STATUS revision bbe8c2e5ffecf3f8a6a697a184b7173c173b6b4a
5bd562b1d7da51cb5715899d32bb4c79c54459b0wroweAPACHE 2.0 STATUS: -*-text-*-
5bd562b1d7da51cb5715899d32bb4c79c54459b0wroweLast modified at [$Date: 2002/03/22 19:35:47 $]
2a6e98ba4ffa30ded5d8831664c5cb2a170a56b6coar 2.0.34 : in development
9a11fa4e07f50f2e5750d078ef3751ddbf441b8ftrawick 2.0.33 : tagged March 6, 2002.
9a11fa4e07f50f2e5750d078ef3751ddbf441b8ftrawick 2.0.32 : released Feburary 16, 2002.
9a11fa4e07f50f2e5750d078ef3751ddbf441b8ftrawick 2.0.31 : rolled Feburary 1, 2002. not released.
1f279dc92a60df9f61bf58468162aab0eef072e4brianp 2.0.30 : tagged January 8, 2002. not rolled.
1f279dc92a60df9f61bf58468162aab0eef072e4brianp 2.0.29 : tagged November 27, 2001. not rolled.
1f279dc92a60df9f61bf58468162aab0eef072e4brianp 2.0.28 : released November 13, 2001
1f279dc92a60df9f61bf58468162aab0eef072e4brianp 2.0.27 : rolled November 6, 2001
2fd0edbd8b2f47a8458322bedd3b82f825faf336trawick 2.0.26 : tagged October 16, 2001. not rolled.
2fd0edbd8b2f47a8458322bedd3b82f825faf336trawick 2.0.25 : rolled August 29, 2001
2fd0edbd8b2f47a8458322bedd3b82f825faf336trawick 2.0.24 : rolled August 18, 2001
39021cf8b495cdb94013ca73531ccb32658fb793rederpj 2.0.23 : rolled August 9, 2001
39021cf8b495cdb94013ca73531ccb32658fb793rederpj 2.0.22 : rolled July 29, 2001
39021cf8b495cdb94013ca73531ccb32658fb793rederpj 2.0.21 : rolled July 20, 2001
39021cf8b495cdb94013ca73531ccb32658fb793rederpj 2.0.20 : rolled July 8, 2001
39021cf8b495cdb94013ca73531ccb32658fb793rederpj 2.0.19 : rolled June 27, 2001
39021cf8b495cdb94013ca73531ccb32658fb793rederpj 2.0.18 : rolled May 18, 2001
39021cf8b495cdb94013ca73531ccb32658fb793rederpj 2.0.17 : rolled April 17, 2001
39021cf8b495cdb94013ca73531ccb32658fb793rederpj 2.0.16 : rolled April 4, 2001
fa16f10bc02e46bc5a6c2c2c6371926cd1dbe2edwrowe 2.0.15 : rolled March 21, 2001
fa16f10bc02e46bc5a6c2c2c6371926cd1dbe2edwrowe 2.0.14 : rolled March 7, 2001
fa16f10bc02e46bc5a6c2c2c6371926cd1dbe2edwrowe 2.0a9 : released December 12, 2000
fa16f10bc02e46bc5a6c2c2c6371926cd1dbe2edwrowe 2.0a8 : released November 20, 2000
fa16f10bc02e46bc5a6c2c2c6371926cd1dbe2edwrowe 2.0a7 : released October 8, 2000
fa16f10bc02e46bc5a6c2c2c6371926cd1dbe2edwrowe 2.0a6 : released August 18, 2000
fa16f10bc02e46bc5a6c2c2c6371926cd1dbe2edwrowe 2.0a5 : released August 4, 2000
fa16f10bc02e46bc5a6c2c2c6371926cd1dbe2edwrowe 2.0a4 : released June 7, 2000
fa16f10bc02e46bc5a6c2c2c6371926cd1dbe2edwrowe 2.0a3 : released April 28, 2000
fa16f10bc02e46bc5a6c2c2c6371926cd1dbe2edwrowe 2.0a2 : released March 31, 2000
db8ac7cbb1fa6cdd6abcc4bb797d4deed32dd269jim 2.0a1 : released March 10, 2000
db8ac7cbb1fa6cdd6abcc4bb797d4deed32dd269jimPlease consult the following STATUS files for information
db8ac7cbb1fa6cdd6abcc4bb797d4deed32dd269jimon related projects:
7b0a3bcc0e689305df49f7d4da7abc35aa891862brianpCURRENT RELEASE NOTES:
7b0a3bcc0e689305df49f7d4da7abc35aa891862brianp * 34 status: Let's get all API changes and showstoppers in this one.
1bae4591a85d90325ecdacedf7e54d1bbfe31037aaronFINAL RELEASE SHOWSTOPPERS:
a6f48cc01ab8f5377e570c61826dcdfc36741936trawick * API changes planned for 2.0 that should happen before the
a6f48cc01ab8f5377e570c61826dcdfc36741936trawick GA release:
a6f48cc01ab8f5377e570c61826dcdfc36741936trawick * Free lists for bucket allocation
2da345202997f8f5860c801d68f7913c02fc05fctrawick * We do not properly substitute the prefix-variables in the configuration
2da345202997f8f5860c801d68f7913c02fc05fctrawick scripts or generated-configs. (i.e. if sysconfdir is etc,
5bd562b1d7da51cb5715899d32bb4c79c54459b0wrowe httpd-std.conf points to conf.)
5bd562b1d7da51cb5715899d32bb4c79c54459b0wrowe Aaron says: This is not a showstopper, these problems have existed
5bd562b1d7da51cb5715899d32bb4c79c54459b0wrowe for as long as I can remember. It would be nice to fix
5bd562b1d7da51cb5715899d32bb4c79c54459b0wrowe them but they are not new.
5bd562b1d7da51cb5715899d32bb4c79c54459b0wrowe Not a showstopper: Jeff, Aaron, BillS, gregames
1c06e98017400874d5ff6ad79f13145ec4589225strikerCURRENT VOTES:
99a041314eb3db0b0cca325c9c40d0a6c5fdf04fwrowe * Should we always build [support*] binaries statically unless otherwise
99a041314eb3db0b0cca325c9c40d0a6c5fdf04fwrowe Message-ID: <20020129210006.B23512@Lithium.MeepZor.Com>
99a041314eb3db0b0cca325c9c40d0a6c5fdf04fwrowe +1: Ken, *wrowe [they are PITAs on OSX]
99a041314eb3db0b0cca325c9c40d0a6c5fdf04fwrowe -1: Justin, Ian
99a041314eb3db0b0cca325c9c40d0a6c5fdf04fwrowe * If the parent process dies, should the remaining child processes
99a041314eb3db0b0cca325c9c40d0a6c5fdf04fwrowe "gracefully" self-terminate. Or maybe we should make it a runtime
6e119e632566d69798ce6cf4e714ed374b72914frederpj option, or have a concept of 2 parent processes (one being a
6e119e632566d69798ce6cf4e714ed374b72914frederpj "hot spare").
6e119e632566d69798ce6cf4e714ed374b72914frederpj See: Message-ID: <3C58232C.FE91F19F@Golux.Com>
6e119e632566d69798ce6cf4e714ed374b72914frederpj Self-destruct: Ken, Martin
bd120542ebe7e09cdbada5daf4924f4690e5ece3trawick Not self-destruct: BrianP, Ian, Cliff, BillS
bd120542ebe7e09cdbada5daf4924f4690e5ece3trawick Make it runtime configurable: Aaron, Jim, Justin
bd120542ebe7e09cdbada5daf4924f4690e5ece3trawick Have 2 parents: +1: Jim
bd120542ebe7e09cdbada5daf4924f4690e5ece3trawick -1: Justin, wrowe [for 2.0]
bd120542ebe7e09cdbada5daf4924f4690e5ece3trawick +0: Martin (while standing by, could it do
68d439bc0482b2e41053480f748edc2574c2ea7btrawick something useful?)
68d439bc0482b2e41053480f748edc2574c2ea7btrawick * Make the worker MPM the default MPM for threaded Unix boxes.
