STATUS revision 106c2cbb223e2388d5470b6c3c34a1f0e78ef94c
f743002678eb67b99bbc29fee116b65d9530fec0wroweAPACHE 2.0 STATUS: -*-text-*-
80833bb9a1bf25dcf19e814438a4b311d2e1f4cffuankgLast modified at [$Date: 2002/03/08 14:32:55 $]
4d9aacb99fa81594665b3024fd15adcad28d2e60ylavic 2.0.33 : in development
4d9aacb99fa81594665b3024fd15adcad28d2e60ylavic 2.0.32 : released Feburary 16, 2002.
4d9aacb99fa81594665b3024fd15adcad28d2e60ylavic 2.0.31 : rolled Feburary 1, 2002. not released.
4f29b65ab4b547ad5dbe506e2d0ff5d12ead9247ylavic 2.0.30 : tagged January 8, 2002. not rolled.
4f29b65ab4b547ad5dbe506e2d0ff5d12ead9247ylavic 2.0.29 : tagged November 27, 2001. not rolled.
e7a83ac97aed916dc6da9441f24d5630caad249fjailletc 2.0.28 : released November 13, 2001
e7a83ac97aed916dc6da9441f24d5630caad249fjailletc 2.0.27 : rolled November 6, 2001
e7a83ac97aed916dc6da9441f24d5630caad249fjailletc 2.0.26 : tagged October 16, 2001. not rolled.
e7a83ac97aed916dc6da9441f24d5630caad249fjailletc 2.0.25 : rolled August 29, 2001
e7a83ac97aed916dc6da9441f24d5630caad249fjailletc 2.0.24 : rolled August 18, 2001
e7a83ac97aed916dc6da9441f24d5630caad249fjailletc 2.0.23 : rolled August 9, 2001
e7a83ac97aed916dc6da9441f24d5630caad249fjailletc 2.0.22 : rolled July 29, 2001
f4c3ded5dd561f8aa9779f227fed41303c15efaaylavic 2.0.21 : rolled July 20, 2001
f4c3ded5dd561f8aa9779f227fed41303c15efaaylavic 2.0.20 : rolled July 8, 2001
f4c3ded5dd561f8aa9779f227fed41303c15efaaylavic 2.0.19 : rolled June 27, 2001
0a0df13b7f1f4f1a74fe295253d89ca3911b301aylavic 2.0.18 : rolled May 18, 2001
0a0df13b7f1f4f1a74fe295253d89ca3911b301aylavic 2.0.17 : rolled April 17, 2001
0a0df13b7f1f4f1a74fe295253d89ca3911b301aylavic 2.0.16 : rolled April 4, 2001
0a0df13b7f1f4f1a74fe295253d89ca3911b301aylavic 2.0.15 : rolled March 21, 2001
206ee035a75fa226ca51d6ee1f118624354803d2ylavic 2.0.14 : rolled March 7, 2001
206ee035a75fa226ca51d6ee1f118624354803d2ylavic 2.0a9 : released December 12, 2000
206ee035a75fa226ca51d6ee1f118624354803d2ylavic 2.0a8 : released November 20, 2000
206ee035a75fa226ca51d6ee1f118624354803d2ylavic 2.0a7 : released October 8, 2000
206ee035a75fa226ca51d6ee1f118624354803d2ylavic 2.0a6 : released August 18, 2000
69301145375a889e7e37caf7cc7321ac0f91801erpluem 2.0a5 : released August 4, 2000
69301145375a889e7e37caf7cc7321ac0f91801erpluem 2.0a4 : released June 7, 2000
69301145375a889e7e37caf7cc7321ac0f91801erpluem 2.0a3 : released April 28, 2000
506bfe33206b2fece40ef25f695af39dd4130facjkaluza 2.0a2 : released March 31, 2000
506bfe33206b2fece40ef25f695af39dd4130facjkaluza 2.0a1 : released March 10, 2000
506bfe33206b2fece40ef25f695af39dd4130facjkaluzaPlease consult the following STATUS files for information
78ae177952e97903c19115e0f1b0383a5ca22eb4ylavicon related projects:
d58a848a016d401b965111e50ef829e1641f7834minfrinCURRENT RELEASE NOTES:
25b59a7ba7fb7ce658ddb3fc31531b0ce414283cylavic * 33 status: tag on 3/7/2002? Ideally, this release should have all
d06d6ad3304fda986f67b5ada248e9ac05848cd9ylavic showstoppers removed. Are we dreaming? Can the horde
33a50f7ac69bc7f9f5c35cc7f0e240ec324ae660minfrin of developers actually focus on removing them in a week?
33a50f7ac69bc7f9f5c35cc7f0e240ec324ae660minfrin (Note: Win32 had its .32 yanked due to a mmap() bug.)
dd993bccb903f6f0b256c4dc973b793848986ef2minfrinFINAL RELEASE SHOWSTOPPERS:
dd993bccb903f6f0b256c4dc973b793848986ef2minfrin * If any request gets to the core handler, without a flag that this
dd993bccb903f6f0b256c4dc973b793848986ef2minfrin r->filename was tested by dir/file_walk, we need to 500 at the very
2e6f4d654c96c98b761fb012fd25c5d5b1558c44sf end of the ap_process_request_internal() processing. This provides
2e6f4d654c96c98b761fb012fd25c5d5b1558c44sf authors of older modules better compatibility, while still improving
2e6f4d654c96c98b761fb012fd25c5d5b1558c44sf the security and robustness of 2.0.
