, 2013
  1. @bdarnell

    Document tornado.concurrent.Future.

    bdarnell authored
    All internal links now point there instead of to concurrent.future.Futures.
Commits on Mar 16, 2013
  1. @bdarnell
  2. @bdarnell
Commits on Mar 15, 2013
  1. @bdarnell

    Rename WebSocketConnect to websocket_connect.

    bdarnell authored
    Add docs for undocumented functions and modules.
Commits on Mar 4, 2013
  1. @bdarnell

    Re-apply autopep8.

    bdarnell authored
  2. @bdarnell

    Introduce TracebackFuture and use it where applicable.

    bdarnell authored
    This allows for better tracebacks when using Futures on Python 2.
  3. @bdarnell
  4. @bdarnell
Commits on Mar 3, 2013
  1. @bdarnell

    return_future and friends now pass the result, not the Future, to the…

    bdarnell authored
    … callback.
    
    Functions and decorators that take an optional callback and return
    a future (return_future, gen.concurrent, and run_executor) no longer pass
    the Future object to the callback.  This results in somewhat less flexible
    error handling, but is more consistent with prevailing practice without
    Futures.
Commits on Mar 2, 2013
  1. @bdarnell

    Make DummyFuture class private.

    bdarnell authored
    It's now only exposed and used when concurrent.futures is not available.
Commits on Feb 23, 2013
  1. @bdarnell
  2. @bdarnell
Commits on Feb 19, 2013
  1. @bdarnell
  2. @bdarnell

    Exceptions in a @return_future function go to both the caller and the…

    bdarnell authored
    … Future.
    
    Add a test for the UnboundLocal error seen here and remove the
    no-longer-needed initialization.  I'm not 100% sure we want to do both
    (due to the risk of callback-doubling), but the old way where the caller
    saw one error and the future resolved to a different error was clearly
    incorrect, and leaving the future unresolved doesn't seem right either.
Commits on Feb 18, 2013
  1. @bdarnell
  2. @bdarnell
Something went wrong with that request. Please try again.