STATUS revision fb585e7ccd9194964442d6b0d89fa93109bcc57d
9174efb969475801d0dc88eee35aae40c748d450ndAPACHE 2.0 STATUS: -*-text-*-
ec79b29695b183f794264bbb578c51e93d1f9b1emartinLast modified at [$Date: 2002/03/22 15:05:21 $]
71f97e4dd8147cbe4c073afd6604089d03497bc5rpluem 2.0.34 : in development
71f97e4dd8147cbe4c073afd6604089d03497bc5rpluem 2.0.33 : tagged March 6, 2002.
ff2e7b35704e8610383152ebc823b94515d9fc2eniq 2.0.32 : released Feburary 16, 2002.
0a44bb1ff02dc8d3623e919c6d59f151bdd92ea0ianh 2.0.31 : rolled Feburary 1, 2002. not released.
0a44bb1ff02dc8d3623e919c6d59f151bdd92ea0ianh 2.0.30 : tagged January 8, 2002. not rolled.
52e3d2373288a4015d4152067c124e4dc9876f08trawick 2.0.29 : tagged November 27, 2001. not rolled.
52e3d2373288a4015d4152067c124e4dc9876f08trawick 2.0.28 : released November 13, 2001
52e3d2373288a4015d4152067c124e4dc9876f08trawick 2.0.27 : rolled November 6, 2001
6ecde05c829c9d0aa24b2b1c18b40c8739997571trawick 2.0.26 : tagged October 16, 2001. not rolled.
6ecde05c829c9d0aa24b2b1c18b40c8739997571trawick 2.0.25 : rolled August 29, 2001
6ecde05c829c9d0aa24b2b1c18b40c8739997571trawick 2.0.24 : rolled August 18, 2001
17d53ea32c4968e47733f1c2c063ae07d280efd6jerenkrantz 2.0.23 : rolled August 9, 2001
17d53ea32c4968e47733f1c2c063ae07d280efd6jerenkrantz 2.0.22 : rolled July 29, 2001
17d53ea32c4968e47733f1c2c063ae07d280efd6jerenkrantz 2.0.21 : rolled July 20, 2001
2d5532b13110a8d85653da92e97795b09cc25cc2trawick 2.0.20 : rolled July 8, 2001
b38565306421ff53e9f7499bc728d6df5cec294dpquerna 2.0.19 : rolled June 27, 2001
b38565306421ff53e9f7499bc728d6df5cec294dpquerna 2.0.18 : rolled May 18, 2001
b38565306421ff53e9f7499bc728d6df5cec294dpquerna 2.0.17 : rolled April 17, 2001
b38565306421ff53e9f7499bc728d6df5cec294dpquerna 2.0.16 : rolled April 4, 2001
3aeb30211790fef38a8297f990b7ad3b2c46ece9colm 2.0.15 : rolled March 21, 2001
3aeb30211790fef38a8297f990b7ad3b2c46ece9colm 2.0.14 : rolled March 7, 2001
bf0dec31a67271a4a67a6d3a3a9b3cdb41278390rpluem 2.0a9 : released December 12, 2000
bf0dec31a67271a4a67a6d3a3a9b3cdb41278390rpluem 2.0a8 : released November 20, 2000
bf0dec31a67271a4a67a6d3a3a9b3cdb41278390rpluem 2.0a7 : released October 8, 2000
bf0dec31a67271a4a67a6d3a3a9b3cdb41278390rpluem 2.0a6 : released August 18, 2000
7612604db51e61f6ec4a6593c188e6d0640646abwrowe 2.0a5 : released August 4, 2000
7612604db51e61f6ec4a6593c188e6d0640646abwrowe 2.0a4 : released June 7, 2000
7612604db51e61f6ec4a6593c188e6d0640646abwrowe 2.0a3 : released April 28, 2000
7612604db51e61f6ec4a6593c188e6d0640646abwrowe 2.0a2 : released March 31, 2000
7612604db51e61f6ec4a6593c188e6d0640646abwrowe 2.0a1 : released March 10, 2000
0a2c6381b34b977101327f412a4b1ddc8a361ae7rpluemPlease consult the following STATUS files for information
0a2c6381b34b977101327f412a4b1ddc8a361ae7rpluemon related projects:
0830215c2542165c027f092c18924f2885c98b49trawickCURRENT RELEASE NOTES:
0830215c2542165c027f092c18924f2885c98b49trawick * 34 status: Let's get all API changes and showstoppers in this one.
bc886ee8559b4feb30d32dc8a50ba2c013c3cc02trawickFINAL RELEASE SHOWSTOPPERS:
9505e1a9012d92bfd2b6228e675156f44a27aa04colm * If any request gets to the core handler, without a flag that this
9505e1a9012d92bfd2b6228e675156f44a27aa04colm r->filename was tested by dir/file_walk, we need to 500 at the very
3aeb30211790fef38a8297f990b7ad3b2c46ece9colm end of the ap_process_request_internal() processing. This provides
3aeb30211790fef38a8297f990b7ad3b2c46ece9colm authors of older modules better compatibility, while still improving
3aeb30211790fef38a8297f990b7ad3b2c46ece9colm the security and robustness of 2.0.
3aeb30211790fef38a8297f990b7ad3b2c46ece9colm Status: still need to decide where this goes, OtherBill comments...
6eb3a057b5e4f046faba79104d7941552de755dacolm Message-ID: <065701c14526$495203b0$96c0b0d0@roweclan.net>
6eb3a057b5e4f046faba79104d7941552de755dacolm we need to look at halting this in the 'default handler' case,
2e02c854b4531480fe3f7ab353154c99c5bd97e8jorton and that implies pushing the 'handler election' into the request
2e02c854b4531480fe3f7ab353154c99c5bd97e8jorton internal processing phase from the run request phase.
2e02c854b4531480fe3f7ab353154c99c5bd97e8jorton Jim asks: would a stopgap be something bogus like adding another
2e02c854b4531480fe3f7ab353154c99c5bd97e8jorton flag to request_rec ala eos_sent and before we OK, if not set
c07be90520f577cdffc0ee59a81e7f2be4ab24c6martin Jeff says: reviewing the original message and the one
2e02c854b4531480fe3f7ab353154c99c5bd97e8jorton follow-up (also from OtherBill) it looks like OtherBill had a
2e02c854b4531480fe3f7ab353154c99c5bd97e8jorton good handle on the problem, though I wonder why not just put a
2e02c854b4531480fe3f7ab353154c99c5bd97e8jorton simple check in default_handler to see if dir/file_walk has
2e02c854b4531480fe3f7ab353154c99c5bd97e8jorton been done (a footprint left by dir/file_walk doesn't have to
2e02c854b4531480fe3f7ab353154c99c5bd97e8jorton be in request_rec; a better place is core_request_config)
70706a4d1ead0d72a443eac1cc446a0f1a4de069colm gregames says: can this happen somehow without a broken module
70706a4d1ead0d72a443eac1cc446a0f1a4de069colm being involved? If not, why waste cycles trying to defend against
70706a4d1ead0d72a443eac1cc446a0f1a4de069colm potential broken modules? It seems futile. Please vote.
70706a4d1ead0d72a443eac1cc446a0f1a4de069colm not a showstopper: gregames
70706a4d1ead0d72a443eac1cc446a0f1a4de069colm showstopper:
9cfe09f7e922f7bdf61bfe10a47e8f25145cdf92colm * API changes planned for 2.0 that should happen before the
9cfe09f7e922f7bdf61bfe10a47e8f25145cdf92colm GA release:
9cfe09f7e922f7bdf61bfe10a47e8f25145cdf92colm * Free lists for bucket allocation
2e02c854b4531480fe3f7ab353154c99c5bd97e8jorton * We do not properly substitute the prefix-variables in the configuration
2e02c854b4531480fe3f7ab353154c99c5bd97e8jorton scripts or generated-configs. (i.e. if sysconfdir is etc,
2e02c854b4531480fe3f7ab353154c99c5bd97e8jorton httpd-std.conf points to conf.)
2e02c854b4531480fe3f7ab353154c99c5bd97e8jorton Aaron says: This is not a showstopper, these problems have existed
2e02c854b4531480fe3f7ab353154c99c5bd97e8jorton for as long as I can remember. It would be nice to fix
2e02c854b4531480fe3f7ab353154c99c5bd97e8jorton them but they are not new.
