=================================
2009-02-19T07:14:44.152645Z v0_88
=================================

  2009-02-19 07:14:19 (r361) by rcaputo
  lib/POE/Component/Client/HTTP.pm M

    Time for a new release.

  2009-02-19 07:09:22 (r360) by rcaputo
  lib/POE/Component/Client/HTTP/RequestFactory.pm M;
  t/57_pravus_progress.t A; MANIFEST M

    Add PRAVUS' test case and patch to return progress events even after
    a redirect. Resolves his rt.cpan.org ticket, #36627. 

  2009-02-19 05:20:37 (r359) by rcaputo
  lib/POE/Component/Client/HTTP/Request.pm M

    Comment the RS_* states.

  2009-02-19 04:00:21 (r358) by rcaputo; t/01_request.t M

    Resolve rt.cpan.org #41278, reported by Andreas J. Knig. The
    component test was timing out on his smoker, so I increased the
    timeout. 

  2009-02-18 17:43:15 (r357) by rcaputo
  lib/POE/Component/Client/HTTP/RequestFactory.pm M;
  lib/POE/Component/Client/HTTP.pm M

    Resolve rt.cpan.org #35951, reported by Dave Miller. We were passing
    Tag to POE::Component::Client::Request->new() when it wasn't
    necessary. 

  2009-02-18 17:36:32 (r356) by rcaputo
  lib/POE/Component/Client/HTTP.pm M

    Resolve rt.cpan.org ticket #27643. When a server disconnects before
    responding, POE::Component::Client::HTTP returns a 406 error to the
    application. The "MaxSize" error was incorrect and has been replaced
    with "Server response is Not Acceptable". The surrounding logic has
    been simplified somewhat as well. 

  2009-02-18 07:29:34 (r355) by rcaputo
  lib/POE/Component/Client/HTTP.pm M

    Clarify the docs per Evan Carroll's suggestion. Resolves rt.cpan.org
    ticket 41312. 

  2009-02-18 06:19:51 (r354) by rcaputo
  t/10_shutdown.t M; lib/POE/Filter/HTTPChunk.pm M; t/04_chunk_filter.t
  M; t/02_keepalive.t M; t/11_cancel.t M

    Clean up several warnings introduced by recent POE changes.
    
    Apply Zhurs' patch to resolve rt.cpan.org tickets 43073 and 25855.
    POE::Filter::HTTPChunk would go into an infinite loop on bad chunk
    trailer data, but now it appears to be fixed. Thank you! 

  2009-01-06 03:11:42 (r353) by rcaputo
  lib/POE/Component/Client/HTTP.pm M

    Doc fix. Client::DNS is used by Client::Keepalive and is currently
    not optional.

=================================
2008-12-09T06:22:46.753700Z v0_87
=================================

  2008-12-09 06:17:03 (r351) by rcaputo
  lib/POE/Component/Client/HTTP.pm M; Makefile.PL M

    Time for a new release. Bump the version and fix dependencies.

  2008-12-09 06:10:50 (r350) by rcaputo; t/11_cancel.t M

    Switch a post() to call() to avoid dispatch timing errors that caused
    some machines to fail this test. I think it's the test's fault for
    being intolerant, and I'm not entirely sure the fix will work
    everywhere. 

  2008-12-09 05:44:46 (r349) by rcaputo; t/01_request.t M

    Fix a spurious test failure. Not all systems implement (or even
    enable) Nagle's algorithm the same way. This change makes
    t/01_request.t not as susceptible to packet transmission timing.
    Previously the test would fail of packets were combined (but succeed
    if, like on my development system, they were sent separately). 

  2008-12-09 05:29:27 (r348) by rcaputo
  lib/POE/Component/Client/HTTP/RequestFactory.pm M

    Don't set a content length if the content is a reference. The content
    length would be the reference's stringified length rather than the
    length of any actual content. In the case of CODE references, content
    will be generated in chunks anyway. 

  2008-12-08 07:54:03 (r347) by rcaputo
  lib/POE/Component/Client/HTTP.pm M

    Clean up the leaky ext_request_to_int_id in more places. Apocalypse
    suggested using the requesting HTTP::Request reference directly as
    the internal ID rather than mapping it. That bears some
    investigation. 

