===========================
2005-06-27 05:36:25 v0_7001
===========================

  2005-06-27 05:33:49 (r181) by rcaputo
  t/03_head_filter.t M; t/04_chunk_filter.t M; HTTP.pm M

    Remove POE::Filter::XML from the tests. It appears not to be needed,
    and it causes the tests to fail because it's not a prerequisite for
    this distribution. Bump the version in HTTP.pm for a minor fix
    release. 

=========================
2005-06-23 01:13:21 v0_70
=========================

  2005-06-23 01:08:34 (r179) by rcaputo
  t/01_request.t M; HTTP.pm M; t/02_keepalive.t M

    t/02_keepalive.t was incorrectly matching responses to their
    corresponding tests. The result was a near-infinite loop between
    tests 6 and 7. This commit resolves that issue, shaving about 120
    seconds off the tests.
    
    t/01_request.t ambiguously used GET and POST. I added parens so that
    the parameters are explicit.
    
    HTTP.pm wasn't passing a timeout parameter from the request through
    to Client::Keepalive. I thought this might also result in tests that
    ran longer than they should. Anyway, it seems to be more proper
    behavior. 

  2005-06-22 16:22:17 (r178) by rcaputo
  t/01_request.t M; t/03_head_filter.t M; t/05_request.t M;
  t/06_factory.t M; HTTPChunk.pm M; Request.pm M; t/04_chunk_filter.t M;
  Makefile.PL M; HTTP.pm M; RequestFactory.pm M; t/02_keepalive.t M;
  HTTPHead.pm M

    Add Id tags, and set the svn:keywords property on files. Changed the
    tabbed indents to hard spaces, and tidy up the resulting mess. 

  2005-06-21 17:53:32 (r177) by rcaputo
  t/03_head_filter.t M; META.yml D; MANIFEST M; Makefile.PL M; CHANGES D;
  mylib D

    The SSL dependencies moved over to POE::Component::Client::Keepalive,
    so we were able to remove ExtUtils::AutoInstall (and the entire mylib
    directory). Cleaned up the MANIFEST and Makefile.PL to reflect that.
    
    Modified Makefile.PL to use svn-log.perl instead of cvs-log.perl.
    
    Removed CHANGES and META.yml from the repository since they're
    automatically generated at "make dist" time. 

  2005-05-11 17:25:13 (r176) by martijn; RequestFactory.pm M

    documentation skeleton for RequestFactory 

  2005-05-11 16:02:13 (r175) by martijn; Request.pm M

    a redirect response can have content, so don't move to RS_DONE 

  2005-05-11 16:02:03 (r174) by martijn
  Request.pm M; RequestFactory.pm M

    use the FollowRedirects parameter instead of hardcoding the number 

  2005-05-11 15:29:08 (r173) by martijn; Request.pm M; HTTP.pm M

    remove useless commented out code, and set the RS_REDIRECTED state in
    Request.pm 

  2005-05-11 15:28:57 (r172) by martijn; Request.pm M; HTTP.pm M

    move _remove_timeout into Request.pm 

  2005-05-11 15:28:48 (r171) by martijn; RequestFactory.pm M

    remove import of Request.pm internals. 

  2005-05-11 15:28:38 (r170) by martijn; RequestFactory.pm M

    remove empty, unused check_redirect. the real one was in Request.pm 

  2005-05-11 15:28:25 (r169) by martijn
  Request.pm M; HTTP.pm M; RequestFactory.pm M

    initiate the cookie frobbing in Request.pm, so we can pass the
    response. yay better encapsulation 

  2005-05-06 19:54:17 (r168) by martijn; HTTP.pm M

    r5569@fangorn: martijn | 2005-05-06 21:54:40 +0200 update copyright
    info 

  2005-05-06 19:07:02 (r167) by martijn
  README M; HTTPChunk.pm M; t/04_chunk_filter.t M; CHANGES.OLD M;
  mylib/ExtUtils/AutoInstall.pm M

    Add more tests to t/04_chunk_filter. Also fix a bug in
    Filter::HTTPChunk which was discovered thanks to these tests. 

  2005-05-06 18:23:15 (r166) by rcaputo
  t/03_head_filter.t M; t/04_chunk_filter.t M

    Simplified the preparation of the DATA filehandle for use with
    Wheel::Readwrite. Worked around a bizarre case where BSD derived
    operating systems would immediately EOF on the first sysread() from
    DATA. Oddly enough, Martijn's code does essentially the same thing
    but on a duplicated filehandle. Using DATA directly seems to fix it. 