2e02c854b4531480fe3f7ab353154c99c5bd97e8jorton Not a showstopper: Jeff, Aaron
a09378e32e823d5619f766394fcdde1cfa16c5dajim * Address popular PRs
a09378e32e823d5619f766394fcdde1cfa16c5dajim * Win32 doesn't install as service correctly [9863, 9914, 9961]
a09378e32e823d5619f766394fcdde1cfa16c5dajimCURRENT VOTES:
a09378e32e823d5619f766394fcdde1cfa16c5dajim * Should we always build binaries statically unless otherwise
ce8e09ffeb8c696f7fc6c5dfa8814c52caaac59bcolm indicated?
ce8e09ffeb8c696f7fc6c5dfa8814c52caaac59bcolm Message-ID: <20020129210006.B23512@Lithium.MeepZor.Com>
5340bb8547e62ec8ede44b1b086b7da37ee64eefcolm -1: Justin, Ian
5340bb8547e62ec8ede44b1b086b7da37ee64eefcolm * If the parent process dies, should the remaining child processes
f6e71f622ffde40c654288851ff65b8505cf7596colm "gracefully" self-terminate. Or maybe we should make it a runtime
f6e71f622ffde40c654288851ff65b8505cf7596colm option, or have a concept of 2 parent processes (one being a
f6e71f622ffde40c654288851ff65b8505cf7596colm "hot spare").
f8ae06c7d0615d8b9441099794fe80a3f2915469colm See: Message-ID: <3C58232C.FE91F19F@Golux.Com>
f8ae06c7d0615d8b9441099794fe80a3f2915469colm Self-destruct: Ken, Martin
6eb3a057b5e4f046faba79104d7941552de755dacolm Not self-destruct: BrianP, Ian, Cliff, BillS
6eb3a057b5e4f046faba79104d7941552de755dacolm Make it runtime configurable: Aaron, Jim, Justin
6eb3a057b5e4f046faba79104d7941552de755dacolm Have 2 parents: +1: Jim
6eb3a057b5e4f046faba79104d7941552de755dacolm +0: Martin (while standing by, could it do
dface6840439ab84a36749f40e7773ebfdc2066ejorton something useful?)
dface6840439ab84a36749f40e7773ebfdc2066ejorton * Make the worker MPM the default MPM for threaded Unix boxes.
dface6840439ab84a36749f40e7773ebfdc2066ejorton +1: Justin, Ian, Cliff
dface6840439ab84a36749f40e7773ebfdc2066ejorton -0: Aaron (premature decision, needs more discussion), Lars
1febae173a82bc2a71c3c0ba4105cf674000791bjimRELEASE NON-SHOWSTOPPERS BUT WOULD BE REAL NICE TO WRAP THESE UP:
1febae173a82bc2a71c3c0ba4105cf674000791bjim * Convert all instances of the old apr_lock_t type to the new
2b409b0583ef844d6f85002160cbc6df95b9afa8jerenkrantz Status: Aaron has converted all but the perchild MPM to
2b409b0583ef844d6f85002160cbc6df95b9afa8jerenkrantz use the new lock API. Since perchild has been put
2b409b0583ef844d6f85002160cbc6df95b9afa8jerenkrantz on the back burner, this is no longer a showstopper.
2b409b0583ef844d6f85002160cbc6df95b9afa8jerenkrantz Aaron will patch perchild as soon as it becomes
454d060a0d7b13606978c536f632be71e311959djerenkrantz buildable again.
454d060a0d7b13606978c536f632be71e311959djerenkrantz * With AP_MODE_EXHAUSTIVE in the core, it is finally clear to me
454d060a0d7b13606978c536f632be71e311959djerenkrantz how the Perchild MPM should be re-written. It hasn't worked
1392b9997000dceb6aca39da0a1e6405ccb6f0a6jerenkrantz correctly since filters were added because it wasn't possible to
1392b9997000dceb6aca39da0a1e6405ccb6f0a6jerenkrantz get the content that had already been written and the socket at
1392b9997000dceb6aca39da0a1e6405ccb6f0a6jerenkrantz the same time. This mode lets us do that, so the MPM can be
4b0a0ae3fd77f8fc8fd2a693be33ffe4d99cb94ajerenkrantz * htpasswd blindly processes the file you give it, and does no
4b0a0ae3fd77f8fc8fd2a693be33ffe4d99cb94ajerenkrantz sanity checking before totally corrupting whatever file it was
4b0a0ae3fd77f8fc8fd2a693be33ffe4d99cb94ajerenkrantz you thought you had. It should check the input file and bail
4b0a0ae3fd77f8fc8fd2a693be33ffe4d99cb94ajerenkrantz if it finds non-comment lines that do not contain exactly 1
ee508128c414648982d1cca7801f63b01a0a4f8aminfrin ':' character.
ee508128c414648982d1cca7801f63b01a0a4f8aminfrin Message-ID: <20020217150457.A31632@clove.org>
ee508128c414648982d1cca7801f63b01a0a4f8aminfrin * Can a static httpd be built reliably?
f07ae092933846055c4ca046c5a65de83d79c9c6pquerna Message-ID: <20020207142751.T31582@clove.org>
f07ae092933846055c4ca046c5a65de83d79c9c6pquerna * [Ken] Test suite failures:
9b88ec397b4155f3ff79bf7c08de5f15d21765b3pquerna o worker is also failing some of the 'cgi' subtests
9b88ec397b4155f3ff79bf7c08de5f15d21765b3pquerna (see <URL:http://Source-Zone.Org/Apache/regression/>):
9b88ec397b4155f3ff79bf7c08de5f15d21765b3pquerna Justin says: "Worker should be fine and passes httpd-test here.
a42c13195a5c68d55f6b2bbf5b5e1e66ab45d5d4pquerna If you can provide evidence that it can be reproduced
a42c13195a5c68d55f6b2bbf5b5e1e66ab45d5d4pquerna outside of httpd-test, then it's a showstopper. I
a42c13195a5c68d55f6b2bbf5b5e1e66ab45d5d4pquerna think it's a perl or a httpd-test problem."
2a90918318af50d5712ae7159e5625476332c184pquerna Not a showstopper: Justin
2a90918318af50d5712ae7159e5625476332c184pquerna * Usage of APR_BRIGADE_NORMALIZE in core_input_filter should be
ede0584e7ec53cc1ea832c5c50281431d9ecc576pquerna removed if possible.
ede0584e7ec53cc1ea832c5c50281431d9ecc576pquerna Message-ID: <Pine.LNX.4.33.0201202232430.318-100000@deepthought.cs.virginia.edu>
ede0584e7ec53cc1ea832c5c50281431d9ecc576pquerna * There is a bug in how we sort some hooks, at least the pre-config
82d2a5debc5a6ed2118ac5916d9ba36ad0b5d78btrawick hook. The first time we call the hooks, they are in the correct
42167da203d969a1402cf7ce09c14586c04af1dfjim order, but the second time, we don't sort them correctly. Currently,
42167da203d969a1402cf7ce09c14586c04af1dfjim the modules/http/config.m4 file has been renamed to
42167da203d969a1402cf7ce09c14586c04af1dfjim modules/http/config2.m4 to work around this problem, it should moved
81ebf4d4f943d1ec988dd10105354e510c2da1d8jim back when this is fixed. rbb
e62b34b26eb8e82bf6f06ff99a08e304b48792ecpquerna Justin says: "Is this really a showstopper? This has been here
e62b34b26eb8e82bf6f06ff99a08e304b48792ecpquerna forever. What's wrong? Does this have to do with
e62b34b26eb8e82bf6f06ff99a08e304b48792ecpquerna autoconf or m4?"
27420b7ba57bf78f7a85b19fb0155053a3aff80dwrowe Not a showstopper: Justin, BrianP, trawick, gregames
27420b7ba57bf78f7a85b19fb0155053a3aff80dwrowe * The Add...Filter and Set...Filter directives do not allow the
ceeb4b77f172b9d0125f1b4b323675d700b2c43bminfrin administrator to order filters, beyond the order of filename (mime)
ceeb4b77f172b9d0125f1b4b323675d700b2c43bminfrin extensions. It isn't clear if Set...Filter(s) should be inserted
ceeb4b77f172b9d0125f1b4b323675d700b2c43bminfrin before or after the Add...Filter(s) which are ordered by sequence of
9152b0c4366af633c4b7982af01bd9a821f410d5pquerna filename extensions. At minimum, some sort of +-[0-10] syntax seems
9152b0c4366af633c4b7982af01bd9a821f410d5pquerna like the quickest fix for a 2.0 gold release.