4e10eefaad0287c0c4ac2185bc3a1615092dbef8ylavic Status: still need to decide where this goes, OtherBill comments...
4e10eefaad0287c0c4ac2185bc3a1615092dbef8ylavic Message-ID: <065701c14526$495203b0$96c0b0d0@roweclan.net>
4e10eefaad0287c0c4ac2185bc3a1615092dbef8ylavic we need to look at halting this in the 'default handler' case,
4e10eefaad0287c0c4ac2185bc3a1615092dbef8ylavic and that implies pushing the 'handler election' into the request
17e6c95f3b22d18acdf8380fb26a8d0e10c80767ylavic internal processing phase from the run request phase.
17e6c95f3b22d18acdf8380fb26a8d0e10c80767ylavic Jim asks: would a stopgap be something bogus like adding another
17e6c95f3b22d18acdf8380fb26a8d0e10c80767ylavic flag to request_rec ala eos_sent and before we OK, if not set
e8bd80a4bb88199d2f9a24a50345688e52d9c116ylavic * We need to find out the right place to add the AddOutputFilterByType
e8bd80a4bb88199d2f9a24a50345688e52d9c116ylavic directive. mod_mime? No. core with fixups? That's what it is now.
e8bd80a4bb88199d2f9a24a50345688e52d9c116ylavic ap_pass_brigade() hook? As a filter that runs at HTTP_HEADER stage?
ac7482772673c7dab4afbf53649e07030fd57a8eylavic This needs to be resolved before the next public release.
8c94893229334f5b79636e4ec53a992a0b9f9828ylavic Message-ID: <018701c1b99c$f0f24310$0a01230a@KOJ>
ac7482772673c7dab4afbf53649e07030fd57a8eylavic * API changes planned for 2.0 that should happen before the
330e16bea8fe9cace4de90c349750c03dfb1fe64ylavic GA release:
330e16bea8fe9cace4de90c349750c03dfb1fe64ylavic * Free lists for bucket allocation
330e16bea8fe9cace4de90c349750c03dfb1fe64ylavic * Pool allocator change
330e16bea8fe9cace4de90c349750c03dfb1fe64ylavic * Address popular PRs
330e16bea8fe9cace4de90c349750c03dfb1fe64ylavic * Win32 doesn't install as service correctly [9863, 9914, 9961]
330e16bea8fe9cace4de90c349750c03dfb1fe64ylavic * Don't be stupid and cd to a blank directory when doing installs
d7205b1a86c51c27b71a2c458dc453fd53a261c1covenerCURRENT VOTES:
d7205b1a86c51c27b71a2c458dc453fd53a261c1covener * Should we always build binaries statically unless otherwise
44ff304057225e944e220e981d434a046d14cf06covener Message-ID: <20020129210006.B23512@Lithium.MeepZor.Com>
291eb44b3adaf8247425286615b4f4b69fbea274minfrin -1: Justin, Ian
291eb44b3adaf8247425286615b4f4b69fbea274minfrin * If the parent process dies, should the remaining child processes
5d1ba75b8794925e67591c209085a49279791de9covener "gracefully" self-terminate. Or maybe we should make it a runtime
5d1ba75b8794925e67591c209085a49279791de9covener option, or have a concept of 2 parent processes (one being a
5d1ba75b8794925e67591c209085a49279791de9covener "hot spare").
032982212dbcc7c3cce95bf89c503bb56e185ac7kbrand See: Message-ID: <3C58232C.FE91F19F@Golux.Com>
032982212dbcc7c3cce95bf89c503bb56e185ac7kbrand Self-destruct: Ken, Martin
032982212dbcc7c3cce95bf89c503bb56e185ac7kbrand Not self-destruct: BrianP, Ian, Cliff, BillS
caad2986f81ab263f7af41467dd622dc9add17f3ylavic Make it runtime configurable: Aaron, Jim, Justin
caad2986f81ab263f7af41467dd622dc9add17f3ylavic Have 2 parents: +1: Jim
caad2986f81ab263f7af41467dd622dc9add17f3ylavic +0: Martin (while standing by, could it do
45a10d38e6051fd7bdf9d742aaae633d97ff02abjailletc something useful?)
f7317ff316c2b141feea31bddb74d5d3fa1584edjorton * Make the worker MPM the default MPM for threaded Unix boxes.
2165214331e4afafca4048f66f303d0253d7b001covener +1: Justin, Ian, Cliff
a34684a59b60a4173c25035d0c627ef17e6dc215rpluem -0: Aaron (premature decision, needs more discussion), Lars
1e2d421a36999d292042a5539971070d54aa6c63ylavicRELEASE NON-SHOWSTOPPERS BUT WOULD BE REAL NICE TO WRAP THESE UP:
1e2d421a36999d292042a5539971070d54aa6c63ylavic * Convert all instances of the old apr_lock_t type to the new
fa7ed98b9dc94c5845cf845aea0a44ecacd290c9humbedooh Status: Aaron has converted all but the perchild MPM to
fa7ed98b9dc94c5845cf845aea0a44ecacd290c9humbedooh use the new lock API. Since perchild has been put
0b67eb8568cd58bb77082703951679b42cf098actrawick on the back burner, this is no longer a showstopper.
0b67eb8568cd58bb77082703951679b42cf098actrawick Aaron will patch perchild as soon as it becomes
0b67eb8568cd58bb77082703951679b42cf098actrawick buildable again.
