STATUS revision f8ce8d962c047327f15cb3a13294f4d988e87fcb
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'AmoreAPACHE 2.0 STATUS: -*-text-*-
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'AmoreLast modified at [$Date: 2002/04/09 21:28:03 $]
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.35 : released April 5, 2002 as GA.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.34 : tagged March 26, 2002.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.33 : tagged March 6, 2002. not released.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.32 : released Feburary 16, 2002 as beta.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.31 : rolled Feburary 1, 2002. not released.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.30 : tagged January 8, 2002. not rolled.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.29 : tagged November 27, 2001. not rolled.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.28 : released November 13, 2001 as beta.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.27 : rolled November 6, 2001
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.26 : tagged October 16, 2001. not rolled.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.25 : rolled August 29, 2001
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.24 : rolled August 18, 2001
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.23 : rolled August 9, 2001
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.22 : rolled July 29, 2001
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.21 : rolled July 20, 2001
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.20 : rolled July 8, 2001
2c30fa4582c5d6c659e059e719c5f6163f7ef1e3Garrett D'Amore 2.0.19 : rolled June 27, 2001
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.18 : rolled May 18, 2001
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.17 : rolled April 17, 2001
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.16 : rolled April 4, 2001
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.15 : rolled March 21, 2001
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0.14 : rolled March 7, 2001
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0a9 : released December 12, 2000
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0a8 : released November 20, 2000
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0a7 : released October 8, 2000
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0a6 : released August 18, 2000
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0a5 : released August 4, 2000
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0a4 : released June 7, 2000
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0a3 : released April 28, 2000
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0a2 : released March 31, 2000
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore 2.0a1 : released March 10, 2000
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'AmorePlease consult the following STATUS files for information
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amoreon related projects:
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'AmoreCURRENT RELEASE NOTES:
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'AmoreRELEASE SHOWSTOPPERS:
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Daedalus segfaults with 2.0.35
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Greg and Cliff are working on this.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'AmoreCURRENT VOTES:
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Should we always build [support*] binaries statically unless otherwise
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Message-ID: <20020129210006.B23512@Lithium.MeepZor.Com>
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore +1: Ken, *wrowe [they are PITAs on OSX]
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore -1: Justin, Ian
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * If the parent process dies, should the remaining child processes
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore "gracefully" self-terminate. Or maybe we should make it a runtime
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore option, or have a concept of 2 parent processes (one being a
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore "hot spare").
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore See: Message-ID: <3C58232C.FE91F19F@Golux.Com>
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Self-destruct: Ken, Martin
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Not self-destruct: BrianP, Ian, Cliff, BillS
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Make it runtime configurable: Aaron, Jim, Justin
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Have 2 parents: +1: Jim
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore -1: Justin, wrowe [for 2.0]
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore +0: Martin (while standing by, could it do
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore something useful?)
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Make the worker MPM the default MPM for threaded Unix boxes.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore +1: Justin, Ian, Cliff
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore -0: Aaron (premature decision, needs more discussion), Lars
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'AmoreRELEASE NON-SHOWSTOPPERS BUT WOULD BE REAL NICE TO WRAP THESE UP:
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * We do not properly substitute the prefix-variables in the configuration
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore scripts or generated-configs. (i.e. if sysconfdir is etc,
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore httpd-std.conf points to conf.)
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * If any request gets through ap_process_request_internal() and is
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore scheduled to be served by the core handler, without a flag that this
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore r->filename was tested by dir/file_walk, we need to 500 at the very
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore end of the ap_process_request_internal() processing so sub_req-esters
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore know this request cannot be run. This provides authors of older
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore modules better compatibility, while still improving the security and
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore robustness of 2.0.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Status: still need to decide where this goes, OtherBill comments...
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Message-ID: <065701c14526$495203b0$96c0b0d0@roweclan.net>
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore [Deleted comments regarding the ap_run_handler phase, as irrelevant
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore as BillS points out that "common case will be caught in
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore default_handler already (with the r->finfo.filetype == 0 check)"
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore and the issue is detecting this -before- we try to run the req.]