68d439bc0482b2e41053480f748edc2574c2ea7btrawick +1: Justin, Ian, Cliff
dddbde8480d265d06c84f2281f01e00f8ef52e94mjc -0: Aaron (premature decision, needs more discussion), Lars
dddbde8480d265d06c84f2281f01e00f8ef52e94mjcRELEASE NON-SHOWSTOPPERS BUT WOULD BE REAL NICE TO WRAP THESE UP:
e06675c51d084791089d79c3ac18aeae8dd465fcrbb * If any request gets through ap_process_request_internal() and is
e06675c51d084791089d79c3ac18aeae8dd465fcrbb scheduled to be served by the core handler, without a flag that this
e06675c51d084791089d79c3ac18aeae8dd465fcrbb r->filename was tested by dir/file_walk, we need to 500 at the very
e06675c51d084791089d79c3ac18aeae8dd465fcrbb end of the ap_process_request_internal() processing so sub_req-esters
481c1206b6065a8f37ab75ca1fc26c947cb37852ianh know this request cannot be run. This provides authors of older
481c1206b6065a8f37ab75ca1fc26c947cb37852ianh modules better compatibility, while still improving the security and
481c1206b6065a8f37ab75ca1fc26c947cb37852ianh robustness of 2.0.
a964f7434f5c7f512a5fa0d0178260ccb74c84berbb Status: still need to decide where this goes, OtherBill comments...
a964f7434f5c7f512a5fa0d0178260ccb74c84berbb Message-ID: <065701c14526$495203b0$96c0b0d0@roweclan.net>
a964f7434f5c7f512a5fa0d0178260ccb74c84berbb [Deleted comments regarding the ap_run_handler phase, as irrelevant
6a7877447bcb8e6ff848d72f82f184c404ef4c0bminfrin as BillS points out that "common case will be caught in
6a7877447bcb8e6ff848d72f82f184c404ef4c0bminfrin default_handler already (with the r->finfo.filetype == 0 check)"
6a7877447bcb8e6ff848d72f82f184c404ef4c0bminfrin and the issue is detecting this -before- we try to run the req.]
9335cbd541cca1ca6038af329bbd1645310aabccminfrin gregames says: can this happen somehow without a broken module
9335cbd541cca1ca6038af329bbd1645310aabccminfrin being involved? If not, why waste cycles trying to defend against
9335cbd541cca1ca6038af329bbd1645310aabccminfrin potential broken modules? It seems futile.
9335cbd541cca1ca6038af329bbd1645310aabccminfrin wrowe counters: no, it shouldn't happen unless the module is broken.
9335cbd541cca1ca6038af329bbd1645310aabccminfrin But the right answer is to fail the request up-front in dir/file
9335cbd541cca1ca6038af329bbd1645310aabccminfrin walk if the path was entirely invalid; and we can't do that either
9335cbd541cca1ca6038af329bbd1645310aabccminfrin or we break modules that are unwilling to hook map_to_storage.
e156db58351d1c040bc72430f3eb072cb6ae7107brianp * Rewrite core_output_filter. It is nearly impossible to support
e156db58351d1c040bc72430f3eb072cb6ae7107brianp it with predictable results as it is implemented now.
1c06e98017400874d5ff6ad79f13145ec4589225striker * Convert all instances of the old apr_lock_t type to the new
1c06e98017400874d5ff6ad79f13145ec4589225striker Status: Aaron has converted all but the perchild MPM to
eb1349e4ab58bd2935f7054a1bfc5c86ab5a5fa3striker use the new lock API. Since perchild has been put
eb1349e4ab58bd2935f7054a1bfc5c86ab5a5fa3striker on the back burner, this is no longer a showstopper.
6a94da925498a20a09fde0a66002607be8d83b1astriker Aaron will patch perchild as soon as it becomes
6a94da925498a20a09fde0a66002607be8d83b1astriker buildable again.
7639aa8b39e0d9dbd096f9cc3379bcd3d5e4003bstriker * With AP_MODE_EXHAUSTIVE in the core, it is finally clear to me
8dc5aa056a586ffa920a6ecd5c31048702371ea6brianp how the Perchild MPM should be re-written. It hasn't worked
4c9d27bfdfea41b388dc705f7cc2b49318ab5344jim correctly since filters were added because it wasn't possible to
4c9d27bfdfea41b388dc705f7cc2b49318ab5344jim get the content that had already been written and the socket at
4c9d27bfdfea41b388dc705f7cc2b49318ab5344jim the same time. This mode lets us do that, so the MPM can be
e8e8ab3cbc3d90f15eb78e094c381a6e908fd6efjerenkrantz * htpasswd blindly processes the file you give it, and does no
f4c472b8dce3c2e559232dbb5b27ed2466922ea4jerenkrantz sanity checking before totally corrupting whatever file it was
f4c472b8dce3c2e559232dbb5b27ed2466922ea4jerenkrantz you thought you had. It should check the input file and bail
8dc5aa056a586ffa920a6ecd5c31048702371ea6brianp if it finds non-comment lines that do not contain exactly 1
8dc5aa056a586ffa920a6ecd5c31048702371ea6brianp ':' character.
8dc5aa056a586ffa920a6ecd5c31048702371ea6brianp Message-ID: <20020217150457.A31632@clove.org>
7e31ef4870c7ef94838585004405e8854fefcc51ianh * Can a static httpd be built reliably?
7e31ef4870c7ef94838585004405e8854fefcc51ianh Message-ID: <20020207142751.T31582@clove.org>
7e31ef4870c7ef94838585004405e8854fefcc51ianh * [Ken] Test suite failures:
bd496a3a7752a55c849e62ed00cacc492d4f6d3erederpj o worker is also failing some of the 'cgi' subtests
bd496a3a7752a55c849e62ed00cacc492d4f6d3erederpj (see <URL:http://Source-Zone.Org/Apache/regression/>):
bd496a3a7752a55c849e62ed00cacc492d4f6d3erederpj Justin says: "Worker should be fine and passes httpd-test here.
bd496a3a7752a55c849e62ed00cacc492d4f6d3erederpj If you can provide evidence that it can be reproduced
bd496a3a7752a55c849e62ed00cacc492d4f6d3erederpj outside of httpd-test, then it's a showstopper. I
a8c401eadf77822e851f19c7740e7ec6dca03daastoddard think it's a perl or a httpd-test problem."
a8c401eadf77822e851f19c7740e7ec6dca03daastoddard Not a showstopper: Justin
a8c401eadf77822e851f19c7740e7ec6dca03daastoddard * Usage of APR_BRIGADE_NORMALIZE in core_input_filter should be
93d7153aa172665f55b04463b831ad556269c3efbrianp removed if possible.
791781f2ccc1f1f1bc1b1643861d3da23edfd147jerenkrantz Message-ID: <Pine.LNX.4.33.0201202232430.318-100000@deepthought.cs.virginia.edu>
2ffa5829cad36f12b0a1fc3481592e85bc210a5bjerenkrantz * There is a bug in how we sort some hooks, at least the pre-config
2ffa5829cad36f12b0a1fc3481592e85bc210a5bjerenkrantz hook. The first time we call the hooks, they are in the correct
2ffa5829cad36f12b0a1fc3481592e85bc210a5bjerenkrantz order, but the second time, we don't sort them correctly. Currently,
02e8590d904653a95eca31bdf8e60866642bb592slive the modules/http/config.m4 file has been renamed to
02e8590d904653a95eca31bdf8e60866642bb592slive modules/http/config2.m4 to work around this problem, it should moved
30a5ee06ae5f7f67a83d6852871f654ae3e14aadslive back when this is fixed. rbb
02e8590d904653a95eca31bdf8e60866642bb592slive Justin says: "Is this really a showstopper? This has been here
a97c4ce2bea3dbf8ddc82c796cf93aa6b46765a4brianp forever. What's wrong? Does this have to do with
a97c4ce2bea3dbf8ddc82c796cf93aa6b46765a4brianp autoconf or m4?"
a97c4ce2bea3dbf8ddc82c796cf93aa6b46765a4brianp Not a showstopper: Justin, BrianP, trawick, gregames
b7838ae85a698af19d90ba4ebe03e10bdc149eacjerenkrantz * The Add...Filter and Set...Filter directives do not allow the
435d2db95b905b0d16d35410e18ff77dc39688aabrianp administrator to order filters, beyond the order of filename (mime)
435d2db95b905b0d16d35410e18ff77dc39688aabrianp extensions. It isn't clear if Set...Filter(s) should be inserted
435d2db95b905b0d16d35410e18ff77dc39688aabrianp before or after the Add...Filter(s) which are ordered by sequence of
1a7728c0205d607d5d87c6b6bf1b8837a9785a99ianh filename extensions. At minimum, some sort of +-[0-10] syntax seems
a574815e2c6db68b9d8139db89921ededf033decianh like the quickest fix for a 2.0 gold release.
a574815e2c6db68b9d8139db89921ededf033decianh Justin says: "Could we delay this for a point release or 2.1?"
