1N/AAPACHE 2.0 STATUS: -*-text-*-
1N/ALast modified at [$Date: 2001/12/27 20:20:21 $]
1N/A 2.0.30 : In development
1N/A 2.0.29 : tagged November 27, 2001
1N/A 2.0.28 : released November 13, 2001
1N/A 2.0.27 : rolled November 6, 2001
1N/A 2.0.26 : tagged October 16, 2001. not rolled.
1N/A 2.0.25 : rolled August 29, 2001
1N/A 2.0.24 : rolled August 18, 2001
1N/A 2.0.23 : rolled August 9, 2001
1N/A 2.0.22 : rolled July 29, 2001
1N/A 2.0.21 : rolled July 20, 2001
1N/A 2.0.20 : rolled July 8, 2001
1N/A 2.0.19 : rolled June 27, 2001
1N/A 2.0.18 : rolled May 18, 2001
1N/A 2.0.17 : rolled April 17, 2001
1N/A 2.0.16 : rolled April 4, 2001
1N/A 2.0.15 : rolled March 21, 2001
1N/A 2.0.14 : rolled March 7, 2001
1N/A 2.0a9 : released December 12, 2000
1N/A 2.0a8 : released November 20, 2000
1N/A 2.0a7 : released October 8, 2000
1N/A 2.0a6 : released August 18, 2000
1N/A 2.0a5 : released August 4, 2000
1N/A 2.0a4 : released June 7, 2000
1N/A 2.0a3 : released April 28, 2000
1N/A 2.0a2 : released March 31, 2000
1N/A 2.0a1 : released March 10, 2000
1N/APlease consult the following STATUS files for information
1N/ARELEASE SHOWSTOPPERS:
1N/A * ap_directory_walk skips some per-dir config merge functions
1N/A if there is no "<Directory />" block in the configuration
1N/A Message-ID: <m3itbdiijq.fsf@rdu163-40-092.nc.rr.com>
1N/A * Test suite failures:
1N/A o perchild doesn't even build
1N/A o all MPMs fail the LimitRequestBody directive test
1N/A o both worker and prefork are failing some of the 'chunked' subtests
1N/A o worker is also failing some of the 'cgi' subtests
1N/A * If any request gets to the core handler, without a flag that this
1N/A end of the ap_process_request_internal() processing. This provides
1N/A authors of older modules better compatibility, while still improving
1N/A the security and robustness of 2.0.
1N/A Status: still need to decide where this goes, OtherBill comments...
1N/A Message-ID: <065701c14526$495203b0$96c0b0d0@roweclan.net>
1N/A we need to look at halting this in the 'default handler' case,
1N/A and that implies pushing the 'handler election' into the request
1N/A internal processing phase from the run request phase.
1N/A * There is a bug in how we sort some hooks, at least the pre-config
1N/A hook. The first time we call the hooks, they are in the correct
1N/A order, but the second time, we don't sort them correctly. Currently,
1N/A back when this is fixed. rbb
1N/A administrator to order filters, beyond the order of filename (mime)
1N/A filename extensions. At minimum, some sort of +-[0-10] syntax seems
1N/A like the quickest fix for a 2.0 gold release.
1N/A * mod_negotiation needs a new option or directive, something like
1N/A ForceLanguagePriority, to fall back to the LanguagePriority
1N/A directive instead of returning a "no acceptable variant" error.
1N/A Status: Bill has some code in his tree that accomplishes
1N/A this, and will commit it Friday after it's tested.
1N/A * Fold mod_auth_db features back into mod_auth_dbm, and depricate it.
1N/A This can't wait until we have a 2.0-gold release, if folks need
1N/A to move over to auth_dbm, we can't do that to them after 2.0 gold.
1N/A Status: Ian says.. auth_dbm can now handle multiple DBM types,
1N/A is this still an issue?