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore gregames says: can this happen somehow without a broken module
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore being involved? If not, why waste cycles trying to defend against
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore potential broken modules? It seems futile.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore wrowe counters: no, it shouldn't happen unless the module is broken.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore But the right answer is to fail the request up-front in dir/file
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore walk if the path was entirely invalid; and we can't do that either
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore or we break modules that are unwilling to hook map_to_storage.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Rewrite core_output_filter. It is nearly impossible to support
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore it with predictable results as it is implemented now.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Convert all instances of the old apr_lock_t type to the new
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * With AP_MODE_EXHAUSTIVE in the core, it is finally clear to me
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore how the Perchild MPM should be re-written. It hasn't worked
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore correctly since filters were added because it wasn't possible to
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore get the content that had already been written and the socket at
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore the same time. This mode lets us do that, so the MPM can be
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * htpasswd blindly processes the file you give it, and does no
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore sanity checking before totally corrupting whatever file it was
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore you thought you had. It should check the input file and bail
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore if it finds non-comment lines that do not contain exactly 1
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore ':' character.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Message-ID: <20020217150457.A31632@clove.org>
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Can a static httpd be built reliably?
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Message-ID: <20020207142751.T31582@clove.org>
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * [Ken] Test suite failures:
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore o worker is also failing some of the 'cgi' subtests
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore (see <URL:http://Source-Zone.Org/Apache/regression/>):
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Justin says: "Worker should be fine and passes httpd-test here.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore If you can provide evidence that it can be reproduced
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore outside of httpd-test, then it's a showstopper. I
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore think it's a perl or a httpd-test problem."
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Not a showstopper: Justin
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Usage of APR_BRIGADE_NORMALIZE in core_input_filter should be
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore removed if possible.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Message-ID: <Pine.LNX.4.33.0201202232430.318-100000@deepthought.cs.virginia.edu>
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * There is a bug in how we sort some hooks, at least the pre-config
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore hook. The first time we call the hooks, they are in the correct
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore order, but the second time, we don't sort them correctly. Currently,
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore the modules/http/config.m4 file has been renamed to
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore modules/http/config2.m4 to work around this problem, it should moved
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore back when this is fixed. rbb
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Justin says: "Is this really a showstopper? This has been here
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore forever. What's wrong? Does this have to do with
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore autoconf or m4?"
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Not a showstopper: Justin, BrianP, trawick, gregames
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * The Add...Filter and Set...Filter directives do not allow the
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore administrator to order filters, beyond the order of filename (mime)
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore extensions. It isn't clear if Set...Filter(s) should be inserted
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore before or after the Add...Filter(s) which are ordered by sequence of
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore filename extensions. At minimum, some sort of +-[0-10] syntax seems
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore like the quickest fix for a 2.0 gold release.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Justin says: "Could we delay this for a point release or 2.1?"
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Not a showstopper: justin, wrowe, trawick, stoddard, Jim, Ian, Aaron,
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Get perchild to work on platforms other than Linux. This
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore will require a portable mechanism to pass data and file/socket
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore descriptors between vhost child groups. An API was proposed
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Message-ID: <20020111115006.K1529@clove.org>
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Try to get libtool inter-library dependency code working on AIX.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Message-ID: <cm3n10lx555.fsf@rdu163-40-092.nc.rr.com>
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Justin says: If we get it working on AIX, we can enable this
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore on all platforms and clean up our build system
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Jeff says: I thought I tested a patch for you sometime in
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore January that you were going to commit within a few
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Handling of %2f in URIs. Currently both 1.3 and 2.0
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore completely disallow %2f in the request URI path (see
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore ap_unescape_url() in util.c). It's permitted and passed
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore through in the query string, however. Roy says the
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore original reason for disallowing it, from five years ago,
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore was to protect CGI scripts that applied PATH_INFO to
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore a filesystem location and which might be tricked by
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore ..%2f..%2f(...). We *should* allow path-info of the
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore form 'http://foo.com/index.cgi/path/to/path%2finfo'.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Since we've revamped a lot of our processing of path
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore segments, it would be nice to allow this, or at least
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore allow it conditionally with a directive.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * FreeBSD, threads, and worker MPM. All seems to work fine
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore if you only have one worker process with many threads. Add
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore a second worker process and the accept lock seems to be
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore lost. This might be an APR issue with how it deals with
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore the child_init hook (i.e. the fcntl lock needs to be resynced).