9152b0c4366af633c4b7982af01bd9a821f410d5pquerna Justin says: "Could we delay this for a point release or 2.1?"
9152b0c4366af633c4b7982af01bd9a821f410d5pquerna Not a showstopper: justin, wrowe, trawick, stoddard, Jim, Ian, Aaron,
424c1a743525e9c5008e29b39a99363723409a1dtrawick * Get perchild to work on platforms other than Linux. This
424c1a743525e9c5008e29b39a99363723409a1dtrawick will require a portable mechanism to pass data and file/socket
424c1a743525e9c5008e29b39a99363723409a1dtrawick descriptors between vhost child groups. An API was proposed
4375cabac8fc22b3717687ffdbce9bbdf095f255trawick on dev@apr:
4375cabac8fc22b3717687ffdbce9bbdf095f255trawick Message-ID: <20020111115006.K1529@clove.org>
39dbd3f60b93f5e0fbf46d9ae237f6742e113442pquerna * CGI single-byte reads
39dbd3f60b93f5e0fbf46d9ae237f6742e113442pquerna BrianP suggests that this is caused by the ap_scan_script_header_err()
89211a3153be8b03353c3bfbca45fed67cb80f0bpquerna routine, which will do single-byte reads until it finds the end
89211a3153be8b03353c3bfbca45fed67cb80f0bpquerna of the header, at which point it constructs a pipe-bucket (buffered)
89211a3153be8b03353c3bfbca45fed67cb80f0bpquerna to read from.
89211a3153be8b03353c3bfbca45fed67cb80f0bpquerna Proposed solution in:
89211a3153be8b03353c3bfbca45fed67cb80f0bpquerna Message-ID: <3C36ADAF.60601@cnet.com>
89211a3153be8b03353c3bfbca45fed67cb80f0bpquerna * Try to get libtool inter-library dependency code working on AIX.
89211a3153be8b03353c3bfbca45fed67cb80f0bpquerna Message-ID: <cm3n10lx555.fsf@rdu163-40-092.nc.rr.com>
a14ccf0f7e9b44c6848334823542a1799577f669pquerna Justin says: If we get it working on AIX, we can enable this
a14ccf0f7e9b44c6848334823542a1799577f669pquerna on all platforms and clean up our build system
43c24b7301a9df7014ce54c0bc55ac4754cf5b0dpquerna Jeff says: I thought I tested a patch for you sometime in
a2e37e48efb113e8e4f9d9ca9c8286aaac6c936cpquerna January that you were going to commit within a few
007ba11dc68651df7f872f35947ae82a595d9e02pquerna * Handling of %2f in URIs. Currently both 1.3 and 2.0
007ba11dc68651df7f872f35947ae82a595d9e02pquerna completely disallow %2f in the request URI path (see
007ba11dc68651df7f872f35947ae82a595d9e02pquerna ap_unescape_url() in util.c). It's permitted and passed
44ca834b970b454b844efb96f219bdf49fee71e5trawick through in the query string, however. Roy says the
44ca834b970b454b844efb96f219bdf49fee71e5trawick original reason for disallowing it, from five years ago,
44ca834b970b454b844efb96f219bdf49fee71e5trawick was to protect CGI scripts that applied PATH_INFO to
eb8430fd3bc941c0b3ba8bba3884b7fc464bf458pquerna a filesystem location and which might be tricked by
eb8430fd3bc941c0b3ba8bba3884b7fc464bf458pquerna ..%2f..%2f(...). We *should* allow path-info of the
eb8430fd3bc941c0b3ba8bba3884b7fc464bf458pquerna form 'http://foo.com/index.cgi/path/to/path%2finfo'.
fde88bb682426885c679198ee130d2d5a29e8c0fbnicholes Since we've revamped a lot of our processing of path
fde88bb682426885c679198ee130d2d5a29e8c0fbnicholes segments, it would be nice to allow this, or at least
fde88bb682426885c679198ee130d2d5a29e8c0fbnicholes allow it conditionally with a directive.
fde88bb682426885c679198ee130d2d5a29e8c0fbnicholes * FreeBSD, threads, and worker MPM. All seems to work fine
fde88bb682426885c679198ee130d2d5a29e8c0fbnicholes if you only have one worker process with many threads. Add
fde88bb682426885c679198ee130d2d5a29e8c0fbnicholes a second worker process and the accept lock seems to be
fde88bb682426885c679198ee130d2d5a29e8c0fbnicholes lost. This might be an APR issue with how it deals with
60e385aa992e11a6cb0504e8d4fc35186e8d848bpquerna the child_init hook (i.e. the fcntl lock needs to be resynced).
60e385aa992e11a6cb0504e8d4fc35186e8d848bpquerna More examination and analysis is required.
60e385aa992e11a6cb0504e8d4fc35186e8d848bpquerna Status: This has also been reported on Cygwin.
d66ac514cc15e99228d72c56c6c3daf25da8d360niq Message-ID: <3C2CC514.8EF3BED1@wapme-systems.de> (cygnus)
d66ac514cc15e99228d72c56c6c3daf25da8d360niq Justin says: So, FreeBSD-CURRENT and Cywin have the same
d66ac514cc15e99228d72c56c6c3daf25da8d360niq problem. Yum. If another platform has this
e23b77006a8b079c0ad52e42ba2029e759455b8fjorton with worker, this becomes a showstopper.
e23b77006a8b079c0ad52e42ba2029e759455b8fjorton Aaron says: I spent some time disecting this and have come to
e23b77006a8b079c0ad52e42ba2029e759455b8fjorton the conclusion that it is not a problem in the worker MPM
e23b77006a8b079c0ad52e42ba2029e759455b8fjorton (or at least, it is not isolated to a problem in worker).
22c347c08cd77d5e2c32536b467b389fd33d631fpquerna I'll list some of the problems I'm seeing in case someone
8c46f9f81ae6357dc97560d6d85154f19cf251aejim else wants to pick up where I've left off:
b6e2b36d6972a0e243368eb567e91d511fb40b1bbnicholes - Delivery of just about any signal to one of the child
b6e2b36d6972a0e243368eb567e91d511fb40b1bbnicholes processes will send it into an infinite loop as well.
b6e2b36d6972a0e243368eb567e91d511fb40b1bbnicholes - Even though the parent is spinning out of control,
b6e2b36d6972a0e243368eb567e91d511fb40b1bbnicholes at first the child or children will appear to work
b6e2b36d6972a0e243368eb567e91d511fb40b1bbnicholes properly. At times it is possible to get it into a state,
ad4a5834dfdf9488af4ac3238ea4d8b18dce49acpquerna however, where a request will hang until another concurrent
e5882a36d7756850cc829f5f2286120b877458b1pquerna request "kicks" the first, at which point the second will
e5882a36d7756850cc829f5f2286120b877458b1pquerna hang. My theory is that this has to do with the
8b7594c66e764f5cd3506b6f2459497ab65a8b03pquerna pthread_cond_*() implementation in FreeBSD, but it's still
8b7594c66e764f5cd3506b6f2459497ab65a8b03pquerna possible that it is in APR.
8c46f9f81ae6357dc97560d6d85154f19cf251aejim Justin adds: Oh, FreeBSD threads are implemented entirely with
8c46f9f81ae6357dc97560d6d85154f19cf251aejim select()/poll()/longjmp(). Welcome to the nightmare.
20f1b1a67eef5ab0f3295608c89964a7dca4fdd1pquerna So, that means a ktrace output also has the thread
20f1b1a67eef5ab0f3295608c89964a7dca4fdd1pquerna scheduling internals in it (since it is all the same to
20f1b1a67eef5ab0f3295608c89964a7dca4fdd1pquerna the kernel). Which makes it hard to distinguish between
20f1b1a67eef5ab0f3295608c89964a7dca4fdd1pquerna our select() calls and their select() calls.