1N/A Vote: Remove mod_auth_db
1N/A * Convert all instances of the old apr_lock_t type to the new
1N/A types (once they are fully supported in APR).
1N/A Status: Aaron is working on converting INTRAPROCESS
1N/A to apr_thread_mutex_t types. Full replacements for
1N/A LOCKALL and CROSS_PROCESS are not yet complete on all
1N/A platforms, and should only be used in MPMs like worker
1N/A with limited OS exposure.
1N/A * ap_create_scoreboard() can exit the process, leaving stuff like
1N/A mod_cgid's daemon process stranded. Either ap_create_scoreboard()
1N/A needs to be called at a different time or the pre-mpm hook needs
1N/A to be able to return an error code.
1N/ARELEASE NON-SHOWSTOPPERS BUT WOULD BE REAL NICE TO WRAP THESE UP:
1N/A * Handling of %2f in URIs. Currently both 1.3 and 2.0
1N/A completely disallow %2f in the request URI path (see
1N/A ap_unescape_url() in
util.c). It's permitted and passed
1N/A through in the query string, however. Roy says the
1N/A original reason for disallowing it, from five years ago,
1N/A was to protect CGI scripts that applied PATH_INFO to
1N/A a filesystem location and which might be tricked by
1N/A ..%2f..%2f(...). We *should* allow path-info of the
1N/A Since we've revamped a lot of our processing of path
1N/A segments, it would be nice to allow this, or at least
1N/A allow it conditionally with a directive.
1N/A * FreeBSD, threads, and worker MPM. All seems to work fine
1N/A if you only have one worker process with many threads. Add
1N/A a second worker process and the accept lock seems to be
1N/A lost. This might be an APR issue with how it deals with
1N/A the child_init hook (
i.e. the fcntl lock needs to be resynced).
1N/A More examination and analysis is required.
1N/A * There is increasing demand from module writers for an API
1N/A that will allow them to control the server � la apachectl.
1N/A Reasons include sole-function servers that need to die if
1N/A an external dependency (
e.g., a database) fails, et cetera.
1N/A Perhaps something in the (ever more abused) scoreboard?
1N/A rbb: I don't believe the scoreboard is the correct mechanism
1N/A for this. We already have a pipe that goes between parent
1N/A and child for graceful shutdown events, along with an API that
1N/A can be used to send a message down that pipe. In threaded MPMs,
1N/A it is easy enough to make that one pipe be used for graceful
1N/A and graceless events, and it is also easy to open that pipe
1N/A to both parent and child for writing. Then we just need to figure
1N/A out how to do graceless on non-threaded MPMs.
1N/A * revamp the input filter behavior, per discussions since
1N/A February (and especially at the hackathon last
1N/A April). Specifically, ap_get_brigade will return a brigade with
1N/A *up to* a specific number of bytes, or a "line" of data. The
1N/A read may be blocking or nonblocking. ap_getline() will be
1N/A refactored into apr_brigade_getline(), and then DECHUNK can use
1N/A f->next (ap_getline will always read "top of input stack"). Also
1N/A fix the bug where request body content will end up closing the
1N/A connection (buggering up persistent conns).
1N/A Status: Justin is working on this as fast as he can.
1N/A The core input filters, HTTP-related filters, mod_ssl, and
1N/A mod_proxy are switched to the new logic.
1N/A However, ap_getline() still needs to be refactored out. But,
1N/A there's a problem there: ap_getline() peeks ahead for MIME
1N/A continuation (first character on line is space or \t) and
1N/A stores unused data in core_request_config which violates the
1N/A abstraction. That's cheating. So, we may not be able to
1N/A implement this without setting some data aside (yuck!).