964f539e766a3301b3e2f767baeffddcf9f6092bjerenkrantz Not a showstopper: justin, wrowe, trawick, stoddard, Jim, Ian, Aaron,
5e538c6ced13aa2f7c358e1a44f651d31dd5fab8brianp * Get perchild to work on platforms other than Linux. This
5e538c6ced13aa2f7c358e1a44f651d31dd5fab8brianp will require a portable mechanism to pass data and file/socket
5e538c6ced13aa2f7c358e1a44f651d31dd5fab8brianp descriptors between vhost child groups. An API was proposed
5e538c6ced13aa2f7c358e1a44f651d31dd5fab8brianp on dev@apr:
61202a45487668abad788c02e339f626176e645fianh Message-ID: <20020111115006.K1529@clove.org>
61202a45487668abad788c02e339f626176e645fianh * CGI single-byte reads
02c49e1f35a4d1a171df2d319e76af0c5163dc4dmartin BrianP suggests that this is caused by the ap_scan_script_header_err()
02c49e1f35a4d1a171df2d319e76af0c5163dc4dmartin routine, which will do single-byte reads until it finds the end
02c49e1f35a4d1a171df2d319e76af0c5163dc4dmartin of the header, at which point it constructs a pipe-bucket (buffered)
6deb8bcfb8511ac38243a8274fc589842841b398ianh to read from.
6deb8bcfb8511ac38243a8274fc589842841b398ianh Proposed solution in:
6deb8bcfb8511ac38243a8274fc589842841b398ianh Message-ID: <3C36ADAF.60601@cnet.com>
49bbbd1939208be54a3eb00b95e61d90d180a606ianh * Try to get libtool inter-library dependency code working on AIX.
93d7153aa172665f55b04463b831ad556269c3efbrianp Message-ID: <cm3n10lx555.fsf@rdu163-40-092.nc.rr.com>
93d7153aa172665f55b04463b831ad556269c3efbrianp Justin says: If we get it working on AIX, we can enable this
93d7153aa172665f55b04463b831ad556269c3efbrianp on all platforms and clean up our build system
bc6600a6207e5d15b895294e370e4e3320a803d8stoddard Jeff says: I thought I tested a patch for you sometime in
bc6600a6207e5d15b895294e370e4e3320a803d8stoddard January that you were going to commit within a few
bc6600a6207e5d15b895294e370e4e3320a803d8stoddard * Handling of %2f in URIs. Currently both 1.3 and 2.0
bc6600a6207e5d15b895294e370e4e3320a803d8stoddard completely disallow %2f in the request URI path (see
4e21f1a207aefa9796dc758bf7274b0f1fea780fstoddard ap_unescape_url() in util.c). It's permitted and passed
7a23067e782dd5612d4d4b539906e1733b664df7jwoolley through in the query string, however. Roy says the
5c214a63f9722864ac4983995da11353779515dbrederpj original reason for disallowing it, from five years ago,
5c214a63f9722864ac4983995da11353779515dbrederpj was to protect CGI scripts that applied PATH_INFO to
5c214a63f9722864ac4983995da11353779515dbrederpj a filesystem location and which might be tricked by
5c214a63f9722864ac4983995da11353779515dbrederpj ..%2f..%2f(...). We *should* allow path-info of the
9f20717d827f2113a23dfa45539813171cf626eaianh Since we've revamped a lot of our processing of path
9f20717d827f2113a23dfa45539813171cf626eaianh segments, it would be nice to allow this, or at least
9f20717d827f2113a23dfa45539813171cf626eaianh allow it conditionally with a directive.
a0db2f093595083300ad3438314f90921405ccf9wrowe * FreeBSD, threads, and worker MPM. All seems to work fine
a0db2f093595083300ad3438314f90921405ccf9wrowe if you only have one worker process with many threads. Add
c66798efb2184ecf904cd8471acd17e65688b1caianh a second worker process and the accept lock seems to be
c66798efb2184ecf904cd8471acd17e65688b1caianh lost. This might be an APR issue with how it deals with
fb59f85aab19883025f619727948b8088232cc4brederpj the child_init hook (i.e. the fcntl lock needs to be resynced).
fb59f85aab19883025f619727948b8088232cc4brederpj More examination and analysis is required.
fb59f85aab19883025f619727948b8088232cc4brederpj Status: This has also been reported on Cygwin.
fb59f85aab19883025f619727948b8088232cc4brederpj Message-ID: <3C2CC514.8EF3BED1@wapme-systems.de> (cygnus)
1a1cf0ee9229ee29e5750b25dd94dbb9b04072cfianh Justin says: So, FreeBSD-CURRENT and Cywin have the same
1a1cf0ee9229ee29e5750b25dd94dbb9b04072cfianh problem. Yum. If another platform has this
1a1cf0ee9229ee29e5750b25dd94dbb9b04072cfianh with worker, this becomes a showstopper.
61afed048a4d67ed923d52e5c865c0f10a8e9e73trawick Aaron says: I spent some time disecting this and have come to
0bbb249eafe9ef9508821f0ef58e7440625ecd62trawick the conclusion that it is not a problem in the worker MPM
0bbb249eafe9ef9508821f0ef58e7440625ecd62trawick (or at least, it is not isolated to a problem in worker).
61afed048a4d67ed923d52e5c865c0f10a8e9e73trawick I'll list some of the problems I'm seeing in case someone
7a23067e782dd5612d4d4b539906e1733b664df7jwoolley else wants to pick up where I've left off:
7a23067e782dd5612d4d4b539906e1733b664df7jwoolley - Delivery of just about any signal to one of the child
7a23067e782dd5612d4d4b539906e1733b664df7jwoolley processes will send it into an infinite loop as well.
7a23067e782dd5612d4d4b539906e1733b664df7jwoolley - Even though the parent is spinning out of control,
6032a7c97a25c52f4bdd78ce23f2010e52c9e81arederpj at first the child or children will appear to work
6032a7c97a25c52f4bdd78ce23f2010e52c9e81arederpj properly. At times it is possible to get it into a state,
6032a7c97a25c52f4bdd78ce23f2010e52c9e81arederpj however, where a request will hang until another concurrent
ba2e14e474516f1c75a96b4f6d1a9dec332175efianh request "kicks" the first, at which point the second will
ba2e14e474516f1c75a96b4f6d1a9dec332175efianh hang. My theory is that this has to do with the
ec69fc6e323eb1f3112966e06e9e37be608d052cianh pthread_cond_*() implementation in FreeBSD, but it's still
f0791c5bdfd36969d292a4092df076aa6d1c34ccwrowe possible that it is in APR.
f0791c5bdfd36969d292a4092df076aa6d1c34ccwrowe Justin adds: Oh, FreeBSD threads are implemented entirely with
f0791c5bdfd36969d292a4092df076aa6d1c34ccwrowe select()/poll()/longjmp(). Welcome to the nightmare.