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore More examination and analysis is required.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Status: This has also been reported on Cygwin.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Message-ID: <3C2CC514.8EF3BED1@wapme-systems.de> (cygnus)
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Justin says: So, FreeBSD-CURRENT and Cywin have the same
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore problem. Yum. If another platform has this
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore with worker, this becomes a showstopper.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Aaron says: I spent some time disecting this and have come to
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore the conclusion that it is not a problem in the worker MPM
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore (or at least, it is not isolated to a problem in worker).
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore I'll list some of the problems I'm seeing in case someone
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore else wants to pick up where I've left off:
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore - Delivery of just about any signal to one of the child
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore processes will send it into an infinite loop as well.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore - Even though the parent is spinning out of control,
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore at first the child or children will appear to work
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore properly. At times it is possible to get it into a state,
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore however, where a request will hang until another concurrent
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore request "kicks" the first, at which point the second will
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore hang. My theory is that this has to do with the
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore pthread_cond_*() implementation in FreeBSD, but it's still
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore possible that it is in APR.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Justin adds: Oh, FreeBSD threads are implemented entirely with
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore select()/poll()/longjmp(). Welcome to the nightmare.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore So, that means a ktrace output also has the thread
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore scheduling internals in it (since it is all the same to
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore the kernel). Which makes it hard to distinguish between
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore our select() calls and their select() calls.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore *bangs head on wall repeatedly* But, some of the libc_r
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore files have a DBG_MSG #define. This is moderately helpful
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore when used with -DNO_DETACH. The kernel scheduler isn't
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore waking up the threads on a select(). Yum. And, I bet
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore those decrementing select calls have to do with the
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore scheduler. Time to brush up on our OS fundamentals.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * There is increasing demand from module writers for an API
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore that will allow them to control the server � la apachectl.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Reasons include sole-function servers that need to die if
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore an external dependency (e.g., a database) fails, et cetera.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Perhaps something in the (ever more abused) scoreboard?
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore rbb: I don't believe the scoreboard is the correct mechanism
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore for this. We already have a pipe that goes between parent
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore and child for graceful shutdown events, along with an API that
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore can be used to send a message down that pipe. In threaded MPMs,
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore it is easy enough to make that one pipe be used for graceful
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore and graceless events, and it is also easy to open that pipe
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore to both parent and child for writing. Then we just need to
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore figure out how to do graceless on non-threaded MPMs.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Allow the DocumentRoot directive within <Location > scopes? This
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore allows the beloved (crusty) Alias /foo/ /somepath/foo/ followed
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore by a <Directory /somepath/foo> to become simply
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore <Location /foo/> DocumentRoot /somefile/foo (IMHO a bit more legible
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore and in-your-face.) DocumentRoot unset would be accepted [and would
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore not permit content to be served, only virtual resources such as
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore server-info or server-status.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore This proposed change would _not_ depricate Alias.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Win32: Rotatelogs sometimes is not terminated when Apache
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore goes down hard. FirstBill was looking at possibly tracking the
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore child's-child processes in the parent process.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore OtherBill asks, wasn't this fixed?