1N/A I believe this is OtherBill's main complaint with the current
1N/A AIUI (correct me if I'm wrong!), OtherBill believes we
1N/A should have a pushback option so that we can return unread
1N/A data - this would solve this case. However, my question to
1N/A him is how do we handle stuff like mod_ssl - we can't "unread"
1N/A data. So, do we have two brigades for each filter? An in
1N/A brigade and a returned brigade? That seems messy. To
1N/A everyone else, can we refactor ap_getline() without pushback
1N/A - socket bucket and core input filter changes. see end of
1N/A message ID (Feb 27): <20010227075326.S2297@lyra.org>
1N/A - fix up ap_get_brigade() semantics, fix bug in DECHUNK /
1N/A ap_getline. many messages (plus their threads) (
Apr/May):
1N/A Message-ID: <20010402101207.J27539@lyra.org>
1N/A Message-ID: <3AF7F921.D2EEC41A@algroup.co.uk>
1N/A Message-ID: <20010508190029.E18404@lyra.org>
1N/A Message-ID: <20010509115445.D1374@lyra.org>
1N/A - thoughts on filter modes:
1N/A Message-ID: <021b01c14dee$09782af0$93c0b0d0@roweclan.net>
1N/A * Allow the DocumentRoot directive within <Location > scopes? This
1N/A and in-your-face.) DocumentRoot unset would be accepted [and would
1N/A not permit content to be served, only virtual resources such as
1N/A server-info or server-status.
1N/A This proposed change would _not_ depricate Alias.
1N/A * Win32: Rotatelogs sometimes is not terminated when Apache
1N/A goes down hard. FirstBill was looking at possibly tracking the
1N/A child's-child processes in the parent process.
1N/A OtherBill asks, wasn't this fixed?
1N/A * Win32: Add a simple hold console open patch (wait for close or
1N/A the ESC key, with a nice message) if the server died a bad
1N/A death (non-zero exit code) in console mode.
1N/A Resolution: bring forward same ugly hacks from 1.3.13-.20
1N/A * Port of mod_ssl to Apache 2.0:
1N/A remaining work includes:
1N/A (2) Enabling the various SSL caching mechanisms (shmcb, shmht)
1N/A (3) Enabling SSL extentions
1N/A (4) Trying to seperate the https filter logic from mod_ssl -
1N/A This is to facilitate other modules that wish to use the https
1N/A filter or the mod_ssl logic or both as required.
1N/A Justin: mod_ssl filter logic is redone, so that should be fine.
1N/A Madhu has submitted a patch for SSL caching - however, I
1N/A am -0 on that patch as I *think* we could implement the
1N/A shared memory another way that is much cleaner (
i.e. 1N/A treat shmem directly as a dbm via APR routines). Justin
1N/A also thinks that the https filter logic may be sufficiently
1N/A decoupled now, but isn't really sure.
1N/A * Performance: Get the SINGLE_LISTEN_UNSERIALIZED_ACCEPT
1N/A optimization working in worker. prefork's new design for how
1N/A to notice data on the pod should be sufficient.
1N/A * Performance & Debug: Eliminate most (and perhaps all) of the
1N/A light weight memory management functions that allow freeing
1N/A memory (putting it back into a memory pool) when it is no
1N/A longer needed. Enabling simple debugging features like guard
1N/A bands, double free detection, etc. would be cool but certainly
1N/A not a hard requirement.
1N/A Status: Cliff started to implement this using SMS as has
1N/A been discussed at length for months, but since
1N/A SMS is not being used anywhere else in the server,
1N/A several people expressed the opinion that we should
1N/A get rid of it entirely, meaning that the buckets
1N/A need their own memory management (free list) functions.
1N/A Cliff will implement that this weekend so we at least
1N/A * Eliminate unnecessary creation of pipes in mod_cgid
1N/A * the autoconf setup should be fixed to default to using the
1N/A in a layout should be overridable on the command line. Plus,
1N/A what we do right now just doesn't seem to fully fit into how autoconf
1N/A works, eg. AC_PREFIX_DEFAULT issues.
1N/A Message-ID: <Pine.BSF.4.20.0104031557420.20876-100000@alive.znep.com>
1N/A * Document mod_file_cache.