  2008-12-08 07:36:19 (r346) by rcaputo
  lib/POE/Component/Client/HTTP.pm M

    New version. RequestFactory doesn not like underscores in version
    numbers.

  2008-12-08 07:27:14 (r345) by rcaputo
  lib/POE/Component/Client/HTTP.pm M

    Internal development release. 

  2008-12-08 07:23:22 (r344) by rcaputo
  t/01_stream.t M; lib/POE/Component/Client/HTTP.pm M;
  lib/POE/Component/Client/HTTP/Request.pm M

    Move data cleanup for request_to_id outside wheel check blocks. The
    ID being cleaned up isn't a wheel's ID. Also renamed the data member
    for future clarity, as well as REQ_REQUEST (related). 

=================================
2008-11-24T22:39:07.052333Z v0_86
=================================

  2008-11-24 22:37:57 (r342) by rcaputo; t/01_stream.t M

    Remove an errant warn in one of the test cases.

  2008-11-24 22:32:18 (r341) by rcaputo
  lib/POE/Component/Client/HTTP.pm M; t/53_response_parser.t M;
  Makefile.PL M

    Bump the version, depend on the latest Client::Keepalive, and fix a
    bit of tab damage. 

  2008-11-21 17:56:00 (r340) by rcaputo; t/01_ssl.t M

    Fix a test error... sourceforge returns a 302 now, not the 200 that
    we expected.

  2008-11-21 08:26:19 (r339) by rcaputo; t/03_head_filter.t M

    Don't use == to test for undef. 

  2008-11-21 08:10:14 (r338) by rcaputo; t/01_request.t M

    Fix an unexpected response triggered by the coderef content feature.
    Two separate packets were sent without coalescing. I have a feeling
    my "fix" isn't going to be good enough in all cases though... some
    systems may coalesce those put()s. 

  2008-10-15 05:44:51 (r337) by rcaputo
  lib/POE/Component/Client/HTTP.pm M

    Apply doc patches, resolving rt.cpan.org #38176. 

  2008-09-03 17:54:09 (r336) by martijn
  t/01_request.t M; t/01_stream.t A; t/10_shutdown.t M;
  t/50_davis_zerolength.t M; t/59_andy_one_keepalive.t D; t/01_ssl.t A;
  MANIFEST M; t/54_hzheng_head_redir.t M; t/09_bad_scheme.t D;
  Makefile.PL M; t/57_joel_immed_cancel.t D; t/58_joel_cancel_multi.t M;
  lib/POE/Component/Client/HTTP/Request.pm M; t/02_keepalive.t M;
  t/11_cancel.t M

    Merge branch 'localtests'
    
    fix merge problems by deleting t/57_joel_immed_cancel.t since my new
    version of t/58_joel_cancel_multi.t does everything it does and more
    and by updating t/58_joel_cancel_multi.t to my version because the
    changes in trunk were assuming the non-local tests. 

=================================
2008-07-13T08:07:37.418057Z v0_85
=================================

  2008-07-13 08:07:22 (r334) by rcaputo
  lib/POE/Component/Client/HTTP.pm M

    Time for another release.

  2008-07-13 06:26:14 (r333) by rcaputo
  lib/POE/Component/Client/HTTP.pm M; t/57_joel_immed_cancel.t M;
  t/58_joel_cancel_multi.t M

    Don't die when canceling requests or shutting down components. 

  2008-07-09 19:44:56 (r332) by martijn; t/04_chunk_filter.t M

    forget about reading from DATA. This way we can control how much data
    we feed to the filter, so we get better coverage. 

  2008-07-09 19:44:41 (r331) by martijn; t/04_chunk_filter.t M

    de-poe-ify 04_chunk_filter.t too. 

  2008-07-09 19:44:21 (r330) by martijn; t/03_head_filter.t M

    get coverage up to same level as with the old tests. 

  2008-07-09 19:44:07 (r329) by martijn; t/03_head_filter.t M

    stop using POE while testing Filter::HTTPHead. 