06bb3e11d3e997937534ae7bd45b3631d3b5e5bacovener * With AP_MODE_EXHAUSTIVE in the core, it is finally clear to me
06bb3e11d3e997937534ae7bd45b3631d3b5e5bacovener how the Perchild MPM should be re-written. It hasn't worked
2165214331e4afafca4048f66f303d0253d7b001covener correctly since filters were added because it wasn't possible to
06bb3e11d3e997937534ae7bd45b3631d3b5e5bacovener get the content that had already been written and the socket at
5ef3c61605a3a021ff71f488983cb0065f8e1a79covener the same time. This mode lets us do that, so the MPM can be
6502b7b32f980cc2093bb3ebce37e5e4dc68fba4ylavic * htpasswd blindly processes the file you give it, and does no
6502b7b32f980cc2093bb3ebce37e5e4dc68fba4ylavic sanity checking before totally corrupting whatever file it was
3060ce7f798fbda7999cd4ddf89b525d2b294185covener you thought you had. It should check the input file and bail
c85eff31536e6bfef1537b2435564d48665435d3rpluem if it finds non-comment lines that do not contain exactly 1
c85eff31536e6bfef1537b2435564d48665435d3rpluem ':' character.
c85eff31536e6bfef1537b2435564d48665435d3rpluem Message-ID: <20020217150457.A31632@clove.org>
c1a63b8fad09c419c1a64f75993feb8a343a6801ylavic * Can a static httpd be built reliably?
c1a63b8fad09c419c1a64f75993feb8a343a6801ylavic Message-ID: <20020207142751.T31582@clove.org>
e6b4bd1113567627ab6bb6c6a7105e1e01a7d889jailletc * [Ken] Test suite failures:
e6b4bd1113567627ab6bb6c6a7105e1e01a7d889jailletc o worker is also failing some of the 'cgi' subtests
e466c40e1801982602ee0200c9e8b61cc148742djailletc (see <URL:http://Source-Zone.Org/Apache/regression/>):
e466c40e1801982602ee0200c9e8b61cc148742djailletc Justin says: "Worker should be fine and passes httpd-test here.
457468b82e59d01eba00dd9d0817309c8f5e414ejim If you can provide evidence that it can be reproduced
457468b82e59d01eba00dd9d0817309c8f5e414ejim outside of httpd-test, then it's a showstopper. I
457468b82e59d01eba00dd9d0817309c8f5e414ejim think it's a perl or a httpd-test problem."
04983e3bd1754764eec7d6bb772fe3b0bf391771jorton Not a showstopper: Justin
15890c9306ba98f6fc243e15a3c4778ddc7d773erpluem * Usage of APR_BRIGADE_NORMALIZE in core_input_filter should be
15660979a30d251681463de2e0584853890082accovener removed if possible.
15660979a30d251681463de2e0584853890082accovener Message-ID: <Pine.LNX.4.33.0201202232430.318-100000@deepthought.cs.virginia.edu>
15660979a30d251681463de2e0584853890082accovener * There is a bug in how we sort some hooks, at least the pre-config
cfd9415521847b2f9394fad04fb701cfb955f503rjung hook. The first time we call the hooks, they are in the correct
cfd9415521847b2f9394fad04fb701cfb955f503rjung order, but the second time, we don't sort them correctly. Currently,
cfd9415521847b2f9394fad04fb701cfb955f503rjung the modules/http/config.m4 file has been renamed to
28c31fb73c1264bd1d0ff932573677030b024c7dwrowe modules/http/config2.m4 to work around this problem, it should moved
28c31fb73c1264bd1d0ff932573677030b024c7dwrowe back when this is fixed. rbb
28c31fb73c1264bd1d0ff932573677030b024c7dwrowe Justin says: "Is this really a showstopper? This has been here
28c31fb73c1264bd1d0ff932573677030b024c7dwrowe forever. What's wrong? Does this have to do with
28c31fb73c1264bd1d0ff932573677030b024c7dwrowe autoconf or m4?"
8491e0600f69b0405e156ea8a419653c065c645bcovener Not a showstopper: Justin, BrianP, trawick, gregames
63b9f1f5880391261705f696d7d65507bbe9ace3covener * The Add...Filter and Set...Filter directives do not allow the
63b9f1f5880391261705f696d7d65507bbe9ace3covener administrator to order filters, beyond the order of filename (mime)
49dacedb6c387b786b7911082ff35121a45f414bcovener extensions. It isn't clear if Set...Filter(s) should be inserted
49dacedb6c387b786b7911082ff35121a45f414bcovener before or after the Add...Filter(s) which are ordered by sequence of
49dacedb6c387b786b7911082ff35121a45f414bcovener filename extensions. At minimum, some sort of +-[0-10] syntax seems
49dacedb6c387b786b7911082ff35121a45f414bcovener like the quickest fix for a 2.0 gold release.
3c990331fc6702119e4f5b8ba9eae3021aea5265jim Justin says: "Could we delay this for a point release or 2.1?"