749011213737e8d0cd6ca78d5eb532ec6f6b9fdfianh So, that means a ktrace output also has the thread
749011213737e8d0cd6ca78d5eb532ec6f6b9fdfianh scheduling internals in it (since it is all the same to
749011213737e8d0cd6ca78d5eb532ec6f6b9fdfianh the kernel). Which makes it hard to distinguish between
ec69fc6e323eb1f3112966e06e9e37be608d052cianh our select() calls and their select() calls.
ec69fc6e323eb1f3112966e06e9e37be608d052cianh *bangs head on wall repeatedly* But, some of the libc_r
e7bf4d6f15d04e86e20002e65f60d7fbf80e5974stoddard files have a DBG_MSG #define. This is moderately helpful
e7bf4d6f15d04e86e20002e65f60d7fbf80e5974stoddard when used with -DNO_DETACH. The kernel scheduler isn't
8ab933f1df663f95c27e2ce5772127d4f3a10e0bstriker waking up the threads on a select(). Yum. And, I bet
de42d3dfd83a4cc62f0dd6b79ee5cbcfa69fd503brianp those decrementing select calls have to do with the
de42d3dfd83a4cc62f0dd6b79ee5cbcfa69fd503brianp scheduler. Time to brush up on our OS fundamentals.
de42d3dfd83a4cc62f0dd6b79ee5cbcfa69fd503brianp * There is increasing demand from module writers for an API
8ab933f1df663f95c27e2ce5772127d4f3a10e0bstriker that will allow them to control the server � la apachectl.
a8dda281113c5038945423320d8c9b42e3d1ddb1jwoolley Reasons include sole-function servers that need to die if
50e60f30bdc074fbc887f0b98f4d570457ac97c9brianp an external dependency (e.g., a database) fails, et cetera.
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe Perhaps something in the (ever more abused) scoreboard?
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe rbb: I don't believe the scoreboard is the correct mechanism
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe for this. We already have a pipe that goes between parent
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe and child for graceful shutdown events, along with an API that
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe can be used to send a message down that pipe. In threaded MPMs,
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe it is easy enough to make that one pipe be used for graceful
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe and graceless events, and it is also easy to open that pipe
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe to both parent and child for writing. Then we just need to
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe figure out how to do graceless on non-threaded MPMs.
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe * Allow the DocumentRoot directive within <Location > scopes? This
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe allows the beloved (crusty) Alias /foo/ /somepath/foo/ followed
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe by a <Directory /somepath/foo> to become simply
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe <Location /foo/> DocumentRoot /somefile/foo (IMHO a bit more legible
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe and in-your-face.) DocumentRoot unset would be accepted [and would
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe not permit content to be served, only virtual resources such as
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe server-info or server-status.
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe This proposed change would _not_ depricate Alias.
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe * Win32: Rotatelogs sometimes is not terminated when Apache
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe goes down hard. FirstBill was looking at possibly tracking the
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe child's-child processes in the parent process.
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe OtherBill asks, wasn't this fixed?
39dde7f4cd79d701cc14e5beac8ea528bc58d038wrowe stoddard: Not fixed. Shared scoreboard might offer a good
268ac122b1fd6fa948b30bdf0d8c0d80e75d68dawrowe way for the parent to keep track of 'other child' processes
268ac122b1fd6fa948b30bdf0d8c0d80e75d68dawrowe and whack them if the child goes down.
268ac122b1fd6fa948b30bdf0d8c0d80e75d68dawrowe Other thoughts on walking the process chain using the NT kernel
268ac122b1fd6fa948b30bdf0d8c0d80e75d68dawrowe have also been proposed on APR.
268ac122b1fd6fa948b30bdf0d8c0d80e75d68dawrowe * Win32: Add a simple hold console open patch (wait for close or
6ba861fd6c705eaeb1f9bb97df86ddea6895e263minfrin the ESC key, with a nice message) if the server died a bad
6ba861fd6c705eaeb1f9bb97df86ddea6895e263minfrin death (non-zero exit code) in console mode.
6ba861fd6c705eaeb1f9bb97df86ddea6895e263minfrin Resolution: bring forward same ugly hacks from 1.3.13-.20
b78ed256f4b99e72836d36fd68d4e7a26dbe032cianh * Port of mod_ssl to Apache 2.0:
b78ed256f4b99e72836d36fd68d4e7a26dbe032cianh The current porting state is summarized in modules/ssl/README. The
698670444b30b79e808155739f98c39bec35f72awrowe remaining work includes:
e7ec1c54206901c9369e40f471b71836c78e017dwrowe (1) stablizing/optimizing the SSL filter logic
698670444b30b79e808155739f98c39bec35f72awrowe (2) Enabling SSL extentions
57bea0f0559e31536af3b7b5859d3681ee29a34cwrowe (3) Trying to seperate the https filter logic from mod_ssl -
57bea0f0559e31536af3b7b5859d3681ee29a34cwrowe This is to facilitate other modules that wish to use the https
57bea0f0559e31536af3b7b5859d3681ee29a34cwrowe filter or the mod_ssl logic or both as required.
57bea0f0559e31536af3b7b5859d3681ee29a34cwrowe * Performance & Debug: Eliminate most (and perhaps all) of the
57bea0f0559e31536af3b7b5859d3681ee29a34cwrowe malloc/free calls in the bucket brigade code. Need some
94e2b2d12fa269af16fa63a6270d3336d9f126f2trawick light weight memory management functions that allow freeing
94e2b2d12fa269af16fa63a6270d3336d9f126f2trawick memory (putting it back into a memory pool) when it is no
94e2b2d12fa269af16fa63a6270d3336d9f126f2trawick longer needed. Enabling simple debugging features like guard
94e2b2d12fa269af16fa63a6270d3336d9f126f2trawick bands, double free detection, etc. would be cool but certainly
35313c8d7368125c3e95d3118238d2be9a613000trawick not a hard requirement.
35313c8d7368125c3e95d3118238d2be9a613000trawick Status: Cliff started to implement this using SMS as has
55da18d54a0ba74dc51aecba5b0daf71a2ed10a7trawick been discussed at length for months, but since
55da18d54a0ba74dc51aecba5b0daf71a2ed10a7trawick SMS is not being used anywhere else in the server,
55da18d54a0ba74dc51aecba5b0daf71a2ed10a7trawick several people expressed the opinion that we should
55da18d54a0ba74dc51aecba5b0daf71a2ed10a7trawick get rid of it entirely, meaning that the buckets
55da18d54a0ba74dc51aecba5b0daf71a2ed10a7trawick need their own memory management (free list) functions.
c51f2b89da23e3371959a74808dee1792d96f5c1wsanchez Cliff will implement that this weekend so we at least
c51f2b89da23e3371959a74808dee1792d96f5c1wsanchez have something to look at/compare with.
c51f2b89da23e3371959a74808dee1792d96f5c1wsanchez * Eliminate unnecessary creation of pipes in mod_cgid
79c9b0ac498d97336874edba0daf9f544ad14671trawick * Combine log_child and piped_log_spawn. Clean up http_log.c.
79c9b0ac498d97336874edba0daf9f544ad14671trawick Common logging API.
79c9b0ac498d97336874edba0daf9f544ad14671trawick * Document mod_file_cache.
5a7d934619b2be92e18be5dd3366f4ac6ddeab43trawick * Platforms that do not support fork (primarily Win32 and AS/400)
5a7d934619b2be92e18be5dd3366f4ac6ddeab43trawick Architect start-up code that avoids initializing all the modules
5a7d934619b2be92e18be5dd3366f4ac6ddeab43trawick in the parent process on platforms that do not support fork.
5a70e5b66eb7758d0e64e070211f699fc83fca70wrowe * Win32: Migrate the MPM over to use APR thread/process calls. This
5a70e5b66eb7758d0e64e070211f699fc83fca70wrowe would eliminate some code in the Win32 branch that essentially
5a70e5b66eb7758d0e64e070211f699fc83fca70wrowe duplicates what is in APR.
84eeb0ab12215fc22577a9a0a9589cea2a445712trawick * There are still a number of places in the code where we are
84eeb0ab12215fc22577a9a0a9589cea2a445712trawick losing error status (i.e. throwing away the error returned by a
1d3fbd2d9f03c0826977d940a2081401edf522d4jerenkrantz system call and replacing it with a generic error code)
b5cc0253789825ace46944dc9cde744be08dd77fjerenkrantz * Mass vhosting version of suEXEC.
e4bb84f3c11f282d3ba66f64940b1b8e13f85e7aslive * All DBMs suffer from confusion in support/dbmmanage (perl script) since
e4bb84f3c11f282d3ba66f64940b1b8e13f85e7aslive the dbmmanage employs the first-matched dbm format. This is not
e4bb84f3c11f282d3ba66f64940b1b8e13f85e7aslive necessarily the library that Apache was built with. Aught to
e4bb84f3c11f282d3ba66f64940b1b8e13f85e7aslive rewrite dbmmanage upon installation to bin/ with the proper library
e4bb84f3c11f282d3ba66f64940b1b8e13f85e7aslive for predictable mod_auth_dbm administration.
e4bb84f3c11f282d3ba66f64940b1b8e13f85e7aslive Questions; htdbm exists, time to kill dbmmanage, or does it remain
e4bb84f3c11f282d3ba66f64940b1b8e13f85e7aslive useful as a perl dbm management example? If we keep it,
ba2bab42e97405dc41c0f8fe3416f7f9a79ed7a9brianp do we address the issue above?
ba2bab42e97405dc41c0f8fe3416f7f9a79ed7a9brianp * Integrate mod_dav.