599c5abcc7fec60611956f56b732eca033c287d2pquerna *bangs head on wall repeatedly* But, some of the libc_r
599c5abcc7fec60611956f56b732eca033c287d2pquerna files have a DBG_MSG #define. This is moderately helpful
599c5abcc7fec60611956f56b732eca033c287d2pquerna when used with -DNO_DETACH. The kernel scheduler isn't
e76fdcdfb8994ad70776526f50fa013b3e9a6033bnicholes waking up the threads on a select(). Yum. And, I bet
e76fdcdfb8994ad70776526f50fa013b3e9a6033bnicholes those decrementing select calls have to do with the
e76fdcdfb8994ad70776526f50fa013b3e9a6033bnicholes scheduler. Time to brush up on our OS fundamentals.
eb9b491d7b262dad572ec2f1f75eea592283f81apquerna * There is increasing demand from module writers for an API
1ef90d1763e489969ec7a01c77bcb158ece5ae83jorton that will allow them to control the server � la apachectl.
c5cb4c9e7c8696907ddebab2a0037717c710b0f6jorton Reasons include sole-function servers that need to die if
6d00a5e2c08ddbff9614ecc17a1c085462811f89slive an external dependency (e.g., a database) fails, et cetera.
6d00a5e2c08ddbff9614ecc17a1c085462811f89slive Perhaps something in the (ever more abused) scoreboard?
6d00a5e2c08ddbff9614ecc17a1c085462811f89slive rbb: I don't believe the scoreboard is the correct mechanism
6d00a5e2c08ddbff9614ecc17a1c085462811f89slive for this. We already have a pipe that goes between parent
6d00a5e2c08ddbff9614ecc17a1c085462811f89slive and child for graceful shutdown events, along with an API that
6d00a5e2c08ddbff9614ecc17a1c085462811f89slive can be used to send a message down that pipe. In threaded MPMs,
6d00a5e2c08ddbff9614ecc17a1c085462811f89slive it is easy enough to make that one pipe be used for graceful
6d00a5e2c08ddbff9614ecc17a1c085462811f89slive and graceless events, and it is also easy to open that pipe
6d00a5e2c08ddbff9614ecc17a1c085462811f89slive to both parent and child for writing. Then we just need to
6d00a5e2c08ddbff9614ecc17a1c085462811f89slive figure out how to do graceless on non-threaded MPMs.
6d00a5e2c08ddbff9614ecc17a1c085462811f89slive * Allow the DocumentRoot directive within <Location > scopes? This
6d00a5e2c08ddbff9614ecc17a1c085462811f89slive allows the beloved (crusty) Alias /foo/ /somepath/foo/ followed
6d00a5e2c08ddbff9614ecc17a1c085462811f89slive by a <Directory /somepath/foo> to become simply
6d00a5e2c08ddbff9614ecc17a1c085462811f89slive <Location /foo/> DocumentRoot /somefile/foo (IMHO a bit more legible
6d00a5e2c08ddbff9614ecc17a1c085462811f89slive and in-your-face.) DocumentRoot unset would be accepted [and would
4f6e78091b3f45a5782389ae25b62516a7a0c842bnicholes not permit content to be served, only virtual resources such as
4f6e78091b3f45a5782389ae25b62516a7a0c842bnicholes server-info or server-status.
4f6e78091b3f45a5782389ae25b62516a7a0c842bnicholes This proposed change would _not_ depricate Alias.
4f6e78091b3f45a5782389ae25b62516a7a0c842bnicholes * Win32: Rotatelogs sometimes is not terminated when Apache
7d15331eeb5429d7148d13d6fd914a641bf1c000pquerna goes down hard. FirstBill was looking at possibly tracking the
7d15331eeb5429d7148d13d6fd914a641bf1c000pquerna child's-child processes in the parent process.
7d15331eeb5429d7148d13d6fd914a641bf1c000pquerna OtherBill asks, wasn't this fixed?
e9be3aacfd6c0a1208e6c91a133be92ed0f94fe1bnicholes stoddard: Not fixed. Shared scoreboard might offer a good
e9be3aacfd6c0a1208e6c91a133be92ed0f94fe1bnicholes way for the parent to keep track of 'other child' processes
e9be3aacfd6c0a1208e6c91a133be92ed0f94fe1bnicholes and whack them if the child goes down.
1d14622beee568462689b2bbc6a9e0e9b6a40583striker * Win32: Add a simple hold console open patch (wait for close or
1d14622beee568462689b2bbc6a9e0e9b6a40583striker the ESC key, with a nice message) if the server died a bad
6e5f5644328bf50c3aa295d0ab20903369010829gregames death (non-zero exit code) in console mode.
6e5f5644328bf50c3aa295d0ab20903369010829gregames Resolution: bring forward same ugly hacks from 1.3.13-.20
e978097e0bf2ae161b6f9dde40eaf089bf046c89ake * Port of mod_ssl to Apache 2.0:
e978097e0bf2ae161b6f9dde40eaf089bf046c89ake The current porting state is summarized in modules/ssl/README. The
275419d6395e6f072962fb701b89accaff1f3690jerenkrantz remaining work includes:
275419d6395e6f072962fb701b89accaff1f3690jerenkrantz (1) stablizing/optimizing the SSL filter logic
275419d6395e6f072962fb701b89accaff1f3690jerenkrantz (2) Enabling SSL extentions
e777da9fa7ff3138fead7860b53ef00e67a40e26jerenkrantz (3) Trying to seperate the https filter logic from mod_ssl -
e777da9fa7ff3138fead7860b53ef00e67a40e26jerenkrantz This is to facilitate other modules that wish to use the https
e777da9fa7ff3138fead7860b53ef00e67a40e26jerenkrantz filter or the mod_ssl logic or both as required.
8bdea88407c848c1c2693655e2f8b23abde12307bnicholes * Performance & Debug: Eliminate most (and perhaps all) of the
8bdea88407c848c1c2693655e2f8b23abde12307bnicholes malloc/free calls in the bucket brigade code. Need some
8bdea88407c848c1c2693655e2f8b23abde12307bnicholes light weight memory management functions that allow freeing
275419d6395e6f072962fb701b89accaff1f3690jerenkrantz memory (putting it back into a memory pool) when it is no
8e7fb6968047a527d1ccde25ad2aed20da5150ddjerenkrantz longer needed. Enabling simple debugging features like guard
8e7fb6968047a527d1ccde25ad2aed20da5150ddjerenkrantz bands, double free detection, etc. would be cool but certainly
8e7fb6968047a527d1ccde25ad2aed20da5150ddjerenkrantz not a hard requirement.
5a9fa4d75086e942f6e850e1a2e96c4c27a845d0jerenkrantz Status: Cliff started to implement this using SMS as has
5a9fa4d75086e942f6e850e1a2e96c4c27a845d0jerenkrantz been discussed at length for months, but since
5a9fa4d75086e942f6e850e1a2e96c4c27a845d0jerenkrantz SMS is not being used anywhere else in the server,
5a9fa4d75086e942f6e850e1a2e96c4c27a845d0jerenkrantz several people expressed the opinion that we should
5a9fa4d75086e942f6e850e1a2e96c4c27a845d0jerenkrantz get rid of it entirely, meaning that the buckets
36c8049de63c446926139936c3d195330a0539cetrawick need their own memory management (free list) functions.
36c8049de63c446926139936c3d195330a0539cetrawick Cliff will implement that this weekend so we at least
36c8049de63c446926139936c3d195330a0539cetrawick have something to look at/compare with.
36c8049de63c446926139936c3d195330a0539cetrawick * Eliminate unnecessary creation of pipes in mod_cgid
0206c121a68a63559b2e843288e81bcf16093e46jerenkrantz * Combine log_child and piped_log_spawn. Clean up http_log.c.
9174efb969475801d0dc88eee35aae40c748d450nd Common logging API.
8ff094bdec6a2e1c355c1e6e95e9952d4fae7766jerenkrantz * Document mod_file_cache.
628cf3411e8a6d09e27b3666312e43832dda93f3jorton * Platforms that do not support fork (primarily Win32 and AS/400)
628cf3411e8a6d09e27b3666312e43832dda93f3jorton Architect start-up code that avoids initializing all the modules
628cf3411e8a6d09e27b3666312e43832dda93f3jorton in the parent process on platforms that do not support fork.
f0d89a5a23222e031db8113478645f28688fa748jorton * Win32: Migrate the MPM over to use APR thread/process calls. This
f0d89a5a23222e031db8113478645f28688fa748jorton would eliminate some code in the Win32 branch that essentially
f0d89a5a23222e031db8113478645f28688fa748jorton duplicates what is in APR.
f0d89a5a23222e031db8113478645f28688fa748jorton * There are still a number of places in the code where we are
f0d89a5a23222e031db8113478645f28688fa748jorton losing error status (i.e. throwing away the error returned by a
f0d89a5a23222e031db8113478645f28688fa748jorton system call and replacing it with a generic error code)
f0d89a5a23222e031db8113478645f28688fa748jorton * Mass vhosting version of suEXEC.