3c990331fc6702119e4f5b8ba9eae3021aea5265jim Not a showstopper: justin, wrowe, trawick, stoddard, Jim, Ian, Aaron,
fc42512879dd0504532f52fe5d0d0383dda96a1eniq * Get perchild to work on platforms other than Linux. This
fc42512879dd0504532f52fe5d0d0383dda96a1eniq will require a portable mechanism to pass data and file/socket
fc42512879dd0504532f52fe5d0d0383dda96a1eniq descriptors between vhost child groups. An API was proposed
0451df5dc50fa5d8b3e07d92ee6a92e36a1181a5niq on dev@apr:
0451df5dc50fa5d8b3e07d92ee6a92e36a1181a5niq Message-ID: <20020111115006.K1529@clove.org>
da0442c0440caef34706e2c2f3af05cb65921cc0jailletc * CGI single-byte reads
983528026996668ea295be95aedb9c7a346af470ylavic BrianP suggests that this is caused by the ap_scan_script_header_err()
da0442c0440caef34706e2c2f3af05cb65921cc0jailletc routine, which will do single-byte reads until it finds the end
da0442c0440caef34706e2c2f3af05cb65921cc0jailletc of the header, at which point it constructs a pipe-bucket (buffered)
06b8f183140c8e02e0974e938a05078b511d1603covener to read from.
06b8f183140c8e02e0974e938a05078b511d1603covener Proposed solution in:
06b8f183140c8e02e0974e938a05078b511d1603covener Message-ID: <3C36ADAF.60601@cnet.com>
259878293a997ff49f5ddfc53d3739cbdc25444ecovener * Try to get libtool inter-library dependency code working on AIX.
259878293a997ff49f5ddfc53d3739cbdc25444ecovener Message-ID: <cm3n10lx555.fsf@rdu163-40-092.nc.rr.com>
259878293a997ff49f5ddfc53d3739cbdc25444ecovener Justin says: If we get it working on AIX, we can enable this
15890c9306ba98f6fc243e15a3c4778ddc7d773erpluem on all platforms and clean up our build system
b54b024c06a19926832d77d40ba35ad8c41e4d3dminfrin Jeff says: I thought I tested a patch for you sometime in
b54b024c06a19926832d77d40ba35ad8c41e4d3dminfrin January that you were going to commit within a few
65967d05f839dbf27cf91d91fa79585eeae19660minfrin * Handling of %2f in URIs. Currently both 1.3 and 2.0
65967d05f839dbf27cf91d91fa79585eeae19660minfrin completely disallow %2f in the request URI path (see
8152945ae46857b170cb227e79bb799f4fc7710dminfrin ap_unescape_url() in util.c). It's permitted and passed
8152945ae46857b170cb227e79bb799f4fc7710dminfrin through in the query string, however. Roy says the
8152945ae46857b170cb227e79bb799f4fc7710dminfrin original reason for disallowing it, from five years ago,
8152945ae46857b170cb227e79bb799f4fc7710dminfrin was to protect CGI scripts that applied PATH_INFO to
75f5c2db254c0167a0e396254460de09b775d203trawick a filesystem location and which might be tricked by
75f5c2db254c0167a0e396254460de09b775d203trawick ..%2f..%2f(...). We *should* allow path-info of the
75f5c2db254c0167a0e396254460de09b775d203trawick form 'http://foo.com/index.cgi/path/to/path%2finfo'.
4f0358189bfa57b8e75bd6b94db264302a8f336amrumph Since we've revamped a lot of our processing of path
4f0358189bfa57b8e75bd6b94db264302a8f336amrumph segments, it would be nice to allow this, or at least
4f0358189bfa57b8e75bd6b94db264302a8f336amrumph allow it conditionally with a directive.
5716f9c6daa92dde5f2f9d11ed63f7c9549c223atrawick * FreeBSD, threads, and worker MPM. All seems to work fine
5716f9c6daa92dde5f2f9d11ed63f7c9549c223atrawick if you only have one worker process with many threads. Add
5716f9c6daa92dde5f2f9d11ed63f7c9549c223atrawick a second worker process and the accept lock seems to be
54d750a84a175d8e338880514d440773eb986b50covener lost. This might be an APR issue with how it deals with
54d750a84a175d8e338880514d440773eb986b50covener the child_init hook (i.e. the fcntl lock needs to be resynced).
54d750a84a175d8e338880514d440773eb986b50covener More examination and analysis is required.
54d750a84a175d8e338880514d440773eb986b50covener Status: This has also been reported on Cygwin.
54d750a84a175d8e338880514d440773eb986b50covener Message-ID: <3C2CC514.8EF3BED1@wapme-systems.de> (cygnus)
54d750a84a175d8e338880514d440773eb986b50covener Justin says: So, FreeBSD-CURRENT and Cywin have the same
54d750a84a175d8e338880514d440773eb986b50covener problem. Yum. If another platform has this
7a3aa12f0eda24793ee26d6a179bd53132e9dae8covener with worker, this becomes a showstopper.
54d750a84a175d8e338880514d440773eb986b50covener Aaron says: I spent some time disecting this and have come to
54d750a84a175d8e338880514d440773eb986b50covener the conclusion that it is not a problem in the worker MPM
83b50288fa7d306324bba68832011ea08f5c7832covener (or at least, it is not isolated to a problem in worker).
4e30ef014533a7e93c92d88306291f5e49c9692ftrawick I'll list some of the problems I'm seeing in case someone
83b50288fa7d306324bba68832011ea08f5c7832covener else wants to pick up where I've left off:
5f066f496cd9f20a2a701255bc67d44e7cb46daetrawick - Delivery of just about any signal to one of the child
5f066f496cd9f20a2a701255bc67d44e7cb46daetrawick processes will send it into an infinite loop as well.