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore stoddard: Not fixed. Shared scoreboard might offer a good
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore way for the parent to keep track of 'other child' processes
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore and whack them if the child goes down.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Other thoughts on walking the process chain using the NT kernel
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore have also been proposed on APR.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Win32: Add a simple hold console open patch (wait for close or
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore the ESC key, with a nice message) if the server died a bad
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore death (non-zero exit code) in console mode.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Resolution: bring forward same ugly hacks from 1.3.13-.20
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Port of mod_ssl to Apache 2.0:
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore The current porting state is summarized in modules/ssl/README. The
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore remaining work includes:
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore (1) stablizing/optimizing the SSL filter logic
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore (2) Enabling SSL extentions
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore (3) Trying to seperate the https filter logic from mod_ssl -
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore This is to facilitate other modules that wish to use the https
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore filter or the mod_ssl logic or both as required.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Eliminate unnecessary creation of pipes in mod_cgid
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Combine log_child and piped_log_spawn. Clean up http_log.c.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Common logging API.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Document mod_file_cache.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Platforms that do not support fork (primarily Win32 and AS/400)
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Architect start-up code that avoids initializing all the modules
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore in the parent process on platforms that do not support fork.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Win32: Migrate the MPM over to use APR thread/process calls. This
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore would eliminate some code in the Win32 branch that essentially
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore duplicates what is in APR.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * There are still a number of places in the code where we are
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore losing error status (i.e. throwing away the error returned by a
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore system call and replacing it with a generic error code)
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Mass vhosting version of suEXEC.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * All DBMs suffer from confusion in support/dbmmanage (perl script) since
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore the dbmmanage employs the first-matched dbm format. This is not
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore necessarily the library that Apache was built with. Aught to
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore rewrite dbmmanage upon installation to bin/ with the proper library
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore for predictable mod_auth_dbm administration.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Questions; htdbm exists, time to kill dbmmanage, or does it remain
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore useful as a perl dbm management example? If we keep it,
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore do we address the issue above?
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Integrate mod_dav.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Some additional items remaining:
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore - case_preserved_filename stuff
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore (use the new canonical name stuff?)
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore - find a new home for ap_text(_header)
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore - is it possible to remove the DAV: namespace stuff from util_xml?
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * ap_core_translate() and its use by mod_mmap_static and mod_file_cache
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore are a bit wonky. The function should probably be exposed as a utility
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore function (such as ap_translate_url2fs() or ap_validate_fs_url() or
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore something). Another approach would be a new hook phase after
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore "translate" which would allow the module to munge what the
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore translation has decided to do.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Status: Greg +1 (volunteers), Ryan +1
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Explore use of a post-config hook for the code in http_main.c which
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore calls ap_fixup_virutal_hosts(), ap_fini_vhost_config(), and
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore ap_sort_hooks() [to reduce the logic in main()]
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * read the config tree just once, and process N times (as necessary)
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * (possibly) use UUIDs in mod_unique_id and/or mod_usertrack
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * (possibly) port the bug fix for PR 6942 (segv when LoadModule is put
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore into a VirtualHost container) to 2.0.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * shift stuff to mod_core.h
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * callers of ap_run_create_request() should check the return value
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore for failure (Doug volunteers)
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Win32: Get Apache working on Windows 95/98. The following work
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore (at least) needs to be done:
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore - Document warning that OSR2 is required (for Crypt functions, in
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore rand.c, at least.) This could be resolved with an SSL library, or
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore randomization in APR itself.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore - Bring the Win9xConHook.dll from 1.3 into 2.0 (no sense till it
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore actually works) and add in a splash of Win9x service code.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * In order to use a DSO version of mod_ssl we have to link with
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore -lssl and -lcrypto. A workaround is in place right now where the
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore entire EXTRA_LIBS macro is being appended to the objects list, but
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore this is a hack. We should either revamp the APACHE_CHECK_SSL_TOOLKIT
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore autoconf function or come up with some other autoconf checks to
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore search for libssl and libcrypto and properly add them to mod_ssl's
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Fix the worker MPM to use POD to kill child processes instead
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore of ap_os_killpg, regardless of how they should die. (Ryan Bloom)
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * mod_cache: handle cache_control: no_cache "field_name" to enable
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore cacheing the response w/o header "field_name"
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore See RFC2616 section 14.9.1
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Scoreboard structures could be changed in the future such that
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore proper alignment is not maintained, leading to segfaults on
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore some systems. Cliff posted a patch to deal with this issue but
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore later recanted. See this message to dev@apr.apache.org:
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Message-ID: <Pine.LNX.4.44.0203011354090.16457-200000@deepthought
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'AmorePRs that have been suspended forever waiting for someone to
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amoreput them into 'the next release':
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#76: general
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore missing call to "setlocale();"
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#78: mod_include
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Additional status for XBitHack directive
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#362: mod_proxy
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Mod_proxy doesn't allow change of error pages
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#370: mod_env
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Modified PATH environemnt variable is not passed, instead
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore system's is used
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#440: mod_proxy
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Proxy doesn't deliver documents if not connected
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#534: mod_proxy
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore proxy converts ~name to %7Ename when name starts with a dot (.)