65d743d7fbb53143636ee2dec8fe8d8a1a581a6bjerenkrantz * All DBMs suffer from confusion in support/dbmmanage (perl script) since
65d743d7fbb53143636ee2dec8fe8d8a1a581a6bjerenkrantz the dbmmanage employs the first-matched dbm format. This is not
afc08f35f5f387896bc625cdee21b88c7fe7699djerenkrantz necessarily the library that Apache was built with. Aught to
a0fd132e01ab69f1c48e3d6a6791447cb6d65e51jerenkrantz rewrite dbmmanage upon installation to bin/ with the proper library
a0fd132e01ab69f1c48e3d6a6791447cb6d65e51jerenkrantz for predictable mod_auth_dbm administration.
a6bb6f2fb734e488a9b6335fabea3431f9dcf253jerenkrantz Questions; htdbm exists, time to kill dbmmanage, or does it remain
a6bb6f2fb734e488a9b6335fabea3431f9dcf253jerenkrantz useful as a perl dbm management example? If we keep it,
a6bb6f2fb734e488a9b6335fabea3431f9dcf253jerenkrantz do we address the issue above?
3ca80e703a960eca0760c23636b7fe502a8f0342bnicholes * Integrate mod_dav.
3ca80e703a960eca0760c23636b7fe502a8f0342bnicholes Some additional items remaining:
3ca80e703a960eca0760c23636b7fe502a8f0342bnicholes - case_preserved_filename stuff
3ca80e703a960eca0760c23636b7fe502a8f0342bnicholes (use the new canonical name stuff?)
5300d4a4fabe3b594da950e4b9ab0f90e076546ejerenkrantz - find a new home for ap_text(_header)
5300d4a4fabe3b594da950e4b9ab0f90e076546ejerenkrantz - is it possible to remove the DAV: namespace stuff from util_xml?
b1306729566b49fb30aed5c46adaf07a637115afjerenkrantz * ap_core_translate() and its use by mod_mmap_static and mod_file_cache
b1306729566b49fb30aed5c46adaf07a637115afjerenkrantz are a bit wonky. The function should probably be exposed as a utility
b1306729566b49fb30aed5c46adaf07a637115afjerenkrantz function (such as ap_translate_url2fs() or ap_validate_fs_url() or
aa50e6503e0c7fbdc7d9bfab7d3a5fa640c2c8c1wrowe something). Another approach would be a new hook phase after
aa50e6503e0c7fbdc7d9bfab7d3a5fa640c2c8c1wrowe "translate" which would allow the module to munge what the
418ee053321d0ee451bb482a9becdfcd3344201fjim translation has decided to do.
5c6cb72bfeee541644cea8177aefce1157571d3bjerenkrantz Status: Greg +1 (volunteers), Ryan +1
5c6cb72bfeee541644cea8177aefce1157571d3bjerenkrantz * Explore use of a post-config hook for the code in http_main.c which
db8551deeb08fa799e7f27e8b748a9397f747bdcjorton calls ap_fixup_virutal_hosts(), ap_fini_vhost_config(), and
db8551deeb08fa799e7f27e8b748a9397f747bdcjorton ap_sort_hooks() [to reduce the logic in main()]
db8551deeb08fa799e7f27e8b748a9397f747bdcjorton * read the config tree just once, and process N times (as necessary)
9e3209bc06ddf32f23e4b254faa45914bc323cc9jim * (possibly) use UUIDs in mod_unique_id and/or mod_usertrack
45ed846f4ed90f05c084f6a33d688e642be4e623jerenkrantz * (possibly) port the bug fix for PR 6942 (segv when LoadModule is put
45ed846f4ed90f05c084f6a33d688e642be4e623jerenkrantz into a VirtualHost container) to 2.0.
54d22ed1c429b903b029bbd62621f11a9e286137minfrin * shift stuff to mod_core.h
54d22ed1c429b903b029bbd62621f11a9e286137minfrin * callers of ap_run_create_request() should check the return value
54d22ed1c429b903b029bbd62621f11a9e286137minfrin for failure (Doug volunteers)
54d22ed1c429b903b029bbd62621f11a9e286137minfrin * Win32: Get Apache working on Windows 95/98. The following work
54d22ed1c429b903b029bbd62621f11a9e286137minfrin (at least) needs to be done:
54d22ed1c429b903b029bbd62621f11a9e286137minfrin - Document warning that OSR2 is required (for Crypt functions, in
92e8e44864d94866eefcbfde0a53fa3d12855149jerenkrantz rand.c, at least.) This could be resolved with an SSL library, or
92e8e44864d94866eefcbfde0a53fa3d12855149jerenkrantz randomization in APR itself.
92e8e44864d94866eefcbfde0a53fa3d12855149jerenkrantz - Bring the Win9xConHook.dll from 1.3 into 2.0 (no sense till it
06106b6b63b112a09de1b66fa29596035c0176ffthommay actually works) and add in a splash of Win9x service code.
06106b6b63b112a09de1b66fa29596035c0176ffthommay * In order to use a DSO version of mod_ssl we have to link with
e335319a08e12eb7daff9afa80e985dc53f652b8jorton -lssl and -lcrypto. A workaround is in place right now where the
e335319a08e12eb7daff9afa80e985dc53f652b8jorton entire EXTRA_LIBS macro is being appended to the objects list, but
e335319a08e12eb7daff9afa80e985dc53f652b8jorton this is a hack. We should either revamp the APACHE_CHECK_SSL_TOOLKIT
e335319a08e12eb7daff9afa80e985dc53f652b8jorton autoconf function or come up with some other autoconf checks to
e335319a08e12eb7daff9afa80e985dc53f652b8jorton search for libssl and libcrypto and properly add them to mod_ssl's
c8794ec1e4c474ae101ce3835080f638136e7860erikabele * Fix the worker MPM to use POD to kill child processes instead
c8794ec1e4c474ae101ce3835080f638136e7860erikabele of ap_os_killpg, regardless of how they should die. (Ryan Bloom)
bb07ee33bce1a448bcc60ca43720b1ab1c413f87minfrin * mod_cache: handle cache_control: no_cache "field_name" to enable
bb07ee33bce1a448bcc60ca43720b1ab1c413f87minfrin cacheing the response w/o header "field_name"
22dda44322067379eeba28d7ec7fc833cb04c0dfminfrin See RFC2616 section 14.9.1
7b6ba9c468f26bdb3492d5e8cb79628a3b04e8c8wrowe * Scoreboard structures could be changed in the future such that
7b6ba9c468f26bdb3492d5e8cb79628a3b04e8c8wrowe proper alignment is not maintained, leading to segfaults on
7b6ba9c468f26bdb3492d5e8cb79628a3b04e8c8wrowe some systems. Cliff posted a patch to deal with this issue but
7b6ba9c468f26bdb3492d5e8cb79628a3b04e8c8wrowe later recanted. See this message to dev@apr.apache.org:
77edcaaccc089335938f3844b752e1044f7eb278trawick Message-ID: <Pine.LNX.4.44.0203011354090.16457-200000@deepthought
77edcaaccc089335938f3844b752e1044f7eb278trawickPRs that have been suspended forever waiting for someone to
59b1b6c3fd51c83c3bb9f02a8f08751335f9fb1dminfrinput them into 'the next release':
59b1b6c3fd51c83c3bb9f02a8f08751335f9fb1dminfrin * PR#76: general
59b1b6c3fd51c83c3bb9f02a8f08751335f9fb1dminfrin missing call to "setlocale();"
9f1a88897168c3f1e5009acb585daf01e38a0299jim * PR#78: mod_include
9f1a88897168c3f1e5009acb585daf01e38a0299jim Additional status for XBitHack directive
616a448c1fca1648622707df5a1aae7316bb3d5djim * PR#362: mod_proxy
616a448c1fca1648622707df5a1aae7316bb3d5djim Mod_proxy doesn't allow change of error pages
5a8bb5948d2a258145174320587706de3219d8a3pquerna * PR#370: mod_env
5a8bb5948d2a258145174320587706de3219d8a3pquerna Modified PATH environemnt variable is not passed, instead
5a8bb5948d2a258145174320587706de3219d8a3pquerna system's is used
e5abee85fe751fc27c5d4fc9a8ebe3b80f3d6603trawick * PR#440: mod_proxy
e5abee85fe751fc27c5d4fc9a8ebe3b80f3d6603trawick Proxy doesn't deliver documents if not connected
ca3a25d12c5fe0926aa97550be39f0d88d0decb1pquerna * PR#534: mod_proxy
ca3a25d12c5fe0926aa97550be39f0d88d0decb1pquerna proxy converts ~name to %7Ename when name starts with a dot (.)