5f066f496cd9f20a2a701255bc67d44e7cb46daetrawick - Even though the parent is spinning out of control,
2e15620d724fb8e3a5be183b917359a2fd6e9468covener at first the child or children will appear to work
2e15620d724fb8e3a5be183b917359a2fd6e9468covener properly. At times it is possible to get it into a state,
2e15620d724fb8e3a5be183b917359a2fd6e9468covener however, where a request will hang until another concurrent
2e15620d724fb8e3a5be183b917359a2fd6e9468covener request "kicks" the first, at which point the second will
1b988c41ee505962781d110a3e4c2c90f1ea0aa4covener hang. My theory is that this has to do with the
1b988c41ee505962781d110a3e4c2c90f1ea0aa4covener pthread_cond_*() implementation in FreeBSD, but it's still
1b988c41ee505962781d110a3e4c2c90f1ea0aa4covener possible that it is in APR.
b8efdc95bec9cf089aa1be0bfd07d46aa1137a7acovener Justin adds: Oh, FreeBSD threads are implemented entirely with
b8efdc95bec9cf089aa1be0bfd07d46aa1137a7acovener select()/poll()/longjmp(). Welcome to the nightmare.
b8efdc95bec9cf089aa1be0bfd07d46aa1137a7acovener So, that means a ktrace output also has the thread
f06e7c4b1bce6b6491e5de0b7998d3f5696b293dchrisd scheduling internals in it (since it is all the same to
f06e7c4b1bce6b6491e5de0b7998d3f5696b293dchrisd the kernel). Which makes it hard to distinguish between
f06e7c4b1bce6b6491e5de0b7998d3f5696b293dchrisd our select() calls and their select() calls.
179565be4043d7e5f9161aa75271fa0a001866d9covener *bangs head on wall repeatedly* But, some of the libc_r
179565be4043d7e5f9161aa75271fa0a001866d9covener files have a DBG_MSG #define. This is moderately helpful
179565be4043d7e5f9161aa75271fa0a001866d9covener when used with -DNO_DETACH. The kernel scheduler isn't
111436a32ba1254291e4883292fb116d15fe8f64covener waking up the threads on a select(). Yum. And, I bet
fce4949fb0b309a5744afcd503c6ed2d35621ee2covener those decrementing select calls have to do with the
fce4949fb0b309a5744afcd503c6ed2d35621ee2covener scheduler. Time to brush up on our OS fundamentals.
fce4949fb0b309a5744afcd503c6ed2d35621ee2covener * There is increasing demand from module writers for an API
7b7430e701e9a31ce809da7c220bb8dfcf68c86etrawick that will allow them to control the server � la apachectl.
7b7430e701e9a31ce809da7c220bb8dfcf68c86etrawick Reasons include sole-function servers that need to die if
7b7430e701e9a31ce809da7c220bb8dfcf68c86etrawick an external dependency (e.g., a database) fails, et cetera.
ccc20788c1e5fc973f36df634399c89acb70deaejerenkrantz Perhaps something in the (ever more abused) scoreboard?
ccc20788c1e5fc973f36df634399c89acb70deaejerenkrantz rbb: I don't believe the scoreboard is the correct mechanism
ccc20788c1e5fc973f36df634399c89acb70deaejerenkrantz for this. We already have a pipe that goes between parent
273e512f20f262e5e2aa8e0e83371d1929fb76adjkaluza and child for graceful shutdown events, along with an API that
273e512f20f262e5e2aa8e0e83371d1929fb76adjkaluza can be used to send a message down that pipe. In threaded MPMs,
273e512f20f262e5e2aa8e0e83371d1929fb76adjkaluza it is easy enough to make that one pipe be used for graceful
efe780dcf13b2b95effabf897d694d8f23feac74trawick and graceless events, and it is also easy to open that pipe
fe83f60b41477b14a37edcfcd1f7f5c5a1ebfe44minfrin to both parent and child for writing. Then we just need to
fe83f60b41477b14a37edcfcd1f7f5c5a1ebfe44minfrin figure out how to do graceless on non-threaded MPMs.
993d1261a278d7322bccef219101220b7b4fb8c5jkaluza * Allow the DocumentRoot directive within <Location > scopes? This
993d1261a278d7322bccef219101220b7b4fb8c5jkaluza allows the beloved (crusty) Alias /foo/ /somepath/foo/ followed
993d1261a278d7322bccef219101220b7b4fb8c5jkaluza by a <Directory /somepath/foo> to become simply
ba050a6f942b9fa0e81ed73437588005c569655ccovener <Location /foo/> DocumentRoot /somefile/foo (IMHO a bit more legible
ba050a6f942b9fa0e81ed73437588005c569655ccovener and in-your-face.) DocumentRoot unset would be accepted [and would
ba050a6f942b9fa0e81ed73437588005c569655ccovener not permit content to be served, only virtual resources such as
ba050a6f942b9fa0e81ed73437588005c569655ccovener server-info or server-status.
135ddda3a989215d2bedbcf1529bfb269c3eda23niq This proposed change would _not_ depricate Alias.
135ddda3a989215d2bedbcf1529bfb269c3eda23niq * Win32: Rotatelogs sometimes is not terminated when Apache
001a44c352f89c9ec332ffd3e0a6927dcd19432chumbedooh goes down hard. FirstBill was looking at possibly tracking the
001a44c352f89c9ec332ffd3e0a6927dcd19432chumbedooh child's-child processes in the parent process.
001a44c352f89c9ec332ffd3e0a6927dcd19432chumbedooh OtherBill asks, wasn't this fixed?
efe780dcf13b2b95effabf897d694d8f23feac74trawick stoddard: Not fixed. Shared scoreboard might offer a good
793214f67dede32edfd9ee96c664ead04d175cbbjfclere way for the parent to keep track of 'other child' processes
cc5a4a08dc9783fcbc52ce86f11e01c281a43810minfrin and whack them if the child goes down.
