Log Message:
Update 0.22 to 0.26
-------------------
0.26 2015/02/10 08:30:00
  - The deprecialted SSL_Advanced option now causes new() to call croak.
    I'll remove any remaining reference to it in v0.27 or v0.28.
  - The POD text no longer mentions the SSL_Advanced option at all.
  - Added to README the new environment variable created for the test scripts
    in the previous release.
  - The RFC links in the SEE ALSO section had all expired.  Located replacement
    links for all referenced documents.  Looks like they were just moved to
    another location on the same web site.  A victim of web-site remodeling.
  - Enhanced _debug_print_hash().  It originally worked with just GLOBs, but
    now works with regular hashes as well.  Also fixed to follow SCALAR
    references.
  - Now dumps to the log file the contents of the SSL_Client_Certificate
    hash if provided.  Was difficult to debug logic when it wasn't logged.
  - Bug Id # 101388, reusing SSL_Client_Certificate options on the data channel.
    Added additional options to go with SSL_reuse_ctx to resolve this issue.
    I checked the IO-Socket-SSL docs & all variables in the bug id are by
    default part of SSL_reuse_ctx.  So some low level code isn't resolving
    the context correctly.  This is a work arround.  (Certificates were not
    being used here.  Just using the context for strict hostname verification.)
    May be a case that when certificates are not used, the context isn't
    fully used either.

0.25 2014/09/05 08:30:00
  WARNING: This is a major upgrade & may not be 100% backwards compatible with
           existing code.  Also the response() redesign may not work for all
           FTPS servers!  It's a work in progress!  So comments are welcome!
  - Upgraded the minimum version of IO-Socket-SSL required from v1.08 to v1.26
    so that I could remove the warnings in the POD for ccc().  But it should
    still work with the earlier versions if you hack the code.  But don't rely
    on this in future releases.  (minus the ccc command)
  - Modified the Synopsis to use Croak since no one was reading my comments
    about using it too closely anyway.
  - Added a new pattern for uput() to check for file names with.
    Also fixed to guarentee it will never return any path info, just the
    base file name itself.  [ This change may break existing programs. ]
  - Another hack to recover from a garbled CCC response.  This hack just
    prevents unnecessary warnings.
  - Hit a weird FTP server that requiers me to flip flop the binary/ASCII
    settings on the server/client.  See mixedModeAI() & mixedModeIA().
  - Added a way to print Perl warnings to the log file. (trapWarn)  It will
    chain things if warnings were already trapped.  Even between multiple open
    Net::FTPSSL object logs.  This is only usefull as a debugging tool when
    reporting on errors via CPAN.  It helps provide context to the warnings
    that affect this code base.
  - Added get_log_filehandle() to allow someone to gain access to the open
    filehandle used to write to the log file generated when Debug is turned on
    and you specified a log file via DebugLogFile.
  - Now allows DebugLogFile to be an open file handle (GLOB).
  - Fixed quot() to echo the command if it's one of those dissallowed.
    Also now strips off any leading spaces from the command before using it.
  - Found out use "sub DESTROY" rather than "sub END" for \ 
objects. (Just a FYI)
    Surprise, Surprise!  Sometimes END is called before DESTROY is!
  - Moved some logic out of quit() into the new DESTROY() method.  No longer
    need to hack object termination logic via quit().
  - Added END to handle final clean up for trapWarn().
  - Fixed warnings in function _feat() caused by strange server returns.
  - Fixed warnings in function _help() caused by strange server returns.
  - Fixed 20-certificate.t bug where $ENV{HOME} is not defined for all OS.
  - Fixed 10-complex.t to use the new warning logic.
  - Fixed 20-certificate.t to use the new warning logic.
  - Fixed t/10-complex.t & t/20-certificate.t to use a 30 second Timeout
    instead of using the default 2 minutes.
  - Reworked response() with regard to Bug # 97608. (related to Bug # 73115)
    Had to redesign how this method worked.  It was getting too convoluted
    to patch any further.  Was the only way to properly fix _help & _feat.
  - Second issue for Bug # 9706 was an issue with very long login messsages
    hit an unexpected Timeout issue (last release enhancement).  Turns out the
    Timeout logic in response() didn't always work properly if you didn't read
    in the entire response via a single call to sysread().  Had two options,
    make the buffer huge for the login response or make the logic more complex,
    to only check for Timeouts on the command channel for new commands.  Turns
    out select() & sysread() don't actually compare notes.  Hense another reason
    to redesign how things worked.
  - This rewrite of response() broke the ccc() hack.  So had to redesign how
    this hack worked as well.  [May break existing programs using CCC!]
  - Added special case Debug=>99 for more detailed logging for debugging
    the new response code base in the future.
  - Added new environment variable (FTPSSL_DEBUG_LEVEL) to t/10-complex.t to
    enable low level debugging of the new response() code.  There is no prompt
    to turn on this feature.  Added more as a reminder that it can be done!
  - Updated the SSL_Advanced depreciated warning that it will soon be removed
    in a future release.  It was depreciated in v0.18 way back in 2011, so I
    feel it's about time to have it removed.  Just giving fair warning here!
    I'll remove it in v0.26 or v0.27.

0.24 2014/06/30 08:30:00
  - Makefile.PL fix, only asks question if Net::HTTPTunel isn't present!
    Also now defaults to "N" instead of "Y" if module not \ 
present.
  - Fixed bug in supported().  Unsupported commands are followed with "*"
    in help.  Most do "CMD*", but found some servers doing "CMD \ 
*" instead.
  - Fixed bug in 20-certificate.t where we were getting a false failure
    on nlst().  Fixed to explicitly check the response code instead of
    assuming an error when nothing was found.  Fixed list() as well.
  - Same fix in 10-complex.t, even though tests designed to always return
    something.
  - Added transfer() to transfer files from one system to another system
    without the need to make a local copy of the file.
  - Added xtransfer() as well.
  - Added transfer & xtransfer validation to 10-complex.t to test things out.
  - Added a new environment variable default for *.t prompts.
  - Bug # 95411 - Patch provided by ian@iansramblings.com.  Fixes hang issue
    when the command channel unexpectedly drops.  Does this by implementing
    a timeout in response().
    #-------  (Should I do this for the many data channel sysreads as well?)

0.23 2013/08/01 08:30:00
  - Added clarification to use of SSL_Client_Certificate option in pod text.
  - Added Bug Id 82094 to support tunneling through a proxy server via the
    ProxyArgs option.  Supports both proxy servers that require a password as
    well as those that don't.  This option requires Net::HTTPTunnel to use.
    But this new module isn't required if you don't use tunneling.
    (perl Makefile.PL will ask.)
  - Added optional proxy server questions if Net::HTTPTunnel is installed for
    the test scripts (t/*.t).  It warns you if it's missing during the test.
  - Fixed bug in login() so the call to help() only happens if the login is
    successfull.
  - Bug Id 84376 uncoverd bug in IO-Socket-SSL versions 1.79 to 1.85 when
    using Certificats for the Data Channel & getting a Man-In-The-Middle (MITM)
    warning that didn't apply.  So skip those releases.
  - Bug Id 85420 same MITM warning as the other bug, but this time without using
    Certificates in the command channel.  This required a code change in new()
    to fix.
  - Fixed masking issue when writing to the log file when the value that needed
    to be masked contained regex special chars.
  - Added clarification that any warnings printed below the 10-complex.t tests
    are just that.  Warnings, not errors preventing things from working.  They
    are collected so the developer can be notified about them for fixing in
    furture releases.  They are also written to the end of the log now as well.