756b54396a86db555817bb52149d91b60d00e35fwrowe Some additional items remaining:
756b54396a86db555817bb52149d91b60d00e35fwrowe - case_preserved_filename stuff
756b54396a86db555817bb52149d91b60d00e35fwrowe (use the new canonical name stuff?)
756b54396a86db555817bb52149d91b60d00e35fwrowe - find a new home for ap_text(_header)
b4251d1fbef86f96e01c68f8de086e0dbb8bcb74trawick - is it possible to remove the DAV: namespace stuff from util_xml?
b4251d1fbef86f96e01c68f8de086e0dbb8bcb74trawick * ap_core_translate() and its use by mod_mmap_static and mod_file_cache
24efed0910118b762a4eb84830875d4714b8d315ianh are a bit wonky. The function should probably be exposed as a utility
24efed0910118b762a4eb84830875d4714b8d315ianh function (such as ap_translate_url2fs() or ap_validate_fs_url() or
50e60f30bdc074fbc887f0b98f4d570457ac97c9brianp something). Another approach would be a new hook phase after
50e60f30bdc074fbc887f0b98f4d570457ac97c9brianp "translate" which would allow the module to munge what the
50e60f30bdc074fbc887f0b98f4d570457ac97c9brianp translation has decided to do.
bdbafc44d060509e86f0cc56ff4d19579438f846striker Status: Greg +1 (volunteers), Ryan +1
bdbafc44d060509e86f0cc56ff4d19579438f846striker * Explore use of a post-config hook for the code in http_main.c which
fbd0c3dbae333ba4a7225dad2d090419ad894e4ctrawick calls ap_fixup_virutal_hosts(), ap_fini_vhost_config(), and
bdbafc44d060509e86f0cc56ff4d19579438f846striker ap_sort_hooks() [to reduce the logic in main()]
6a82dfd37385024d0e94e71edd2f46b609796cfdwrowe * read the config tree just once, and process N times (as necessary)
6d0ec39a3ef89ce485f23008efa399b7b35bf1fdjwoolley * (possibly) use UUIDs in mod_unique_id and/or mod_usertrack
6d0ec39a3ef89ce485f23008efa399b7b35bf1fdjwoolley * (possibly) port the bug fix for PR 6942 (segv when LoadModule is put
0fdf8c342123fde84405b885fb1720ebc652e10djerenkrantz into a VirtualHost container) to 2.0.
0fdf8c342123fde84405b885fb1720ebc652e10djerenkrantz * shift stuff to mod_core.h
0fdf8c342123fde84405b885fb1720ebc652e10djerenkrantz * callers of ap_run_create_request() should check the return value
6b87b6eee6a43f40ef6bead9ef3173979b4cd76crbb for failure (Doug volunteers)
6b87b6eee6a43f40ef6bead9ef3173979b4cd76crbb * Win32: Get Apache working on Windows 95/98. The following work
6b87b6eee6a43f40ef6bead9ef3173979b4cd76crbb (at least) needs to be done:
6b87b6eee6a43f40ef6bead9ef3173979b4cd76crbb - Document warning that OSR2 is required (for Crypt functions, in
bfd2cedbf2918fcb95daa9f850ecdf5e24765c22jerenkrantz rand.c, at least.) This could be resolved with an SSL library, or
bfd2cedbf2918fcb95daa9f850ecdf5e24765c22jerenkrantz randomization in APR itself.
bfd2cedbf2918fcb95daa9f850ecdf5e24765c22jerenkrantz - Bring the Win9xConHook.dll from 1.3 into 2.0 (no sense till it
f9a773d26994c3b267589e404cdb5b760f83e888jerenkrantz actually works) and add in a splash of Win9x service code.
f9a773d26994c3b267589e404cdb5b760f83e888jerenkrantz * In order to use a DSO version of mod_ssl we have to link with
a250599aab6669d5877edf158032efd2538e5820trawick -lssl and -lcrypto. A workaround is in place right now where the
a250599aab6669d5877edf158032efd2538e5820trawick entire EXTRA_LIBS macro is being appended to the objects list, but
a250599aab6669d5877edf158032efd2538e5820trawick this is a hack. We should either revamp the APACHE_CHECK_SSL_TOOLKIT
a250599aab6669d5877edf158032efd2538e5820trawick autoconf function or come up with some other autoconf checks to
41338e6ead3fa8d60ad3841d069f4b47e71d9177wrowe search for libssl and libcrypto and properly add them to mod_ssl's
41338e6ead3fa8d60ad3841d069f4b47e71d9177wrowe link flags.
41338e6ead3fa8d60ad3841d069f4b47e71d9177wrowe * Fix the worker MPM to use POD to kill child processes instead
92b0ffb9cbc04b3d9c7ce6becadc0c3d88dea2d9wrowe of ap_os_killpg, regardless of how they should die. (Ryan Bloom)
92b0ffb9cbc04b3d9c7ce6becadc0c3d88dea2d9wrowe * mod_cache: handle cache_control: no_cache "field_name" to enable
961ff00a8f1fe79a8ac8b18617b40a404e28cb35brianp cacheing the response w/o header "field_name"
961ff00a8f1fe79a8ac8b18617b40a404e28cb35brianp See RFC2616 section 14.9.1
92b0ffb9cbc04b3d9c7ce6becadc0c3d88dea2d9wrowe * Scoreboard structures could be changed in the future such that
6a82dfd37385024d0e94e71edd2f46b609796cfdwrowe proper alignment is not maintained, leading to segfaults on
6a82dfd37385024d0e94e71edd2f46b609796cfdwrowe some systems. Cliff posted a patch to deal with this issue but
6a82dfd37385024d0e94e71edd2f46b609796cfdwrowe later recanted. See this message to dev@apr.apache.org:
6a82dfd37385024d0e94e71edd2f46b609796cfdwrowe Message-ID: <Pine.LNX.4.44.0203011354090.16457-200000@deepthought
6a82dfd37385024d0e94e71edd2f46b609796cfdwrowePRs that have been suspended forever waiting for someone to
6a82dfd37385024d0e94e71edd2f46b609796cfdwroweput them into 'the next release':
c43fd8f8f90a7549bffe1e581eedbd087db1163estoddard * PR#76: general
c43fd8f8f90a7549bffe1e581eedbd087db1163estoddard missing call to "setlocale();"
854cc4d3451547c2359c27870a3c354ad385a49bianh * PR#78: mod_include
854cc4d3451547c2359c27870a3c354ad385a49bianh Additional status for XBitHack directive
02ec77ed8e15b4b601de98a322e4bd8d7d3e1ec2trawick * PR#362: mod_proxy
49ada1eac7c4cae429ba193273b7f40f355d9c7ejwoolley Mod_proxy doesn't allow change of error pages
88425bd3442321915195ac9dfa9a80ffcd968fa4brianp * PR#370: mod_env
88425bd3442321915195ac9dfa9a80ffcd968fa4brianp Modified PATH environemnt variable is not passed, instead
88425bd3442321915195ac9dfa9a80ffcd968fa4brianp system's is used
a8dda281113c5038945423320d8c9b42e3d1ddb1jwoolley * PR#440: mod_proxy
a8dda281113c5038945423320d8c9b42e3d1ddb1jwoolley Proxy doesn't deliver documents if not connected
37b8494ffaeb4ee9a9a2f9917d334078c16d4212jwoolley * PR#534: mod_proxy
bf3d1782a29630335a1df535eb395355ab1cd154jwoolley proxy converts ~name to %7Ename when name starts with a dot (.)