33124689065ade0dfc8c54d8ebb734f9439cb89btrawick * Win32: Add a simple hold console open patch (wait for close or
33124689065ade0dfc8c54d8ebb734f9439cb89btrawick the ESC key, with a nice message) if the server died a bad
9b0076ddd1103e5fa9c1f9bafde4b06ce244fbaecovener death (non-zero exit code) in console mode.
9b0076ddd1103e5fa9c1f9bafde4b06ce244fbaecovener Resolution: bring forward same ugly hacks from 1.3.13-.20
249d09d51808cb7981af99762c3b3736ca126cd5jkaluza * Port of mod_ssl to Apache 2.0:
249d09d51808cb7981af99762c3b3736ca126cd5jkaluza The current porting state is summarized in modules/ssl/README. The
249d09d51808cb7981af99762c3b3736ca126cd5jkaluza remaining work includes:
56589be3d7a3e9343370df240010c6928cc78b39jkaluza (1) stablizing/optimizing the SSL filter logic
56589be3d7a3e9343370df240010c6928cc78b39jkaluza (2) Enabling the various SSL caching mechanisms (shmcb, shmht)
56589be3d7a3e9343370df240010c6928cc78b39jkaluza (3) Enabling SSL extentions
77ca16c5676da23155311e13cee61e7eaba9fa3ejailletc (4) Trying to seperate the https filter logic from mod_ssl -
77ca16c5676da23155311e13cee61e7eaba9fa3ejailletc This is to facilitate other modules that wish to use the https
77ca16c5676da23155311e13cee61e7eaba9fa3ejailletc filter or the mod_ssl logic or both as required.
77ca16c5676da23155311e13cee61e7eaba9fa3ejailletc Justin: mod_ssl filter logic is redone, so that should be fine.
f87299dab99bc04b51a6b8cad51b6795db862c0atrawick Madhu has submitted a patch for SSL caching - however, I
f87299dab99bc04b51a6b8cad51b6795db862c0atrawick am -0 on that patch as I *think* we could implement the
f87299dab99bc04b51a6b8cad51b6795db862c0atrawick shared memory another way that is much cleaner (i.e.
4d12805e6c18253040223ea637acd6b3b3c18f60jorton treat shmem directly as a dbm via APR routines). Justin
4d12805e6c18253040223ea637acd6b3b3c18f60jorton also thinks that the https filter logic may be sufficiently
4d12805e6c18253040223ea637acd6b3b3c18f60jorton decoupled now, but isn't really sure.
4d12805e6c18253040223ea637acd6b3b3c18f60jorton * Performance & Debug: Eliminate most (and perhaps all) of the
e5d909f2b06bd880fb3675cd49363df981caa631trawick malloc/free calls in the bucket brigade code. Need some
a4df2cd1e1391575a327c2a90ba4315f805a0a78covener light weight memory management functions that allow freeing
a4df2cd1e1391575a327c2a90ba4315f805a0a78covener memory (putting it back into a memory pool) when it is no
a4df2cd1e1391575a327c2a90ba4315f805a0a78covener longer needed. Enabling simple debugging features like guard
cb666b29f81df1d11d65002250153353568021fccovener bands, double free detection, etc. would be cool but certainly
cb666b29f81df1d11d65002250153353568021fccovener not a hard requirement.
6a80c3c6f4b8ea7ba5e89402b8b779b09ce020e0covener Status: Cliff started to implement this using SMS as has
1c2cab00d988fc48cbe59032cf76cc0bab20d6f7covener been discussed at length for months, but since
6a80c3c6f4b8ea7ba5e89402b8b779b09ce020e0covener SMS is not being used anywhere else in the server,
75a230a728338d84dcfe81edd375352f34de22d0covener several people expressed the opinion that we should
75a230a728338d84dcfe81edd375352f34de22d0covener get rid of it entirely, meaning that the buckets
75a230a728338d84dcfe81edd375352f34de22d0covener need their own memory management (free list) functions.
1f50dc34ae069adeed20b2986e5ffdefa5c410e0covener Cliff will implement that this weekend so we at least
1f50dc34ae069adeed20b2986e5ffdefa5c410e0covener have something to look at/compare with.
63a5ea80bddcc84a462e40f402b4f330e0e05411covener * Eliminate unnecessary creation of pipes in mod_cgid
63a5ea80bddcc84a462e40f402b4f330e0e05411covener Status: Ken asks, didn't Jeff fix this when he fixed the queue
65a4e663b82f8bce28ac22ab2edfd7502de36998sf * Combine log_child and piped_log_spawn. Clean up http_log.c.
65a4e663b82f8bce28ac22ab2edfd7502de36998sf Common logging API.
65a4e663b82f8bce28ac22ab2edfd7502de36998sf * Document mod_file_cache.