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#537: mod_access
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore mod_access syntax allows hosts that should be restricted
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#557: mod_auth-any
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore ~UserHome directories are not honored in absolute pathname
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore requests (.htaccess)
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#612: mod_proxy
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Proxy FTP Authentication Fails
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#623: mod_include
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore A smarter "Last Modified" value for SSI documents (see PR number 600)
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#628: config
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Request of "Options SymLinksIfGroupMatch"
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#700: mod_proxy
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Proxy doesn't do links right for OpenVMS files through ftp:
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#759: mod_imap
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore imap should read <MAP><AREA>*</MAP> too!
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#793: general
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore RLimitCPU and RLimitMEM don't apply to all children like they should
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#921: suexec
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Uses cwd before filling it in, doesn't use syslog
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#922: config
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore it is useful to allow specifiction that root-owned symlinks
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore should always be followed
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#980: mod_proxy
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Controlling Access to Remote Proxies would be nice...
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#994: mod_proxy
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Adding authentication "on the fly" through the proxy module
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1004: apache-api
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore request_config field in request_rec is moderately bogus
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1028: other
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore DoS attacks involving memory consumption
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1050: mod_log-any
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Logging of virtual server to error_log as well
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1085: mod_proxy
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore ProxyRemote make a dead cycle.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1117: mod_auth-any
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Using NIS passwd.byname dbm files with AuthDBMUserFile
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1120: suexec
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore suexec does not parse arguments to #exec cmd
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1145: mod_include
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Allow for Last-Modified: without resorting to XBitHack
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1158: apache-api
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore improvements to child spawning API
2c30fa4582c5d6c659e059e719c5f6163f7ef1e3Garrett D'Amore * PR#1166: mod_proxy
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore ``nph-'' not honored (no buffering) for ProxyRemote mapping
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1176: mod_cgi
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Apache cannot handle continuation line in headers
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1191: general
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore setlogin() is not called, causing problems with e.g. identd
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1204: general
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore regerror() exists, use it
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1233: apache-api
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore there is no way to keep per-connection per-module state
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1263: mod_autoexec
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Add frame-safe anchor attribute to mod_autoindex links
a702341c8e6cc834a456108a8bf5d22f031da3beGarrett D'Amore * PR#1268: suexec
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore CGI scripts running as Apache user: security (suexec etc.)
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1285: suexec
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Error messages could be easier to spot in cgi.log file for suexec.c
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1287: mod_access
a702341c8e6cc834a456108a8bf5d22f031da3beGarrett D'Amore add allow,deny/deny,allow warning to mod_access
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1290: mod_proxy
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Need to know "hit-rate" on proxy cache
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1358: mod_log-any
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Selective url-encode of log fields (or maybe a pseudo
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore log_rewrite module?)
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1383: mod_headers
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore I make mod_headers to modify request headers as well as
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore response ones.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1532: mod_proxy
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Proxy transfer logging
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1547: mod_proxy
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore No HTTP_X_FORWARDED_FOR set...