da16bea08c6ff10ceb8d250ff23e8e81a372cef8jwoolley * PR#537: mod_access
da16bea08c6ff10ceb8d250ff23e8e81a372cef8jwoolley mod_access syntax allows hosts that should be restricted
99f692732327e0c200fd639105dbf9940bd229f1rbb * PR#557: mod_auth-any
80f73246cc14f02d50bfac5306c079464c2dd1c6rbb ~UserHome directories are not honored in absolute pathname
80f73246cc14f02d50bfac5306c079464c2dd1c6rbb requests (.htaccess)
80f73246cc14f02d50bfac5306c079464c2dd1c6rbb * PR#612: mod_proxy
dcdc78fce34f06533df4829abbc726f7fbf207fejwoolley Proxy FTP Authentication Fails
d4a93d608a28bf331625544a2896fa20bef4a2b4rbb * PR#623: mod_include
d4a93d608a28bf331625544a2896fa20bef4a2b4rbb A smarter "Last Modified" value for SSI documents (see PR number 600)
d4a93d608a28bf331625544a2896fa20bef4a2b4rbb * PR#628: config
36fcd3d96b9bf9a2d4af424e64584b5dede3e3e6brianp Request of "Options SymLinksIfGroupMatch"
4b34d6a5b70303010612df6c87da3ee91ae86078rbb * PR#700: mod_proxy
4b34d6a5b70303010612df6c87da3ee91ae86078rbb Proxy doesn't do links right for OpenVMS files through ftp:
4b34d6a5b70303010612df6c87da3ee91ae86078rbb * PR#759: mod_imap
4b34d6a5b70303010612df6c87da3ee91ae86078rbb imap should read <MAP><AREA>*</MAP> too!
84bdb86d57d2a2f828b17e77ac2379fed551c2adtrawick * PR#793: general
84bdb86d57d2a2f828b17e77ac2379fed551c2adtrawick RLimitCPU and RLimitMEM don't apply to all children like they should
46603605c2edcc1cc84fa45634e19a395134078atrawick * PR#921: suexec
46603605c2edcc1cc84fa45634e19a395134078atrawick Uses cwd before filling it in, doesn't use syslog
86826d685f83170ca07d56550db9f0c2922a916btrawick * PR#922: config
86826d685f83170ca07d56550db9f0c2922a916btrawick it is useful to allow specifiction that root-owned symlinks
86826d685f83170ca07d56550db9f0c2922a916btrawick should always be followed
4f412c60e9c2af999619d11b236068a0e0e94944trawick * PR#980: mod_proxy
4f412c60e9c2af999619d11b236068a0e0e94944trawick Controlling Access to Remote Proxies would be nice...
4f6effa17a5084085c9104b0bb97c2ba1622cfa6jerenkrantz * PR#994: mod_proxy
4f6effa17a5084085c9104b0bb97c2ba1622cfa6jerenkrantz Adding authentication "on the fly" through the proxy module
bfa5c37259833629155d486bb1571d39a57de64dbnicholes * PR#1004: apache-api
bfa5c37259833629155d486bb1571d39a57de64dbnicholes request_config field in request_rec is moderately bogus
da16bea08c6ff10ceb8d250ff23e8e81a372cef8jwoolley * PR#1028: other
c88a70518d7d0b5bccb22a68d446c2d6589790dcjwoolley DoS attacks involving memory consumption
e59e4b703b7e19c4b35030e4baac8a96a8d4b504dougm * PR#1050: mod_log-any
e59e4b703b7e19c4b35030e4baac8a96a8d4b504dougm Logging of virtual server to error_log as well
5717c6b0b97a065a84fba32cebeee959a5fe4f15dougm * PR#1085: mod_proxy
5717c6b0b97a065a84fba32cebeee959a5fe4f15dougm ProxyRemote make a dead cycle.
5717c6b0b97a065a84fba32cebeee959a5fe4f15dougm * PR#1117: mod_auth-any
835836eaf9e2a23192a262307b08f626e50e2180trawick Using NIS passwd.byname dbm files with AuthDBMUserFile
81dddb023f9dd43b350f782972c1f75a88a2d93ftrawick * PR#1120: suexec
81dddb023f9dd43b350f782972c1f75a88a2d93ftrawick suexec does not parse arguments to #exec cmd
9b9e0eca165f5f464e357bb2a9b8bbfc9621067cwrowe * PR#1145: mod_include
9b9e0eca165f5f464e357bb2a9b8bbfc9621067cwrowe Allow for Last-Modified: without resorting to XBitHack
9b9e0eca165f5f464e357bb2a9b8bbfc9621067cwrowe * PR#1158: apache-api
b26781e595625911fc8fc8215133ad2285ed75d8jim improvements to child spawning API
5117466ef123b1efbc2feba168f37069ef6f230bianh * PR#1166: mod_proxy
5117466ef123b1efbc2feba168f37069ef6f230bianh ``nph-'' not honored (no buffering) for ProxyRemote mapping
9c39f8fb982df4dbce5304e49385568e6d35bfa8trawick * PR#1176: mod_cgi
9c39f8fb982df4dbce5304e49385568e6d35bfa8trawick Apache cannot handle continuation line in headers
bdd4aa13a97de79596cd19708f1516e8fa92700ewrowe * PR#1191: general
bdd4aa13a97de79596cd19708f1516e8fa92700ewrowe setlogin() is not called, causing problems with e.g. identd
1d50c90ddb7e3d144ec8a2bd848ca1e7bbf8e534bnicholes * PR#1204: general
bdd4aa13a97de79596cd19708f1516e8fa92700ewrowe regerror() exists, use it
1d50c90ddb7e3d144ec8a2bd848ca1e7bbf8e534bnicholes * PR#1233: apache-api
24e361af20a3107dc934b4895911ce6bcce0603ejwoolley there is no way to keep per-connection per-module state
4657f9b12af4b123b80e15c73fa03c190e47a8bftrawick * PR#1263: mod_autoexec
4657f9b12af4b123b80e15c73fa03c190e47a8bftrawick Add frame-safe anchor attribute to mod_autoindex links
c36bac9a918f59b2dbf5dcd7d67b50c1da04c89drbb * PR#1268: suexec
c36bac9a918f59b2dbf5dcd7d67b50c1da04c89drbb CGI scripts running as Apache user: security (suexec etc.)
c36bac9a918f59b2dbf5dcd7d67b50c1da04c89drbb * PR#1285: suexec
c36bac9a918f59b2dbf5dcd7d67b50c1da04c89drbb Error messages could be easier to spot in cgi.log file for suexec.c
ad668861e40da485f2eea24dc4c1f9940e470698rbb * PR#1287: mod_access
ad668861e40da485f2eea24dc4c1f9940e470698rbb add allow,deny/deny,allow warning to mod_access
89211a5d592cdf0170d2b541946661b1a2a279c5trawick * PR#1290: mod_proxy
89211a5d592cdf0170d2b541946661b1a2a279c5trawick Need to know "hit-rate" on proxy cache
5caa0a5c428439b566a4fcc711747e2053bcfd1ajerenkrantz * PR#1358: mod_log-any
34d672a81f3e72f30568462135ddf6d71dcfa8d8bnicholes Selective url-encode of log fields (or maybe a pseudo
34d672a81f3e72f30568462135ddf6d71dcfa8d8bnicholes log_rewrite module?)
b1d2d2797866636f792717f96401292481697145wrowe * PR#1383: mod_headers
b1d2d2797866636f792717f96401292481697145wrowe I make mod_headers to modify request headers as well as
b1d2d2797866636f792717f96401292481697145wrowe response ones.
b1d2d2797866636f792717f96401292481697145wrowe * PR#1532: mod_proxy
b1d2d2797866636f792717f96401292481697145wrowe Proxy transfer logging
4a98c07ecf4f27a9b18963bbe85260857664d03bjerenkrantz * PR#1547: mod_proxy
8abcc73436888a98721b10f0c09206f6fea68c55jerenkrantz No HTTP_X_FORWARDED_FOR set...