74e7f6c55fd67b10cb400b3f6d1dc718a303d944minfrin * Platforms that do not support fork (primarily Win32 and AS/400)
74e7f6c55fd67b10cb400b3f6d1dc718a303d944minfrin Architect start-up code that avoids initializing all the modules
74e7f6c55fd67b10cb400b3f6d1dc718a303d944minfrin in the parent process on platforms that do not support fork.
a511a29faf2ff7ead3b67680154a624effb31aafminfrin * Win32: Migrate the MPM over to use APR thread/process calls. This
a511a29faf2ff7ead3b67680154a624effb31aafminfrin would eliminate some code in the Win32 branch that essentially
a511a29faf2ff7ead3b67680154a624effb31aafminfrin duplicates what is in APR.
a511a29faf2ff7ead3b67680154a624effb31aafminfrin * There are still a number of places in the code where we are
63921358ef93fcb41bc71d9894221ba3d7fbb87bminfrin losing error status (i.e. throwing away the error returned by a
63921358ef93fcb41bc71d9894221ba3d7fbb87bminfrin system call and replacing it with a generic error code)
deec48c67d4786bc77112ffbf3a4e70b931097edminfrin * Mass vhosting version of suEXEC.
6d601599d3d65df0410eae6e573e75b2dbfb1fb4minfrin * All DBMs suffer from confusion in support/dbmmanage (perl script) since
6d601599d3d65df0410eae6e573e75b2dbfb1fb4minfrin the dbmmanage employs the first-matched dbm format. This is not
6d601599d3d65df0410eae6e573e75b2dbfb1fb4minfrin necessarily the library that Apache was built with. Aught to
684e0cfc200f66287a93bbd1708d1dd8a92a7eefcovener rewrite dbmmanage upon installation to bin/ with the proper library
684e0cfc200f66287a93bbd1708d1dd8a92a7eefcovener for predictable mod_auth_dbm administration.
5c43d2fb853f84497b5ece2d414ef9484aa87e5fsf Questions; htdbm exists, time to kill dbmmanage, or does it remain
05a5a9c3e16f21566e1b61f4bd68025ce1b741ccjoes useful as a perl dbm management example? If we keep it,
05a5a9c3e16f21566e1b61f4bd68025ce1b741ccjoes do we address the issue above?
26c5829347f6a355c00f1ba0301d575056b69536niq * Integrate mod_dav.
ef82e8fa164e0a1f8b813f7deb6b7ead96018c94niq Some additional items remaining:
ef82e8fa164e0a1f8b813f7deb6b7ead96018c94niq - case_preserved_filename stuff
ef82e8fa164e0a1f8b813f7deb6b7ead96018c94niq (use the new canonical name stuff?)
ef82e8fa164e0a1f8b813f7deb6b7ead96018c94niq - find a new home for ap_text(_header)
ef82e8fa164e0a1f8b813f7deb6b7ead96018c94niq - is it possible to remove the DAV: namespace stuff from util_xml?
413ee814748f37be168ff12407fa6dba0ceeabe6trawick * ap_core_translate() and its use by mod_mmap_static and mod_file_cache
c12917da693bae4028a1d5a5e8224bceed8c739dsf are a bit wonky. The function should probably be exposed as a utility
c12917da693bae4028a1d5a5e8224bceed8c739dsf function (such as ap_translate_url2fs() or ap_validate_fs_url() or
eafcc0ebf263d0ba69855b6e10958c4c1a2361bdsf something). Another approach would be a new hook phase after
eafcc0ebf263d0ba69855b6e10958c4c1a2361bdsf "translate" which would allow the module to munge what the
eafcc0ebf263d0ba69855b6e10958c4c1a2361bdsf translation has decided to do.
eafcc0ebf263d0ba69855b6e10958c4c1a2361bdsf Status: Greg +1 (volunteers), Ryan +1
d7ffd2da16d58b1a0de212e4d56f7aebb72bef26sf * Explore use of a post-config hook for the code in http_main.c which
d7ffd2da16d58b1a0de212e4d56f7aebb72bef26sf calls ap_fixup_virutal_hosts(), ap_fini_vhost_config(), and
d7ffd2da16d58b1a0de212e4d56f7aebb72bef26sf ap_sort_hooks() [to reduce the logic in main()]
4576c1a9ef54cd1e5555ee07d016a7f559f80338sf * read the config tree just once, and process N times (as necessary)
9811aed12bbc71783d2e544ccb5fecd193843eadsf * (possibly) use UUIDs in mod_unique_id and/or mod_usertrack
9811aed12bbc71783d2e544ccb5fecd193843eadsf * (possibly) port the bug fix for PR 6942 (segv when LoadModule is put
88fac54d9d64f85bbdab5d7010816f4377f95bd7rjung into a VirtualHost container) to 2.0.
bd3f5647b96d378d9c75c954e3f13582af32c643sf * shift stuff to mod_core.h
bd3f5647b96d378d9c75c954e3f13582af32c643sf * callers of ap_run_create_request() should check the return value
bd3f5647b96d378d9c75c954e3f13582af32c643sf for failure (Doug volunteers)
2a7beea91d46beb41f043a84eaad060047ee04aafabien * Win32: Get Apache working on Windows 95/98. The following work
2a7beea91d46beb41f043a84eaad060047ee04aafabien (at least) needs to be done:
2a7beea91d46beb41f043a84eaad060047ee04aafabien - Document warning that OSR2 is required (for Crypt functions, in
2a7beea91d46beb41f043a84eaad060047ee04aafabien rand.c, at least.) This could be resolved with an SSL library, or
584a85dd4047e38d3ed3a29b6662fcc9d100ae4csf randomization in APR itself.