74def8815c725f8128a4e76ab1f5704df80b024ajerenkrantz * PR#537: mod_access
4d553781254e46f1dfc8d86b79667a74fb8a3eb5brianp mod_access syntax allows hosts that should be restricted
8f868f15482c7406db01216b6e4778ddabb26898trawick * PR#557: mod_auth-any
8f868f15482c7406db01216b6e4778ddabb26898trawick ~UserHome directories are not honored in absolute pathname
8f868f15482c7406db01216b6e4778ddabb26898trawick requests (.htaccess)
f902601ea431a9b56106e0f5f641dd5fd7efbc30jorton * PR#612: mod_proxy
f902601ea431a9b56106e0f5f641dd5fd7efbc30jorton Proxy FTP Authentication Fails
1e1e4f4f810b99732f06fc05141f42ca1965a9edpquerna * PR#623: mod_include
1e1e4f4f810b99732f06fc05141f42ca1965a9edpquerna A smarter "Last Modified" value for SSI documents (see PR number 600)
78fcc425fc9fc58202a1693fe40dd0ce75c031ffjorton * PR#628: config
78fcc425fc9fc58202a1693fe40dd0ce75c031ffjorton Request of "Options SymLinksIfGroupMatch"
d2ffb32434f79782ff7a364ffa31064698c5c645jorton * PR#700: mod_proxy
d2ffb32434f79782ff7a364ffa31064698c5c645jorton Proxy doesn't do links right for OpenVMS files through ftp:
01847067cfc639c14e1aa77d3b3e98f239447a12jerenkrantz * PR#759: mod_imap
01847067cfc639c14e1aa77d3b3e98f239447a12jerenkrantz imap should read <MAP><AREA>*</MAP> too!
9fe74ffcdea85800f04a7222f716f78ae60cce51jerenkrantz * PR#793: general
9fe74ffcdea85800f04a7222f716f78ae60cce51jerenkrantz RLimitCPU and RLimitMEM don't apply to all children like they should
dae3cb64cc6681b5f6b0fd12e7f8f6296ffaa19abnicholes * PR#921: suexec
d2ffb32434f79782ff7a364ffa31064698c5c645jorton Uses cwd before filling it in, doesn't use syslog
d2ffb32434f79782ff7a364ffa31064698c5c645jorton * PR#922: config
81540a0eb1da2c687e22de3367d8ded55e17e330pquerna it is useful to allow specifiction that root-owned symlinks
ce66c67eba79a20118d8664b65b0c9eeec1bebdabnicholes should always be followed
ce66c67eba79a20118d8664b65b0c9eeec1bebdabnicholes * PR#980: mod_proxy
ce66c67eba79a20118d8664b65b0c9eeec1bebdabnicholes Controlling Access to Remote Proxies would be nice...
6cfc2ed5a8e633c5a40fec65775868d53952d992trawick * PR#994: mod_proxy
6cfc2ed5a8e633c5a40fec65775868d53952d992trawick Adding authentication "on the fly" through the proxy module
6cfc2ed5a8e633c5a40fec65775868d53952d992trawick * PR#1004: apache-api
5159c40648868a58745aa11981f706948ff0f0d1pquerna request_config field in request_rec is moderately bogus
c178b761acd6bffa199c2fd28c4469492b989699nd * PR#1028: other
9174efb969475801d0dc88eee35aae40c748d450nd DoS attacks involving memory consumption
9c6bbd67082b5a47fb17cfa5b61b8a7e1fb01875pquerna * PR#1050: mod_log-any
9c6bbd67082b5a47fb17cfa5b61b8a7e1fb01875pquerna Logging of virtual server to error_log as well
572f5b8a84bb399e51b02a562776f4aec119aa95pquerna * PR#1085: mod_proxy
572f5b8a84bb399e51b02a562776f4aec119aa95pquerna ProxyRemote make a dead cycle.
1b03ca18c41f51a25dcf1a623a8f558bd779e0a4jerenkrantz * PR#1117: mod_auth-any
1b03ca18c41f51a25dcf1a623a8f558bd779e0a4jerenkrantz Using NIS passwd.byname dbm files with AuthDBMUserFile
1b03ca18c41f51a25dcf1a623a8f558bd779e0a4jerenkrantz * PR#1120: suexec
1b03ca18c41f51a25dcf1a623a8f558bd779e0a4jerenkrantz suexec does not parse arguments to #exec cmd
db443e0132f14dac789ab97ec23ce124360d74c2nd * PR#1145: mod_include
db443e0132f14dac789ab97ec23ce124360d74c2nd Allow for Last-Modified: without resorting to XBitHack
4ac7a7c09ac5732b09f8bf28873f7e9efcab34d5bnicholes * PR#1158: apache-api
4ac7a7c09ac5732b09f8bf28873f7e9efcab34d5bnicholes improvements to child spawning API
4ac7a7c09ac5732b09f8bf28873f7e9efcab34d5bnicholes * PR#1166: mod_proxy
4ac7a7c09ac5732b09f8bf28873f7e9efcab34d5bnicholes ``nph-'' not honored (no buffering) for ProxyRemote mapping
ce8490f3812311582d1deee96b012c377311b317minfrin * PR#1176: mod_cgi
ce8490f3812311582d1deee96b012c377311b317minfrin Apache cannot handle continuation line in headers
38f6ebaca968b7b23c25c0b30d0be1c7aad7412bjorton * PR#1191: general
38f6ebaca968b7b23c25c0b30d0be1c7aad7412bjorton setlogin() is not called, causing problems with e.g. identd
cc7d8b55b16eee88be925a090473ca94b0a6e770jorton * PR#1204: general
f2b43354f9ac8496ab7003ae01211af739efba1atrawick regerror() exists, use it
2555a6b5da21d61804f47084d8fcc98eb4acbc42wrowe * PR#1233: apache-api
2555a6b5da21d61804f47084d8fcc98eb4acbc42wrowe there is no way to keep per-connection per-module state
2555a6b5da21d61804f47084d8fcc98eb4acbc42wrowe * PR#1263: mod_autoexec
2555a6b5da21d61804f47084d8fcc98eb4acbc42wrowe Add frame-safe anchor attribute to mod_autoindex links
eb6e3a3e784021abc1e28247158db3a8b07186b9minfrin * PR#1268: suexec
eb6e3a3e784021abc1e28247158db3a8b07186b9minfrin CGI scripts running as Apache user: security (suexec etc.)
f5098e2eb6ed9dfd7052f25761bb5e97779b8460minfrin * PR#1285: suexec
f5098e2eb6ed9dfd7052f25761bb5e97779b8460minfrin Error messages could be easier to spot in cgi.log file for suexec.c
d86551ca954d666d5b15ad38e9daf45680516c25jerenkrantz * PR#1287: mod_access
c79b5caf7c5462ef707363c69cc63ea2606310abtrawick add allow,deny/deny,allow warning to mod_access
1551c2366551d7a361bedbf7ae54f1abd74ef822minfrin * PR#1290: mod_proxy
1551c2366551d7a361bedbf7ae54f1abd74ef822minfrin Need to know "hit-rate" on proxy cache
f1826b4cf1fbeb049c569761cc95b04e3496efd6bnicholes * PR#1358: mod_log-any
f1826b4cf1fbeb049c569761cc95b04e3496efd6bnicholes Selective url-encode of log fields (or maybe a pseudo
f1826b4cf1fbeb049c569761cc95b04e3496efd6bnicholes log_rewrite module?)
103a93c625bcde1a6a7a5155b64dcda36f612180pquerna * PR#1383: mod_headers
103a93c625bcde1a6a7a5155b64dcda36f612180pquerna I make mod_headers to modify request headers as well as
87a528a7622973988232079ba02763748c9c7071pquerna response ones.