=================================
2008-05-28T17:35:59.670622Z v0_84
=================================

  2008-05-28 17:33:06 (r327) by rcaputo
  lib/POE/Component/Client/HTTP.pm M; t/14_gzipped_content.t M;
  lib/POE/Component/Client/HTTP/Request.pm M

    Resolve Content-Encoding issue surfaced by recent changes to
    HTTP::Response. Applies the fix recommeded by Yuri Karaban in
    rt.cpan.org ticket 35538. 

  2008-05-04 06:57:15 (r326) by rcaputo
  lib/POE/Component/Client/HTTP/Request.pm M

    Resolve character encoding issues arising from a change in
    HTTP::Message's decoded_content() behavior. Tentatively resolves
    rt.cpan.org ticket 35538, reported by Yuri Karaban, with additional
    commentary and a test case by "OverlordQ". 

  2008-04-20 00:34:15 (r325) by rcaputo; MANIFEST M

    Add t/59_andy_one_keepalive.t 

=================================
2008-04-20T00:30:17.408340Z v0_83
=================================

  2008-04-20 00:29:46 (r323) by rcaputo
  lib/POE/Component/Client/HTTP.pm M

    Release time! 

  2008-04-20 00:26:49 (r322) by rcaputo
  lib/POE/Component/Client/HTTP.pm M

    Apply Kim Scheibel's patch to support a BindAddress parameter. Useful
    for running on multi-homed machines. Resolves rt.cpan.org ticket
    34837. 

  2008-04-19 22:56:52 (r321) by rcaputo
  lib/POE/Component/Client/HTTP.pm M

    Apply Yuri Karaban's patch to avoid losing responses when a redirect
    fails. Resolves rt.cpan.org ticket 25577. 

  2008-04-19 22:46:23 (r320) by rcaputo
  lib/POE/Filter/HTTPChunk.pm M; lib/POE/Component/Client/HTTP.pm M

    Apply Yuri Karaban's patch to respond correctly if the server closes
    connection before sending headers. Resolves rt.cpan.org ticket 25920. 

  2008-03-25 07:51:12 (r319) by rcaputo; lib/POE/Filter/HTTPHead.pm M

    Apply Jeffrey Posnick's patch to support duplicate headers. Before
    his patch, subsequent headers of the same type would overwrite
    previous ones. For example, a second Cookie header would clobber the
    first. Resolves rt.cpan.org 27262. 

  2008-03-25 07:42:11 (r318) by rcaputo
  lib/POE/Component/Client/HTTP/Request.pm M

    Apply Dan Sterling's patch to use Time::HiRes when it's available. He
    was running into premature timeouts when using subsecond timeouts.
    Resolves rt.cpan.org 30466. 

  2008-03-24 18:05:00 (r317) by rcaputo; Makefile.PL M

    Rely on the latest POE::Component::Client::Keepalive, which includes
    a fix for a common noisy POE::Kernel trap. 

  2008-03-24 18:03:29 (r316) by rcaputo
  lib/POE/Component/Client/HTTP/Request.pm M

    Apply Andrew Sterling Hanenkamp's patch to not pass cookies across
    redirects. Previously cookie headers would bleed with a redirect,
    which is extremely bad. Resolves rt.cpan.org ticket 30400. 

  2008-03-24 17:53:15 (r315) by rcaputo
  lib/POE/Component/Client/HTTP/Request.pm M

    Apply Andrew Sterling Hanenkamp's suggestion not to bail out of
    redirect if we reach the same URL twice. It was doing the wrong thing
    with certain site authentication schemes. Resolves rt.cpan.org ticket
    30399. 

  2008-03-24 17:44:53 (r314) by rcaputo
  lib/POE/Component/Client/HTTP/Request.pm M

    Apply Andrew Sterling Hannenkamp's patch to frob redirect cookies
    (they count!). Resolves rt.cpan.org ticket 30401. 

  2007-08-17 13:54:25 (r313) by andyg; t/59_andy_one_keepalive.t A

    A failing test for reusing a single keep-alive connection (run with
    DEBUG)

  2007-05-20 12:50:02 (r312) by martijn
  lib/POE/Component/Client/HTTP/Request.pm M

    r955@kazad-dum: martijn | 2007-05-20 12:06:14 +0200 spell my name
    right :) 

==============
End of Excerpt
==============