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1567: mod_proxy
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore ProxyRemote proxy requests fail authentication by firewall
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1582: mod_rewrite
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore mod_rewrite forms REQUEST_URI different than mod_cgi does
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1677: mod_headers
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore mod_headers should allow mod_log_config-style formats in
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore header values
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1702: mod_proxy
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore mod_proxy to support persistent conns?
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1803: mod_include
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore patches to mod_include to allow for file tests
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1809: mod_auth-any
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Suggestion for improving authentication modules and core source
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore code, problem with 401 and ErrorDocument
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1878: mod_proxy
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore listing of proxy cache content
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#1905: suexec
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Allow modules to set user:group for execution.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2024: apache-api
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore adding auth_why to conn_rec
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2073: mod_log-any
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore pipelined connections are not logged correctly
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2074: mod_rewrite
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore mod_rewrite doesn't pass Proxy Throughput on internal subrequests
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2113: config
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore HTTP Server Rebuild Line Needs Changing for the better
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2138: mod_status
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore mod_status always displays 256 possible connection slots
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2221: documentation
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Make online documentation search link back to my installation
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2284: general
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Can not POST to ErrorDocument - Apache/1.3b6
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2314: mod_proxy
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore patterns in ProxyRemote
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2343: mod_status
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Status module averages are for entire uptime
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2360: suexec
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore suexec for general access of user content?
260256302ac564fcfed9ebfa5a6c98a2e6ef3808Garrett D'Amore * PR#2396: general
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Proposal for TimeZone directive
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2415: mod_info
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore /server-info doesn't check for the virtual host to list the info
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2421: config
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore problem specifying ndbm library for build ?with autoconfigure
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2431: general
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore A small addition to rotatelogs.c to improve program functionality.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2446: config
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore AllowOverride FileInfo is too coarse
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2460: mod_cgi
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore TimeOut applies to output of CGI scripts
260256302ac564fcfed9ebfa5a6c98a2e6ef3808Garrett D'Amore * PR#2512: mod_access
260256302ac564fcfed9ebfa5a6c98a2e6ef3808Garrett D'Amore <IfDenied> directive wanted
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2573: suexec
260256302ac564fcfed9ebfa5a6c98a2e6ef3808Garrett D'Amore CGI's for general use still have to be run as another user
260256302ac564fcfed9ebfa5a6c98a2e6ef3808Garrett D'Amore * PR#2648: general
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Cache file names in Proxy module
260256302ac564fcfed9ebfa5a6c98a2e6ef3808Garrett D'Amore * PR#2760: config
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore [PATCH] User/Group for <Directory> and <Location> i.e. not only
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore in global and <Virtual>.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2763: general
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore mailto tags and bundling bug report script
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2785: os-aix
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Support for System Resource Controller
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2793: protocol
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore When will Apache support P3P? Any Plans?
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2873: config
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2889: general
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Inclusion of RPM spec file in CVS/distributions
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2906: general
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Propose that Apache recommend $UNIQUE_ID for all "session id"
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#2907: config
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore suggestion: power up your Include directive :)
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#3018: general
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore cannot limit some HTTP methods
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#3143: apache-api
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore No module specific data hook for per-connection data
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#3191: mod_negotiation
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore no way to set global quality-of-source (qs) coneg values
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore with multiviews
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#3568: mod_proxy
260256302ac564fcfed9ebfa5a6c98a2e6ef3808Garrett D'Amore Accessing URL through proxy server corrupts data.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#3605: mod_proxy
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Some anonymous FTP URLs ask for authentication
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#3677: general
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore New ErrorDocumentMatch directive
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#4241: config
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Need to be able to override shebang line to make CGI scripts
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore more portable.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#4244: config
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore "Files" and "FilesMatch" regexp does not recognize bang as
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore negation operator
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#4448: mod_log-any
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Please allow CGI env variables (QUERY_STRING, ...) to be logged
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#4459: mod_include
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Suggestion for better handling of Last-modified headers
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#4490: mod_cgi
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore mod_cgi prevents handling of OPTIONS requests
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#5713: os-windows
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore [PATCH] install as win32 service with domain account
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Status: Cannot accept password-as-arg, we should prompt the
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore user when -k install/-k config with a user argument.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * PR#5993: general
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore AllowOverride should have a 'CheckNone' and 'AllowNone' argument
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore instead of only 'None'
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'AmoreOther bugs that need fixing:
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * MaxRequestsPerChild measures connections, not requests.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Until someone has a better way, we'll probably just rename it
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore "MaxConnectionsPerChild".