  2005-05-06 14:37:59 (r165) by martijn; Request.pm M

    r5514@fangorn: martijn | 2005-05-06 16:35:21 +0200 Another bit of
    carefulness changing REQ_STATE I forgot 

  2005-05-06 14:37:50 (r164) by martijn; HTTP.pm M

    r5513@fangorn: martijn | 2005-05-06 16:33:36 +0200 some doc updates 

  2005-05-06 14:37:41 (r163) by martijn; HTTP.pm M

    r5512@fangorn: martijn | 2005-05-06 16:24:20 +0200 oops, forgot a
    bit. remove a superfluous if/else, and be more careful about changing
    state 

  2005-05-06 14:37:23 (r162) by martijn
  t/01_request.t M; Request.pm M; HTTP.pm M; RequestFactory.pm M

    r5511@fangorn: martijn | 2005-05-06 15:59:41 +0200 Move lots of
    processing code into Request.pm, which allows for more encapsulation.
    Use Test::More for 01_request.t and restructure the tests in the
    process 

  2005-05-06 14:37:13 (r161) by martijn; Request.pm M

    r5510@fangorn: martijn | 2005-05-06 15:24:46 +0200 put use statement
    at the top where it belongs, and use HTTP::Status's response code
    constant instead of a numeric 

  2005-05-06 14:37:01 (r160) by martijn; Request.pm M; HTTP.pm M

    r5509@fangorn: martijn | 2005-05-06 15:14:54 +0200 argh. run tests
    more often 

  2005-05-06 14:36:50 (r159) by martijn; HTTP.pm M

    r5508@fangorn: martijn | 2005-05-06 15:12:23 +0200 forgot this 

  2005-05-06 14:36:39 (r158) by martijn; Request.pm M; HTTP.pm M

    r5507@fangorn: martijn | 2005-05-06 15:05:10 +0200 Handle redirects
    better. Not perfect yet 

  2005-05-06 14:36:30 (r157) by martijn; HTTPChunk.pm M

    r5506@fangorn: martijn | 2005-05-06 14:45:50 +0200 Return a
    HTTP::Headers object instead of needing the HTTP::Response object
    returned by Filter::HTTPHead. Also fix some logic errors 

  2005-05-06 14:36:19 (r156) by martijn; HTTP.pm M

    r5505@fangorn: martijn | 2005-05-06 14:28:45 +0200 add comment about
    bug #12363 

  2005-05-06 14:36:09 (r155) by martijn; HTTP.pm M

    r5504@fangorn: martijn | 2005-05-06 14:25:25 +0200 prepare for other
    transfer-encoding filters 

  2005-05-06 14:35:59 (r154) by martijn; RequestFactory.pm M

    r5503@fangorn: martijn | 2005-05-06 14:13:56 +0200 make single agent
    string work 

  2005-05-06 14:35:48 (r153) by martijn; Request.pm M

    r5502@fangorn: martijn | 2005-05-06 14:11:59 +0200 remove code
    duplication for error messages 

  2005-05-06 14:35:36 (r152) by martijn; Request.pm M

    r5501@fangorn: martijn | 2005-05-06 14:09:22 +0200 remove unused
    field from Request 

  2005-05-06 14:35:24 (r151) by martijn; Request.pm M; HTTP.pm M

    r5500@fangorn: martijn | 2005-05-06 14:06:22 +0200 merge RS_IN_STATUS
    and RS_IN_HEADERS into RS_IN_HEAD 

  2005-05-06 14:35:13 (r150) by martijn; HTTP.pm M

    r5499@fangorn: martijn | 2005-05-06 13:56:24 +0200 remove \n from
    warn messages 

  2005-05-06 14:34:59 (r149) by martijn; Request.pm M; HTTP.pm M

    r5498@fangorn: martijn | 2005-05-06 13:52:04 +0200 add accessor for
    REQ_TIMER remove useless parameter from _finish_request 

  2005-05-06 14:34:48 (r148) by martijn; HTTPHead.pm M

    r5497@fangorn: martijn | 2005-05-06 13:15:52 +0200 add a bit of
    documentation, and comment out some unused methods for now 

  2005-05-06 11:03:58 (r147) by martijn; RequestFactory.pm M

    r5490@fangorn: martijn | 2005-05-06 13:04:31 +0200 Put the correct
    version in the agent string 

  2005-05-06 11:03:50 (r146) by martijn; HTTP.pm M; RequestFactory.pm M

    r5489@fangorn: martijn | 2005-05-06 13:00:42 +0200 keep the timeout
    in RequestFactory 

  2005-05-06 11:03:42 (r145) by martijn
  Request.pm M; RequestFactory.pm M

    r5488@fangorn: martijn | 2005-05-06 12:55:28 +0200 Keep reference to
    its factory in a Request. This will be used later on, I promise :) 