584a85dd4047e38d3ed3a29b6662fcc9d100ae4csf - Bring the Win9xConHook.dll from 1.3 into 2.0 (no sense till it
584a85dd4047e38d3ed3a29b6662fcc9d100ae4csf actually works) and add in a splash of Win9x service code.
f21e9e3d0bfb7a507ecc5bc963f2159d693503d1sf * In order to use a DSO version of mod_ssl we have to link with
f21e9e3d0bfb7a507ecc5bc963f2159d693503d1sf -lssl and -lcrypto. A workaround is in place right now where the
f6b9c755a0b793e8a3a3aebd327ca20a86478117sf entire EXTRA_LIBS macro is being appended to the objects list, but
f6b9c755a0b793e8a3a3aebd327ca20a86478117sf this is a hack. We should either revamp the APACHE_CHECK_SSL_TOOLKIT
f6b9c755a0b793e8a3a3aebd327ca20a86478117sf autoconf function or come up with some other autoconf checks to
132ee6ac1c26d6e8953836316ba50734eefab47bsf search for libssl and libcrypto and properly add them to mod_ssl's
132ee6ac1c26d6e8953836316ba50734eefab47bsf link flags.
85eacfc96a04547ef25aabbc06440039715084c2jorton * Fix the worker MPM to use POD to kill child processes instead
85eacfc96a04547ef25aabbc06440039715084c2jorton of ap_os_killpg, regardless of how they should die. (Ryan Bloom)
536d2e7cd1fdec1255b8c3bdf41fdc714c506a54trawick * mod_cache: handle cache_control: no_cache "field_name" to enable
536d2e7cd1fdec1255b8c3bdf41fdc714c506a54trawick cacheing the response w/o header "field_name"
536d2e7cd1fdec1255b8c3bdf41fdc714c506a54trawick See RFC2616 section 14.9.1
79c5787b92ac5f0e1cc82393816c77a006399316trawickPRs that have been suspended forever waiting for someone to
79c5787b92ac5f0e1cc82393816c77a006399316trawickput them into 'the next release':
79c5787b92ac5f0e1cc82393816c77a006399316trawick * PR#76: general
c967bf3bc89e8aa60dbd30d9da388e448ddc1cc4trawick missing call to "setlocale();"
79c5787b92ac5f0e1cc82393816c77a006399316trawick * PR#78: mod_include
79c5787b92ac5f0e1cc82393816c77a006399316trawick Additional status for XBitHack directive
7b395e4e878c28a4784919cfd2e704ddd14a3390jorton * PR#362: mod_proxy
7b395e4e878c28a4784919cfd2e704ddd14a3390jorton Mod_proxy doesn't allow change of error pages
536e48c08d674acac5d44929318f2ad928edc361jorton * PR#370: mod_env
e81785da447b469da66f218b3f0244aab507958djorton Modified PATH environemnt variable is not passed, instead
e81785da447b469da66f218b3f0244aab507958djorton system's is used
3e4e54d4e3fc0123c63d57aa84ac7ad7a8c73ff8jorton * PR#440: mod_proxy
53e9b27aba029b18be814df40bcf6f0428771d1efuankg Proxy doesn't deliver documents if not connected
53e9b27aba029b18be814df40bcf6f0428771d1efuankg * PR#534: mod_proxy
53e9b27aba029b18be814df40bcf6f0428771d1efuankg proxy converts ~name to %7Ename when name starts with a dot (.)
ca61ccd0c306c2c72df153688ba1b49f3eceed80sf * PR#537: mod_access
6bb524f1895f30265a1431afc460977d391cb36bsf mod_access syntax allows hosts that should be restricted
e6dd71992459d05a676b98b7963423dc5dc1e24aminfrin * PR#557: mod_auth-any
e6dd71992459d05a676b98b7963423dc5dc1e24aminfrin ~UserHome directories are not honored in absolute pathname
23f1535d6a60817d2846bac0aea230ea475d7dccminfrin requests (.htaccess)
23f1535d6a60817d2846bac0aea230ea475d7dccminfrin * PR#612: mod_proxy
ec7520b24cd80d34d82bbcaca153cbb23cc04bc0rjung Proxy FTP Authentication Fails
ec7520b24cd80d34d82bbcaca153cbb23cc04bc0rjung * PR#623: mod_include
ec7520b24cd80d34d82bbcaca153cbb23cc04bc0rjung A smarter "Last Modified" value for SSI documents (see PR number 600)
ec7520b24cd80d34d82bbcaca153cbb23cc04bc0rjung * PR#628: config
6249dfa569d3b4f1f539665b979a80c6e335d93etrawick Request of "Options SymLinksIfGroupMatch"
6249dfa569d3b4f1f539665b979a80c6e335d93etrawick * PR#700: mod_proxy
ae600ca541efc686b34f8b1f21bd3d0741d37674covener Proxy doesn't do links right for OpenVMS files through ftp:
74499a117b3b2cd9666715a14f90c0e5d1a4ee8ajim * PR#759: mod_imap
cfa64348224b66dd1c9979b809406c4d15b1c137fielding imap should read <MAP><AREA>*</MAP> too!
74499a117b3b2cd9666715a14f90c0e5d1a4ee8ajim * PR#793: general
cfa64348224b66dd1c9979b809406c4d15b1c137fielding RLimitCPU and RLimitMEM don't apply to all children like they should
Using NIS passwd.byname dbm files with AuthDBMUserFile
setlogin() is not called, causing problems with e.g. identd
add allow,deny/deny,allow warning to mod_access
A small addition to rotatelogs.c to improve program functionality.
Feedback/Comment on APACI
Inclusion of RPM spec file in CVS/distributions
* orig_ct in the byterange/multipart handling may not be