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Regex containers don't work in an intutive way
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Status: No one has come up with an efficient way to fix this
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore behavior. Dean has suggested getting rid of regex containers
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore OtherBill suggests: We at least seem to agree on eliminating
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore the <Container ~ foo> forms, and using only
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore <ContainerMatch foo> semantics.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * SIGSEGV on Linux (glibc 2.1.2) isn't caught properly by a
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore sigwaiting thread. We need to work around this, perhaps unless
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore there is hope soon for a fixed glibc.
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * orig_ct in the byterange/multipart handling may not be
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore needed. Apache 1.3 just never stashed "multipart" into
260256302ac564fcfed9ebfa5a6c98a2e6ef3808Garrett D'Amore r->content_type. We should probably follow suit since the
260256302ac564fcfed9ebfa5a6c98a2e6ef3808Garrett D'Amore byterange stuff doesn't want the rest of the code to see the
260256302ac564fcfed9ebfa5a6c98a2e6ef3808Garrett D'Amore multipart content-type; the other code should still think it is
260256302ac564fcfed9ebfa5a6c98a2e6ef3808Garrett D'Amore dealing with the <orig_ct> stuff.
260256302ac564fcfed9ebfa5a6c98a2e6ef3808Garrett D'Amore Status: Greg volunteers to investigate (esp. since he was most
260256302ac564fcfed9ebfa5a6c98a2e6ef3808Garrett D'Amore likely the one to break it :-)
260256302ac564fcfed9ebfa5a6c98a2e6ef3808Garrett D'AmoreBinaries (2.0.35):
260256302ac564fcfed9ebfa5a6c98a2e6ef3808Garrett D'Amore Platform Avail. Volunteer
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore ------------------------------------------------------------------
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore AIX 4.3.3 Bill Stoddard
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Mandrake 8.1 no Ryan Bloom
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore FreeBSD 4.1 yes Ryan Bloom
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore i386-unknown-freebsd4.5 yes Aaron Bannert
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore OS X 10.1.3/Darwin 5.3 yes Jim Jagielski
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Solaris 8.x/sparc yes Jim Jagielski
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore i686-pc-linux-gnu-rh70 yes Aaron Bannert
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore i686-pc-linux-gnu-rh72 yes Aaron Bannert
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore i386-pc-solaris2.8 yes Aaron Bannert
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore powerpc-unknown-linux-gnu yes Graham Leggett
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore NetWare no Brad Nicholes
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Win32-x86 no William Rowe
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'AmoreOther features that need writing:
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Finish infrastructure in core for async MPMs
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Status: post 2.0
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * TODO in source -- just do an egrep on "TODO" and see what's there
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'AmoreAvailable Patches:
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Martin Sojka <msojka@gmx.de>'s patch to add error reporting for failed
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore htpasswd actions due to a full /tmp volume (other programs may have
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore similar problems?)
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Mike Abbott's <mja@trudge.engr.sgi.com> patches to improve
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore Status: These were written for 1.3, and are awaiting a port to
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Jim Winstead's <jimw@trainedmonkey.com> patch to add CookieDomain and
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore other small mod_usertrack features
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Dan Rench's <drench@xnet.com> patch to add allow the errmsg and timefmt
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore of SSI's to be modified in the config file. Patch is available in
88447a05f537aabe9a1bc3d5313f22581ec992a7Garrett D'Amore * Which MPMs will be included with Apache 2.0?