  2005-05-06 11:03:36 (r144) by martijn; Request.pm M

    r5487@fangorn: martijn | 2005-05-06 12:52:00 +0200 Add parameter
    checks 

  2005-05-06 11:03:29 (r143) by martijn; RequestFactory.pm M

    r5486@fangorn: martijn | 2005-05-06 12:49:50 +0200 croak if the
    arguments aren't in a hashref. This makes the prematurely committed
    2nd and 3rd test of t/06_factory.t pass 

  2005-05-06 11:03:20 (r142) by martijn
  MANIFEST M; Makefile.PL M; HTTP.pm M

    r5485@fangorn: martijn | 2005-05-06 12:44:42 +0200 Remove
    HTTP::Headparser, which we stopped using, and fix up requirements 

  2005-05-05 07:00:59 (r141) by rcaputo; Request.pm M

    Turn off debugging. 

  2005-05-05 07:00:33 (r140) by rcaputo; HTTP.pm M

    Turn off debugging, and wrap a debug warning in DEBUG so it'll turn
    off. 

  2005-05-05 06:40:52 (r139) by rcaputo; Makefile.PL M

    Remove SSL.pm from the indexer. Also reformat the file for a little
    more sanity. 

  2005-05-05 06:39:19 (r138) by rcaputo; MANIFEST M

    Fix paths of things, also remove SSL.pm, so Makefile.PL is happy. 

  2005-05-04 19:54:14 (r137) by martijn; HTTPHead.pm M

    r5449@fangorn: martijn | 2005-05-04 17:19:12 +0200 add a DEBUG and
    use it for some warns

  2005-05-04 19:54:04 (r136) by martijn; chunked D

    r5448@fangorn: martijn | 2005-05-04 17:15:50 +0200 remove old data
    for chunk filter tests

  2005-05-04 19:53:54 (r135) by martijn; t/05_request.t M

    r5447@fangorn: martijn | 2005-05-04 17:15:29 +0200 Adapt tests to now
    needing a Factory parameter in Request

  2005-05-04 19:53:43 (r134) by martijn; t/04_chunk_filter.t M

    r5446@fangorn: martijn | 2005-05-04 17:15:16 +0200 use Test::More
    instead, and put the test data after __DATA__ instead of in a
    seperate file

  2005-05-04 19:53:33 (r133) by martijn; t/02_keepalive.t M

    r5445@fangorn: martijn | 2005-05-04 17:14:45 +0200 indent

  2005-05-04 19:53:22 (r132) by martijn; SSL.pm D

    r5444@fangorn: martijn | 2005-05-04 17:14:28 +0200 remove SSL.pm, as
    this is handled through SSLify in Client::Keepalive

  2005-05-04 19:53:14 (r131) by martijn; t/06_factory.t A

    r5443@fangorn: martijn | 2005-05-04 17:14:16 +0200 add some tests for
    RequestFactory

  2005-05-04 19:53:05 (r130) by martijn; t/03_head_filter.t M

    r5442@fangorn: martijn | 2005-05-04 17:14:01 +0200 Add a test for
    just a status line (no headers). This makes sure that bug #12363
    isn't caused by HTTPHead.pm

  2005-05-04 19:52:55 (r129) by martijn; t/03_head_filter.t M; foo D

    r5441@fangorn: martijn | 2005-05-04 17:13:35 +0200 Better tests for
    HTTPHead.pm. It's now almost completely covered

  2005-05-04 19:52:43 (r128) by martijn; t/05_request.t A

    r5440@fangorn: martijn | 2005-05-04 17:13:17 +0200 add some tests for
    Request.pm

  2005-05-04 19:52:34 (r127) by martijn; Request.pm M; HTTP.pm M

    r5439@fangorn: martijn | 2005-05-04 17:12:30 +0200 Remove unused
    HTTP::HeadParser code

  2005-05-04 19:52:22 (r126) by martijn; chunked A; foo A

    r5438@fangorn: martijn | 2005-05-04 17:12:15 +0200 files needed for
    the filter tests

  2005-05-04 19:52:12 (r125) by martijn
  t/01_request.t M; HTTP.pm M; t/02_keepalive.t M

    r5437@fangorn: martijn | 2005-05-04 17:11:02 +0200 Make it use
    Client::Keepalive

  2005-05-04 19:51:58 (r124) by martijn
  Request.pm A; RequestFactory.pm A

    r5436@fangorn: martijn | 2005-05-04 17:09:19 +0200 Add Request.pm and
    RequestFactory.pm that I forgot in the previous commit

  2005-05-04 19:51:47 (r123) by martijn
  t/01_request.t M; t/03_head_filter.t M; HTTPChunk.pm M;
  t/04_chunk_filter.t M; Makefile.PL M; HTTP.pm M; t/02_keepalive.t M;
  HTTPHead.pm M

    r5435@fangorn: martijn | 2005-05-04 17:07:36 +0200 bad commit. done
    when I had too many changes, when I wanted to push things into an
    accessible repository. Moves things from HTTP.pm into Request.pm and
    RequestFactory.pm, but I forgot to add those files.