b24d065530fdf97376f390522396be5a4469fcf4jerenkrantz * PR#1567: mod_proxy
a981bcb9c3b780184e75ff069d67fe84da6c3d3dbnicholes ProxyRemote proxy requests fail authentication by firewall
a981bcb9c3b780184e75ff069d67fe84da6c3d3dbnicholes * PR#1582: mod_rewrite
a981bcb9c3b780184e75ff069d67fe84da6c3d3dbnicholes mod_rewrite forms REQUEST_URI different than mod_cgi does
b2cff333bc23b8e74c6aad9ee97973df02cca180aaron * PR#1677: mod_headers
b2cff333bc23b8e74c6aad9ee97973df02cca180aaron mod_headers should allow mod_log_config-style formats in
7eb55be5bcc75f2acf789aeca38d88a9c75d001ejwoolley header values
d31e50dddb975275ef388ab64e380b6d4e5d9710jerenkrantz * PR#1702: mod_proxy
d31e50dddb975275ef388ab64e380b6d4e5d9710jerenkrantz mod_proxy to support persistent conns?
d31e50dddb975275ef388ab64e380b6d4e5d9710jerenkrantz * PR#1803: mod_include
d5eaaee4a1e5faaf21e7111fd61732c6e7dbe8b2jwoolley patches to mod_include to allow for file tests
c88a70518d7d0b5bccb22a68d446c2d6589790dcjwoolley * PR#1809: mod_auth-any
c88a70518d7d0b5bccb22a68d446c2d6589790dcjwoolley Suggestion for improving authentication modules and core source
c88a70518d7d0b5bccb22a68d446c2d6589790dcjwoolley code, problem with 401 and ErrorDocument
c64c364bf863ad985309ef10d68caaa93e8d09ccstoddard * PR#1878: mod_proxy
c64c364bf863ad985309ef10d68caaa93e8d09ccstoddard listing of proxy cache content
d8d240df2f2b23455be6b01343daedebaa6c4f96trawick * PR#1905: suexec
d8d240df2f2b23455be6b01343daedebaa6c4f96trawick Allow modules to set user:group for execution.
b5b5e8cc4668ab29d8f08f590d829dcfaeda9d33brianp * PR#2024: apache-api
b5b5e8cc4668ab29d8f08f590d829dcfaeda9d33brianp adding auth_why to conn_rec
e2653756d0d14a9a620b24bd04a6ab1182178462brianp * PR#2073: mod_log-any
e2653756d0d14a9a620b24bd04a6ab1182178462brianp pipelined connections are not logged correctly
d980a97cfc68de0ef35432ce03f1c1e468a19877jwoolley * PR#2074: mod_rewrite
d980a97cfc68de0ef35432ce03f1c1e468a19877jwoolley mod_rewrite doesn't pass Proxy Throughput on internal subrequests
aa3510b82ec5d82ddbf4748829ec90f1ce71398ebrianp * PR#2113: config
aa3510b82ec5d82ddbf4748829ec90f1ce71398ebrianp HTTP Server Rebuild Line Needs Changing for the better
51b1d7f8eaa74807ab14479edde4421e77f5d1d7brianp * PR#2138: mod_status
08aff55373b2ae69182a58055a5c1b3a12d927b0slive mod_status always displays 256 possible connection slots
4f50bfb0367b91396c0fe85b80536b760080d39etrawick * PR#2221: documentation
4f50bfb0367b91396c0fe85b80536b760080d39etrawick Make online documentation search link back to my installation
5a63340978acb9dd7e87724be57d2bde1cf1f629trawick * PR#2284: general
5a63340978acb9dd7e87724be57d2bde1cf1f629trawick Can not POST to ErrorDocument - Apache/1.3b6
8565f742f1ee3d514b2d48a1f2a5af5d7437c3b9trawick * PR#2314: mod_proxy
8565f742f1ee3d514b2d48a1f2a5af5d7437c3b9trawick patterns in ProxyRemote
8843b75d1c70af3da9a7306c4aede3b3e9346deajwoolley * PR#2343: mod_status
8843b75d1c70af3da9a7306c4aede3b3e9346deajwoolley Status module averages are for entire uptime
dcecfbe7e63603d0fb95e4811e015039e9401990stoddard * PR#2360: suexec
dcecfbe7e63603d0fb95e4811e015039e9401990stoddard suexec for general access of user content?
4844d314b0b6e55a309621b84c1786446c5fb418ianh * PR#2396: general
4844d314b0b6e55a309621b84c1786446c5fb418ianh Proposal for TimeZone directive
127c81d6b2c9ba8932273c2e1b13e3ea3d673b07trawick * PR#2415: mod_info
127c81d6b2c9ba8932273c2e1b13e3ea3d673b07trawick /server-info doesn't check for the virtual host to list the info
59d01b9ed1de043745262e49bb44bde7095c4723jwoolley * PR#2421: config
59d01b9ed1de043745262e49bb44bde7095c4723jwoolley problem specifying ndbm library for build ?with autoconfigure
e6e65585927961caf45d4e9e932bb1f4e9e89ca1jerenkrantz * PR#2431: general
e6e65585927961caf45d4e9e932bb1f4e9e89ca1jerenkrantz A small addition to rotatelogs.c to improve program functionality.
b068b71651c802cc4e0b835495ad0e41c65e1174trawick * PR#2446: config
34a1e566d4af2735c87e2d3e6ac5e505855518aawrowe AllowOverride FileInfo is too coarse
34a1e566d4af2735c87e2d3e6ac5e505855518aawrowe * PR#2460: mod_cgi
34a1e566d4af2735c87e2d3e6ac5e505855518aawrowe TimeOut applies to output of CGI scripts
34a1e566d4af2735c87e2d3e6ac5e505855518aawrowe * PR#2512: mod_access
34a1e566d4af2735c87e2d3e6ac5e505855518aawrowe <IfDenied> directive wanted
34a1e566d4af2735c87e2d3e6ac5e505855518aawrowe * PR#2573: suexec
34a1e566d4af2735c87e2d3e6ac5e505855518aawrowe CGI's for general use still have to be run as another user
34a1e566d4af2735c87e2d3e6ac5e505855518aawrowe with suExec
34a1e566d4af2735c87e2d3e6ac5e505855518aawrowe * PR#2648: general
34a1e566d4af2735c87e2d3e6ac5e505855518aawrowe Cache file names in Proxy module
dd39efc1f7ed97cf526aefa24359f0be2ac5c3f4trawick * PR#2760: config
dd39efc1f7ed97cf526aefa24359f0be2ac5c3f4trawick [PATCH] User/Group for <Directory> and <Location> i.e. not only
dd39efc1f7ed97cf526aefa24359f0be2ac5c3f4trawick in global and <Virtual>.
f53367106769f90696d9c1f0ffcf9fbb4db883c2jerenkrantz * PR#2763: general
9a940e8e90f9b163737e23bbf72a3f2c67a39220brianp mailto tags and bundling bug report script
8496c88debb9962575dac2b1ef9b81984d7bd759brianp * PR#2785: os-aix
8496c88debb9962575dac2b1ef9b81984d7bd759brianp Support for System Resource Controller
7c8747b339a5e47ea8301907051a9974d15b23b7brianp * PR#2793: protocol
7c8747b339a5e47ea8301907051a9974d15b23b7brianp When will Apache support P3P? Any Plans?
0b493ad526b8bbc2ff21ce19510fc32238abb816jwoolley * PR#2873: config
1376737cb6afa24d3e12f3a223318fe1bd71bb1fslive * PR#2889: general
1376737cb6afa24d3e12f3a223318fe1bd71bb1fslive Inclusion of RPM spec file in CVS/distributions
ec1719a5748717f67dcd279bb64bd0da424ae450jerenkrantz * PR#2906: general
ec1719a5748717f67dcd279bb64bd0da424ae450jerenkrantz Propose that Apache recommend $UNIQUE_ID for all "session id"
3f0220bf6e864d982a10348c9cc269bfe798d65eaaron * PR#2907: config
3f0220bf6e864d982a10348c9cc269bfe798d65eaaron suggestion: power up your Include directive :)
75d133afee2e3636b07366fd62102b13b67b1b1djerenkrantz * PR#3018: general
75d133afee2e3636b07366fd62102b13b67b1b1djerenkrantz cannot limit some HTTP methods
2ae7ad6f58b7e3188ed4bd375a8930808529ba71jwoolley * PR#3143: apache-api
2ae7ad6f58b7e3188ed4bd375a8930808529ba71jwoolley No module specific data hook for per-connection data
5b30f835d627766b024500189c35bb55e686e890brianp * PR#3191: mod_negotiation
5b30f835d627766b024500189c35bb55e686e890brianp no way to set global quality-of-source (qs) coneg values
d45481dadf7f1f0ffd95b38b1c5b0ea6b2d57888jerenkrantz with multiviews
0c7592379b55b7736dd70d2d87f53af9a2347614jerenkrantz * PR#3568: mod_proxy
0c7592379b55b7736dd70d2d87f53af9a2347614jerenkrantz Accessing URL through proxy server corrupts data.
