A add much more useful select(2) implementation than is provided by
windows sockets. The winsock implementation will only work with sockets;
our implementation works with sockets and file de

A add much more useful select(2) implementation than is provided by
windows sockets. The winsock implementation will only work with sockets;
our implementation works with sockets and file descriptors.
By association, stream_select() will now operate correctly with files, pipes and sockets.

This change required linking against the winsock2 library. In terms of
compatibility, only older versions of windows 95 do not have winsock2
installed by default. It is available as a redistributable file, and is most likely installed by any OS patches (eg: Internet Explorer) applied by the user.

Also, add a win32 compatible pipe test when opening a stream from a pipe. This test will only work on NT, win2k and XP platforms. Without this test, interleaved fread() and select() calls would cause the read buffer to be clobbered. I will be working on a fix for this issue for win9x.

show more ...

Fixed zlib.output_compression so it can work even if zlib extension is built as shared

Removed pointless memory allocation checks.

Bump year.

Fix for #21304 (gzopen did not report errors on open; "a" mode broken)

Fix ZTS build.

Fixed bug #21228 (broken check for ob_gzhandler).
Fixed a bug that made ob_start return incorrect return value.

Comment about maximal length now matches code.

Merge streams changes from branch.

Fixed OnUpdate_zlib_output_compression() behaviour

Move wrong output buffer usage check to ob_gzhandler init.
Export some output buffer functions.

Revisted Wez patch: chunk_size 0 means cahce the whole output. So
we must apply the default before calling php_enable_output_compression().
I have left the default setting in the rinit functi

Revisted Wez patch: chunk_size 0 means cahce the whole output. So
we must apply the default before calling php_enable_output_compression().
I have left the default setting in the rinit function even though i do think
it is not necessary.

show more ...

Fix infinite recursion bug when using zlib output compression.
Cause: the chunk size is taken from the zlib.output_compression setting,
which is 0 or 1. This causes the block_size for output

Fix infinite recursion bug when using zlib output compression.
Cause: the chunk size is taken from the zlib.output_compression setting,
which is 0 or 1. This causes the block_size for output buffer to be set
to 0 (1 / 2) and thus causes infinite recursion in php_ob_allocate().
Solution: use a value of 0 for the chunk size which will use the default
sizes. Also add a sanity check which will default the block_size to 1
if it ends up as 0.

show more ...

Use zend_ini_string instead of the 'alias'.

Fix ob_gzhandler()'s handling of requests w/o compression but that do have
the Accept-Encoding header

Fix error messages.

Fixed prototype.
Fixed double printing of function name in docref error messages.

change to php_error_docref()
#documentation about output relevant ini settings will be in ref.outcontrol

Changed handling of output buffer sizes.

Changed my mind.
Let users decided whether users want larger chunk size or not.
@ Changed zlib.output_compression behavior. If larger chunk is preferred,
enable output_buffering also. (

Changed my mind.
Let users decided whether users want larger chunk size or not.
@ Changed zlib.output_compression behavior. If larger chunk is preferred,
enable output_buffering also. (Marcus, Yasuo)

show more ...

Generally speaking, sending packets less than 1KB is not efficient.
We're better to take some benchmarks see if removing extra buffer
is actually a good idea or not.

corrected handling of internal output buffers
minor fixs

no need to start additional first buffer when output compression is first buffer

should fix all windows warnings

Improved handling of output buffers (see news)\n#No trim for the string parameter...