  2005-05-04 19:51:28 (r122) by martijn; HTTP.pm M

    r5434@fangorn: martijn | 2005-05-04 16:55:06 +0200 break stuff by
    adding the chunk filter

  2005-05-04 19:51:18 (r121) by martijn
  t/03_head_filter.t A; t/04_chunk_filter.t A

    r5433@fangorn: martijn | 2005-05-04 16:51:13 +0200 add tests for new
    filters

  2005-05-04 19:51:07 (r120) by martijn
  HTTPChunk.pm A; MANIFEST M; Makefile.PL M; HTTP.pm M; HTTPHead.pm A

    r5432@fangorn: martijn | 2005-05-04 16:49:35 +0200 add and use a
    filter for the HEAD portion of a HTTP response also add a filter for
    content-encoding chunked, but that isn't used yet

  2005-05-04 19:50:53 (r119) by martijn; t/02_keepalive.t M

    r5431@fangorn: martijn | 2005-05-04 16:48:50 +0200

    r3236@fangorn: martijn | 2004-11-04 20:40:00 +0100

    remove traces of Spiffy.pm

  2005-05-04 19:50:44 (r118) by martijn
  t/01_request.t M; t/02_keepalive.t A

    r5429@fangorn: martijn | 2005-05-04 16:44:33 +0200

    r3235@fangorn: martijn | 2004-11-04 20:38:26 +0100

    updated tests for HTTP/1.1

  2005-05-04 19:50:31 (r117) by martijn; HTTP.pm M

    r5428@fangorn: martijn | 2005-05-04 16:43:26 +0200

    r3234@fangorn: martijn | 2004-11-04 20:37:58 +0100

    LordVorp's HTTP/1.1 patch

  2005-05-04 19:50:21 (r116) by martijn
  README M; t/01_request.t M; CHANGES.OLD M; MANIFEST M; META.yml A;
  Makefile.PL M; mylib/ExtUtils/AutoInstall.pm M; CHANGES A; HTTP.pm M;
  SSL.pm M

    r5427@fangorn: martijn | 2005-05-04 16:41:46 +0200

    r3233@fangorn: martijn | 2004-11-04 20:17:31 +0100

    initial import

=========================
2004-10-02 15:37:11 v0_65
=========================

  2004-10-02 15:37:11 (r114) by rcaputo; HTTP.pm M

    Bump version to 0.65 for release. 

  2004-10-01 04:04:39 (r113) by rcaputo; HTTP.pm M

    David Davis discovered that HTTP response headers may be discarded if
    they are incomplete at the end of a read buffer. This commits a fix
    for his findings: Don't discard the remaining read buffer if we're
    still processing headers. 

=========================
2004-07-13 18:03:43 v0_64
=========================

  2004-07-13 18:03:43 (r111) by rcaputo; mylib/ExtUtils/AutoInstall.pm M

    Upgrade the included ExtUtils::AutoInstall. 

  2004-07-13 18:02:37 (r110) by rcaputo; HTTP.pm M

    Oh, duh. Increase the version number for a new release. 

  2004-07-13 18:01:56 (r109) by rcaputo; HTTP.pm M

    Matt Cashner and an unidentified AOL Employee that some call... Tim
    pointed out that SSL wasn't working. "But it was for me!" I said. So
    we hunted it down to an incorrect bitwise operator in the
    set-it-blocking code for non-ActiveState Perl versions prior to
    5.8.0. This bug TOTALLY BROKE SSL for these Perl versions, just going
    to show that you can't test too much or on too many platforms. Or
    something. 

  2004-05-02 19:42:43 (r108) by rcaputo; HTTP.pm M

    Scott Beck reported that the streaming code uses two different hash
    keys for the same heap member. This patch makes them consistent, and
    it resolves rt.cpan.org ticket #5810. 

  2004-05-02 19:25:25 (r107) by rcaputo; HTTP.pm M

    Applied Scott Beck's patch to handle interrupted status lines. This
    commit resolves rt.cpan.org ticket #5816. 

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