6d62bc22f8f22a43e66e2e4d8860a5a144316b1astoddard * PR#3605: mod_proxy
6d62bc22f8f22a43e66e2e4d8860a5a144316b1astoddard Some anonymous FTP URLs ask for authentication
71eda8b1c14e4d0adb2f294f22299587cd3d74a9striker * PR#3677: general
922c0ad0014590bb10d13674012683eef44c0bbarederpj New ErrorDocumentMatch directive
922c0ad0014590bb10d13674012683eef44c0bbarederpj * PR#4241: config
aec964227053fab7e59deb26709b94726ce67224rederpj Need to be able to override shebang line to make CGI scripts
aec964227053fab7e59deb26709b94726ce67224rederpj more portable.
1e27f530f81c6058d1a11944ae1e2da45977fc7bjerenkrantz * PR#4244: config
1e27f530f81c6058d1a11944ae1e2da45977fc7bjerenkrantz "Files" and "FilesMatch" regexp does not recognize bang as
1e27f530f81c6058d1a11944ae1e2da45977fc7bjerenkrantz negation operator
469549ac22c6f7b9ecdd9df2565925563e4df84djwoolley * PR#4448: mod_log-any
469549ac22c6f7b9ecdd9df2565925563e4df84djwoolley Please allow CGI env variables (QUERY_STRING, ...) to be logged
7a9f3bef545b0bbc93c2ed758119b0d2e2647c31trawick * PR#4459: mod_include
0be05c658c7e6e5a05fd2d4068d8ac0f030d4752jwoolley Suggestion for better handling of Last-modified headers
223b367115aefc20f1c32fe2d4e2bfcc4bfe108fjwoolley * PR#4490: mod_cgi
223b367115aefc20f1c32fe2d4e2bfcc4bfe108fjwoolley mod_cgi prevents handling of OPTIONS requests
639cf068707865a81ad2c610e72d57b043d20dd6wrowe * PR#5713: os-windows
639cf068707865a81ad2c610e72d57b043d20dd6wrowe [PATCH] install as win32 service with domain account
639cf068707865a81ad2c610e72d57b043d20dd6wrowe Status: Cannot accept password-as-arg, we should prompt the
639cf068707865a81ad2c610e72d57b043d20dd6wrowe user when -k install/-k config with a user argument.
f3f0c2a5f6009d06350341219d1f38a1644708bcbrianp * PR#5993: general
f3f0c2a5f6009d06350341219d1f38a1644708bcbrianp AllowOverride should have a 'CheckNone' and 'AllowNone' argument
33b095d35c5dac22358ce63c32635ab59ce7cb25aaron instead of only 'None'
33b095d35c5dac22358ce63c32635ab59ce7cb25aaronOther bugs that need fixing:
100e6f5dec61d108494565f8f3bfa894afadc33ajerenkrantz * MaxRequestsPerChild measures connections, not requests.
100e6f5dec61d108494565f8f3bfa894afadc33ajerenkrantz Until someone has a better way, we'll probably just rename it
100e6f5dec61d108494565f8f3bfa894afadc33ajerenkrantz "MaxConnectionsPerChild".
5d5d5ca04c57c7ab865924f4648e8f80de27adfebrianp * Regex containers don't work in an intutive way
5d5d5ca04c57c7ab865924f4648e8f80de27adfebrianp Status: No one has come up with an efficient way to fix this
4f3ec2ec53c7ca5ca308d85e2e0beb933a2fd0f9aaron behavior. Dean has suggested getting rid of regex containers
4f3ec2ec53c7ca5ca308d85e2e0beb933a2fd0f9aaron completely.
4f3ec2ec53c7ca5ca308d85e2e0beb933a2fd0f9aaron OtherBill suggests: We at least seem to agree on eliminating
4f3ec2ec53c7ca5ca308d85e2e0beb933a2fd0f9aaron the <Container ~ foo> forms, and using only
4f3ec2ec53c7ca5ca308d85e2e0beb933a2fd0f9aaron <ContainerMatch foo> semantics.
3c0c3ae288166a8736593093c636768702abf263aaron * SIGSEGV on Linux (glibc 2.1.2) isn't caught properly by a
3c0c3ae288166a8736593093c636768702abf263aaron sigwaiting thread. We need to work around this, perhaps unless
b8a843847aae3d9a1838fb8f1de84cd66212f48atrawick there is hope soon for a fixed glibc.
c8411888a54dd5dbfd8a5d337ebf0e911a789063trawick * orig_ct in the byterange/multipart handling may not be
c8411888a54dd5dbfd8a5d337ebf0e911a789063trawick needed. Apache 1.3 just never stashed "multipart" into
c8411888a54dd5dbfd8a5d337ebf0e911a789063trawick r->content_type. We should probably follow suit since the
28caffb98f18c0b9562ac20870f7ab91f3d9a01fjerenkrantz byterange stuff doesn't want the rest of the code to see the
28caffb98f18c0b9562ac20870f7ab91f3d9a01fjerenkrantz multipart content-type; the other code should still think it is
28caffb98f18c0b9562ac20870f7ab91f3d9a01fjerenkrantz dealing with the <orig_ct> stuff.
d757628a07145aa711ba75b4e7c7c01a33133ca1jerenkrantz Status: Greg volunteers to investigate (esp. since he was most
d757628a07145aa711ba75b4e7c7c01a33133ca1jerenkrantz likely the one to break it :-)
1585c2979d749ee71edb6a1d57bc9ff6db12a426jerenkrantzOther features that need writing:
1585c2979d749ee71edb6a1d57bc9ff6db12a426jerenkrantz * Finish infrastructure in core for async MPMs
6bdb2c094666367615890147775bb18761216c8dminfrin Status: post 2.0
6bdb2c094666367615890147775bb18761216c8dminfrin * TODO in source -- just do an egrep on "TODO" and see what's there
c069757188a3cd9ab19b05169b005a824b60b6fcaaronAvailable Patches:
c069757188a3cd9ab19b05169b005a824b60b6fcaaron * Jon Travis's <jtravis@covalent.net> patch to deal with thread-safe
c069757188a3cd9ab19b05169b005a824b60b6fcaaron issues with inet_ntoa. See message <20001201163220.A12827@covalent.net>
11ce97701b22d795fd1a5ec3769a3367bc220921wrowe Status: This is being set aside until the IPv6 work is finished
11ce97701b22d795fd1a5ec3769a3367bc220921wrowe so that we know exactly what is required.
11ce97701b22d795fd1a5ec3769a3367bc220921wrowe * Martin Sojka <msojka@gmx.de>'s patch to add error reporting for failed
11ce97701b22d795fd1a5ec3769a3367bc220921wrowe htpasswd actions due to a full /tmp volume (other programs may have
6cdd5cd95d8323de5cfb912d8fc0df8c5e3d02cejerenkrantz similar problems?)
239ab9d14d3851c2efc1312b3b42c838073533f4jerenkrantz * Mike Abbott's <mja@trudge.engr.sgi.com> patches to improve
239ab9d14d3851c2efc1312b3b42c838073533f4jerenkrantz Status: These were written for 1.3, and are awaiting a port to
c5fb76caa5498c60b991fcba8dfbf3c229f0f4d5rbb * Jim Winstead's <jimw@trainedmonkey.com> patch to add CookieDomain and
c5fb76caa5498c60b991fcba8dfbf3c229f0f4d5rbb other small mod_usertrack features
c5fb76caa5498c60b991fcba8dfbf3c229f0f4d5rbb * Dan Rench's <drench@xnet.com> patch to add allow the errmsg and timefmt
b9cf7102006ac2ccfebcb78174585986ff127ba9jwoolley of SSI's to be modified in the config file. Patch is available in
2b13bc45632d72cdf50ac42149e4fc8bc0d05bf2stoddardOpen issues:
2b13bc45632d72cdf50ac42149e4fc8bc0d05bf2stoddard * Which MPMs will be included with Apache 2.0?