1N/A * OS/2: Make mod_status work for spmt_os2 MPM.
1N/A * Platforms that do not support fork (primarily Win32 and AS/400)
1N/A Architect start-up code that avoids initializing all the modules
1N/A in the parent process on platforms that do not support fork.
1N/A would eliminate some code in the Win32 branch that essentially
1N/A duplicates what is in APR.
1N/A * There are still a number of places in the code where we are
1N/A losing error status (
i.e. throwing away the error returned by a
1N/A system call and replacing it with a generic error code)
1N/A * Mass vhosting version of suEXEC.
1N/A the dbmmanage employs the first-matched dbm format. This is not
1N/A necessarily the library that Apache was built with. Aught to
1N/A rewrite dbmmanage upon installation to bin/ with the proper library
1N/A Status: Mladen Turk has posted several patches and ideas.
1N/A Key question, part of htpasswd, or a seperate utility?
1N/A prefer seperate: OtherBill
1N/A * use apu_dbm in mod_auth_dbm
1N/A Status: Greg +1 (low-priority volunteer)
1N/A Justin says: "Seems like this is already there, so should we just
1N/A remove the other DBM code in that file? If you want
1N/A to use gdbm, or dbm, etc, you should tell apr-util."
1N/A Will says: "bs - I may choose the fastest - most efficient native
1N/A dbm implementation, for shared proc caches, ssl session
1N/A caching, etc, but that has nothing to do with maintaining
1N/A a userlist via dbm, which has to remain readable between
1N/A for apr-util would let us do this with just apr, though."
1N/A Ian says: "multi-dbm is in, but it still has ndbm support hardcoded
1N/A is this still required? isn't ndbm supported via gdbm?"
1N/A * Integrate mod_dav.
1N/A Some additional items remaining:
1N/A - case_preserved_filename stuff
1N/A (use the new canonical name stuff?)
1N/A - find a new home for ap_text(_header)
1N/A - is it possible to remove the DAV: namespace stuff from util_xml?
1N/A * ap_core_translate() and its use by mod_mmap_static and mod_file_cache
1N/A are a bit wonky. The function should probably be exposed as a utility
1N/A function (such as ap_translate_url2fs() or ap_validate_fs_url() or
1N/A something). Another approach would be a new hook phase after
1N/A "translate" which would allow the module to munge what the
1N/A translation has decided to do.
1N/A Status: Greg +1 (volunteers), Ryan +1
1N/A calls ap_fixup_virutal_hosts(), ap_fini_vhost_config(), and
1N/A ap_sort_hooks() [to reduce the logic in main()]
1N/A * read the config tree just once, and process N times (as necessary)
1N/A * (possibly) use UUIDs in mod_unique_id
and/or mod_usertrack
1N/A * (possibly) port the bug fix for PR 6942 (segv when LoadModule is put
1N/A into a VirtualHost container) to 2.0.
1N/A * callers of ap_run_create_request() should check the return value
1N/A for failure (Doug volunteers)
1N/A * Win32: Get Apache working on Windows 95/98. The following work
1N/A (at least) needs to be done:
1N/A - winnt MPM: Fix 95/98 code paths in the winnt MPM. There is some NT
1N/A specific code that is still not in NT only code paths
1N/A - IOL binds to APR sendfile, implemented with TransmitFile, which
1N/A is not available on 95/98.
1N/A - Document warning that OSR2 is required (for Crypt functions, in
1N/A rand.c, at least.) This could be resolved with an SSL library, or
1N/A randomization in APR itself.
1N/A actually works) and add in a splash of Win9x service code.