1a986bbad9314beb8739401cac822e87bb04bbfepquerna * PR#1532: mod_proxy
1a986bbad9314beb8739401cac822e87bb04bbfepquerna Proxy transfer logging
1a986bbad9314beb8739401cac822e87bb04bbfepquerna * PR#1547: mod_proxy
752ec04fdfa6e936047811b9e506cf829d7a69f7pquerna No HTTP_X_FORWARDED_FOR set...
752ec04fdfa6e936047811b9e506cf829d7a69f7pquerna * PR#1567: mod_proxy
cf82d17373559b6163dbd6c844c1af82cab56883niq ProxyRemote proxy requests fail authentication by firewall
cf82d17373559b6163dbd6c844c1af82cab56883niq * PR#1582: mod_rewrite
c4e6f5320f87eff25838c21c107a24b688d539d0niq mod_rewrite forms REQUEST_URI different than mod_cgi does
c4e6f5320f87eff25838c21c107a24b688d539d0niq * PR#1677: mod_headers
c4e6f5320f87eff25838c21c107a24b688d539d0niq mod_headers should allow mod_log_config-style formats in
c4e6f5320f87eff25838c21c107a24b688d539d0niq header values
174a8f105b1c11aa50a05993ea8ba2d117d3df6end * PR#1702: mod_proxy
174a8f105b1c11aa50a05993ea8ba2d117d3df6end mod_proxy to support persistent conns?
3ad0f071604de33c0b3733b649cd0445ba00f54dake * PR#1803: mod_include
3ad0f071604de33c0b3733b649cd0445ba00f54dake patches to mod_include to allow for file tests
9174efb969475801d0dc88eee35aae40c748d450nd * PR#1809: mod_auth-any
c2549f0b237ac86f3623a601a766969d805dbc2and Suggestion for improving authentication modules and core source
7efe7de73c89c26518714a504359244d03cfbbc5jorton code, problem with 401 and ErrorDocument
f84d3d83a741c21154d42e0ebdec9b9b37efeedcjorton * PR#1878: mod_proxy
7efe7de73c89c26518714a504359244d03cfbbc5jorton listing of proxy cache content
d60027cde3289d3ef35bc9d815882975eb60e6abjorton * PR#1905: suexec
d60027cde3289d3ef35bc9d815882975eb60e6abjorton Allow modules to set user:group for execution.
504498fd20729dcc29dfbfb78e64a8521c6a767ajerenkrantz * PR#2024: apache-api
d406be0d6a4f9e61bbdfa6c8cde2e09338dd650atrawick adding auth_why to conn_rec
269e0a31aa0f6aafe7292e932a182ccb535e55a9trawick * PR#2073: mod_log-any
c1ba3b0ac00c4fee2f4b950dfdb167b39611b661nd pipelined connections are not logged correctly
b95a84193578c904426cef6dda84f7118a400a16jim * PR#2074: mod_rewrite
b95a84193578c904426cef6dda84f7118a400a16jim mod_rewrite doesn't pass Proxy Throughput on internal subrequests
03a3ed87983471816561562f957390ed935d7b3bnd * PR#2113: config
03a3ed87983471816561562f957390ed935d7b3bnd HTTP Server Rebuild Line Needs Changing for the better
c533ecac2227dc228070e686fb14dc6860f497f8nd * PR#2138: mod_status
56cefde0af6b8db6fda0f1d95d8cdca54f397cd0nd mod_status always displays 256 possible connection slots
41369ed0bc7f2db6272278c27025f6aabf97fe63nd * PR#2221: documentation
9174efb969475801d0dc88eee35aae40c748d450nd Make online documentation search link back to my installation
496f8f3966319d43455675630a849bae019d2a32nd * PR#2284: general
9e8c2603790f490398a0fabf97866b6815748a54ianh Can not POST to ErrorDocument - Apache/1.3b6
9e8c2603790f490398a0fabf97866b6815748a54ianh * PR#2314: mod_proxy
a21b3b9d8ebb12fd51fa1d17e44d5644a35a9a5fnd patterns in ProxyRemote
1fbcf7a8e7b18be1b0e77b7bc38b7b71f8dfb052nd * PR#2343: mod_status
9174efb969475801d0dc88eee35aae40c748d450nd Status module averages are for entire uptime
f769c33501f474aed3e0f7c769477c8c4f478783geoff * PR#2360: suexec
f769c33501f474aed3e0f7c769477c8c4f478783geoff suexec for general access of user content?
d7c7669331357296719d67d1963d40d713ed455atrawick * PR#2396: general
d7c7669331357296719d67d1963d40d713ed455atrawick Proposal for TimeZone directive
92d95be777d4365eb79444a7a558355a7a92081ajorton * PR#2415: mod_info
92d95be777d4365eb79444a7a558355a7a92081ajorton /server-info doesn't check for the virtual host to list the info
af8dee354a287249dd9f3f77bbe850108e5afe43trawick * PR#2421: config
af8dee354a287249dd9f3f77bbe850108e5afe43trawick problem specifying ndbm library for build ?with autoconfigure
14763a0db22322626dd8cd59dfbc3a4fcc655d99trawick * PR#2431: general
f78505c7d260473bf11002f5028186f27d0ed8a0geoff A small addition to rotatelogs.c to improve program functionality.
f78505c7d260473bf11002f5028186f27d0ed8a0geoff * PR#2446: config
f78505c7d260473bf11002f5028186f27d0ed8a0geoff AllowOverride FileInfo is too coarse
ea981941b1e5ecfb1a27b9e9aff32b9c98640ed9nd * PR#2460: mod_cgi
ea981941b1e5ecfb1a27b9e9aff32b9c98640ed9nd TimeOut applies to output of CGI scripts
d57b793ddd42a997c2cb3bd389bbcf1161226fd8trawick * PR#2512: mod_access
d57b793ddd42a997c2cb3bd389bbcf1161226fd8trawick <IfDenied> directive wanted
9174efb969475801d0dc88eee35aae40c748d450nd * PR#2573: suexec
d11bf78168abbb07d8ed9f54d7ea0953d46aad83nd CGI's for general use still have to be run as another user
bf28a00d8218aa76a56eddabca308e8d0e691626nd with suExec
9174efb969475801d0dc88eee35aae40c748d450nd * PR#2648: general
bf28a00d8218aa76a56eddabca308e8d0e691626nd Cache file names in Proxy module
70f665a8945f3bdfe8bbeaa063de2aa722fc9b29nd * PR#2760: config
c08f5a0cf2b8903255186ba998a4481dfac9c796nd [PATCH] User/Group for <Directory> and <Location> i.e. not only
c08f5a0cf2b8903255186ba998a4481dfac9c796nd in global and <Virtual>.
5bcdf2cd27ab0c91a3c7eaffbff4cb8505389dc1nd * PR#2763: general
9174efb969475801d0dc88eee35aae40c748d450nd mailto tags and bundling bug report script
9174efb969475801d0dc88eee35aae40c748d450nd * PR#2785: os-aix
02b0660d7f44ff8106bd5095653e9c19e39ec37dnd Support for System Resource Controller
5b8e35ad88268c0210d93288dad57c2f1d3e8811nd * PR#2793: protocol
a9ee8e9bd3dffd23ca49be8d0bdf0e33cd0bcce2jorton When will Apache support P3P? Any Plans?
3b86be5b30d5cbacc1f942b05dff8a9365449d30jorton * PR#2873: config
9026fcf39ea5fbacee72b6c4f68e6c778dbad478nd * PR#2889: general
8aad5258d3ecde3751559bc685a3185e63cbde9aianh Inclusion of RPM spec file in CVS/distributions
71f3601de4983bc2a6aaffcf37dc1d35c8674a34coar * PR#2906: general
71f3601de4983bc2a6aaffcf37dc1d35c8674a34coar Propose that Apache recommend $UNIQUE_ID for all "session id"
71f3601de4983bc2a6aaffcf37dc1d35c8674a34coar algorithms
49a82db11388cff9b29ff861b4241bbce69c76fdtrawick * PR#2907: config
49a82db11388cff9b29ff861b4241bbce69c76fdtrawick suggestion: power up your Include directive :)
f23c50b3628a8571da2ff3277ae58ef1d675554ftrawick * PR#3018: general
f23c50b3628a8571da2ff3277ae58ef1d675554ftrawick cannot limit some HTTP methods
8dc154408549195c828b823e9dc7396f107f2512jorton * PR#3143: apache-api
ca2504b59d48a926af23f6b18af550c1e892d8a6jorton No module specific data hook for per-connection data
6126fad8adeca94e9813812c691747afeca164dftrawick * PR#3191: mod_negotiation
6126fad8adeca94e9813812c691747afeca164dftrawick no way to set global quality-of-source (qs) coneg values
6126fad8adeca94e9813812c691747afeca164dftrawick with multiviews
7ed5992392d5babab20be4ce4029ad17ae062b5aianh * PR#3568: mod_proxy
b7973a1e572a3ff7cdedb48ed1f9481ece700bf9trawick Accessing URL through proxy server corrupts data.
5c4e29c5f77fbc967d78886ab378b9500267b0fbtrawick * PR#3605: mod_proxy
5c4e29c5f77fbc967d78886ab378b9500267b0fbtrawick Some anonymous FTP URLs ask for authentication
1fbf6ba0f5207e6637b49f9a9dfcc779bbe952a9trawick * PR#3677: general
6c4c113ce85934b11c9e78399e1bb8ec7a568af9nd New ErrorDocumentMatch directive
9174efb969475801d0dc88eee35aae40c748d450nd * PR#4241: config
6c4c113ce85934b11c9e78399e1bb8ec7a568af9nd Need to be able to override shebang line to make CGI scripts
81ef3d132c8f335006465f297c42319f8734825dnd more portable.
73291de2e17c97d3549f2f8aa085ca43d27e108and * PR#4244: config
73291de2e17c97d3549f2f8aa085ca43d27e108and "Files" and "FilesMatch" regexp does not recognize bang as
9174efb969475801d0dc88eee35aae40c748d450nd negation operator
3caed8f01e7d67f3ad5d47cb28bb4e43131652fdjerenkrantz * PR#4448: mod_log-any
3caed8f01e7d67f3ad5d47cb28bb4e43131652fdjerenkrantz Please allow CGI env variables (QUERY_STRING, ...) to be logged
e127d82e8afc2f885fe2183b68d4a110580d21dfjwoolley * PR#4459: mod_include
e127d82e8afc2f885fe2183b68d4a110580d21dfjwoolley Suggestion for better handling of Last-modified headers
fd99717a52886f7d9f9f73be2080d3fdc2ec4bd6nd * PR#4490: mod_cgi
9174efb969475801d0dc88eee35aae40c748d450nd mod_cgi prevents handling of OPTIONS requests
9174efb969475801d0dc88eee35aae40c748d450nd * PR#5713: os-windows
5fb2e8564729922524cd011b11e2d4f5d76f5a8dnd [PATCH] install as win32 service with domain account
987c2c2ff2f19f306357f79d3280b347d96c470fnd Status: Cannot accept password-as-arg, we should prompt the
987c2c2ff2f19f306357f79d3280b347d96c470fnd user when -k install/-k config with a user argument.
987c2c2ff2f19f306357f79d3280b347d96c470fnd * PR#5993: general
a4dd3688dd6645faf0c1c1bfb22017c8f03d5b24nd AllowOverride should have a 'CheckNone' and 'AllowNone' argument
9174efb969475801d0dc88eee35aae40c748d450nd instead of only 'None'
f2fe083bb9fabd2af7eb39d4f154f27a451bb5a4ndOther bugs that need fixing:
f2fe083bb9fabd2af7eb39d4f154f27a451bb5a4nd * MaxRequestsPerChild measures connections, not requests.
7de6cb79f71a9007ba2b1e786cbad8b84f05d101nd Until someone has a better way, we'll probably just rename it
7de6cb79f71a9007ba2b1e786cbad8b84f05d101nd "MaxConnectionsPerChild".
7de6cb79f71a9007ba2b1e786cbad8b84f05d101nd * Regex containers don't work in an intutive way
aa9b03a5f32732c0caaef03a7ed78ffb290e29e4trawick Status: No one has come up with an efficient way to fix this
36f2369ee230c310fed3b600e9e99cf5770804a3wrowe behavior. Dean has suggested getting rid of regex containers
aa9b03a5f32732c0caaef03a7ed78ffb290e29e4trawick completely.
ab56518227d474ee08f039e4c5540011c1b8a913trawick OtherBill suggests: We at least seem to agree on eliminating
36f2369ee230c310fed3b600e9e99cf5770804a3wrowe the <Container ~ foo> forms, and using only
ab56518227d474ee08f039e4c5540011c1b8a913trawick <ContainerMatch foo> semantics.
36f2369ee230c310fed3b600e9e99cf5770804a3wrowe * SIGSEGV on Linux (glibc 2.1.2) isn't caught properly by a
a18a308d7a942230edcde1bf00d7ae9b4b143c90trawick sigwaiting thread. We need to work around this, perhaps unless
3b872593fd5f61981d9dd69a4b0b5d5f5f668929trawick there is hope soon for a fixed glibc.
3b872593fd5f61981d9dd69a4b0b5d5f5f668929trawick * orig_ct in the byterange/multipart handling may not be
fc25339741311efd7d460f18b6287ef38d76bbe6madhum needed. Apache 1.3 just never stashed "multipart" into
fc25339741311efd7d460f18b6287ef38d76bbe6madhum r->content_type. We should probably follow suit since the
36f2369ee230c310fed3b600e9e99cf5770804a3wrowe byterange stuff doesn't want the rest of the code to see the
fc25339741311efd7d460f18b6287ef38d76bbe6madhum multipart content-type; the other code should still think it is
fcdca175a52fe517f2317ba0e2b6e6d14522b869madhum dealing with the <orig_ct> stuff.
19fdbc5566bf67dde644be9e8d38d62db4dd0ba5jerenkrantz Status: Greg volunteers to investigate (esp. since he was most
92a2439559cf1161742650ed9c50c6483bd029cemadhum likely the one to break it :-)
36f2369ee230c310fed3b600e9e99cf5770804a3wroweOther features that need writing:
ebecc16986604cce1369d5075eff65032e3dd0deianh * Finish infrastructure in core for async MPMs
9174efb969475801d0dc88eee35aae40c748d450nd Status: post 2.0
764315969cef40e50cdc6a5e9638454e10c1c06end * TODO in source -- just do an egrep on "TODO" and see what's there
9174efb969475801d0dc88eee35aae40c748d450ndAvailable Patches:
b92cba59a0890be43b14aaf1ce30606140be9593nd * Jon Travis's <jtravis@covalent.net> patch to deal with thread-safe
b92cba59a0890be43b14aaf1ce30606140be9593nd issues with inet_ntoa. See message <20001201163220.A12827@covalent.net>
9174efb969475801d0dc88eee35aae40c748d450nd Status: This is being set aside until the IPv6 work is finished
b92cba59a0890be43b14aaf1ce30606140be9593nd so that we know exactly what is required.
402d23baca89e8c4fcb4e52ad8b2d66a6904baaetrawick * Martin Sojka <msojka@gmx.de>'s patch to add error reporting for failed
402d23baca89e8c4fcb4e52ad8b2d66a6904baaetrawick htpasswd actions due to a full /tmp volume (other programs may have
4caa28863a3418d26cc20a998dc368c3de3b7e19jerenkrantz similar problems?)
a3f2646ef3d8a3a5234a5601de0f95f10308c2a6jerenkrantz * Mike Abbott's <mja@trudge.engr.sgi.com> patches to improve
9e398d701dd430f073ff5418fb720642e064046ajerenkrantz Status: These were written for 1.3, and are awaiting a port to
1a5b9e0071f0c662036250b482d566ad87ff0b4bjerenkrantz * Jim Winstead's <jimw@trainedmonkey.com> patch to add CookieDomain and
1a5b9e0071f0c662036250b482d566ad87ff0b4bjerenkrantz other small mod_usertrack features
a7ac9b52c3d9f7ce937f078a0d585023db626c55jerenkrantz * Dan Rench's <drench@xnet.com> patch to add allow the errmsg and timefmt
a7ac9b52c3d9f7ce937f078a0d585023db626c55jerenkrantz of SSI's to be modified in the config file. Patch is available in
db5837bbc9bef214303e755fa52122140366cb6fianhOpen issues:
aac2b82fe4f1ac117e2a0702438d6615542642dand * Which MPMs will be included with Apache 2.0?