1N/A * In order to use a DSO version of mod_ssl we have to link with
1N/A -lssl and -lcrypto. A workaround is in place right now where the
1N/A entire EXTRA_LIBS macro is being appended to the objects list, but
1N/A this is a hack. We should either revamp the APACHE_CHECK_SSL_TOOLKIT
1N/A autoconf function or come up with some other autoconf checks to
1N/A search for libssl and libcrypto and properly add them to mod_ssl's
1N/A * Make the worker MPM the default MPM for threaded Unix boxes.
1N/A +1: Justin, Jeff, Ian
1N/A -0: Aaron (premature decision, needs more discussion)
1N/A -0: Cliff (I think the default config should be the safest possible)
1N/A * Fix the worker MPM to use POD to kill child processes instead
1N/A of ap_os_killpg, regardless of how they should die. (Ryan Bloom)
1N/APRs that have been suspended forever waiting for someone to
1N/Aput them into 'the next release':
1N/A missing call to "setlocale();"
1N/A * PR#78: mod_include
1N/A Additional status for XBitHack directive
1N/A Mod_proxy doesn't allow change of error pages
1N/A Modified PATH environemnt variable is not passed, instead
1N/A Proxy doesn't deliver documents if not connected
1N/A proxy converts ~name to %7Ename when name starts with a dot (.)
1N/A * PR#537: mod_access
1N/A mod_access syntax allows hosts that should be restricted
1N/A * PR#557: mod_auth-any
1N/A ~UserHome directories are not honored in absolute pathname
1N/A requests (.htaccess)
1N/A Proxy FTP Authentication Fails
1N/A * PR#623: mod_include
1N/A A smarter "Last Modified" value for SSI documents (see PR number 600)
1N/A Request of "Options SymLinksIfGroupMatch"
1N/A Proxy doesn't do links right for OpenVMS files through ftp:
1N/A imap should read <MAP><AREA>*</MAP> too!
1N/A RLimitCPU and RLimitMEM don't apply to all children like they should
1N/A Uses cwd before filling it in, doesn't use syslog
1N/A it is useful to allow specifiction that root-owned symlinks
1N/A should always be followed
1N/A Controlling Access to Remote Proxies would be nice...
1N/A Adding authentication "on the fly" through the proxy module
1N/A * PR#1004: apache-api
1N/A request_config field in request_rec is moderately bogus
1N/A DoS attacks involving memory consumption
1N/A * PR#1050: mod_log-any
1N/A Logging of virtual server to error_log as well
1N/A * PR#1085: mod_proxy
1N/A ProxyRemote make a dead cycle.
1N/A * PR#1117: mod_auth-any
1N/A suexec does not parse arguments to #exec cmd
1N/A * PR#1145: mod_include
1N/A Allow for Last-Modified: without resorting to XBitHack
1N/A * PR#1158: apache-api
1N/A improvements to child spawning API
1N/A * PR#1166: mod_proxy
1N/A ``nph-'' not honored (no buffering) for ProxyRemote mapping
1N/A Apache cannot handle continuation line in headers
1N/A setlogin() is not called, causing problems with
e.g. identd
1N/A regerror() exists, use it
1N/A * PR#1233: apache-api
1N/A there is no way to keep per-connection per-module state
1N/A * PR#1263: mod_autoexec
1N/A Add frame-safe anchor attribute to mod_autoindex links
1N/A CGI scripts running as Apache user: security (suexec etc.)
1N/A * PR#1287: mod_access
1N/A * PR#1290: mod_proxy
1N/A Need to know "hit-rate" on proxy cache
1N/A * PR#1358: mod_log-any
1N/A Selective url-encode of log fields (or maybe a pseudo
1N/A log_rewrite module?)
1N/A * PR#1383: mod_headers
1N/A I make mod_headers to modify request headers as well as
1N/A * PR#1532: mod_proxy
1N/A Proxy transfer logging
1N/A * PR#1547: mod_proxy
1N/A No HTTP_X_FORWARDED_FOR set...
1N/A * PR#1567: mod_proxy
1N/A ProxyRemote proxy requests fail authentication by firewall
1N/A * PR#1582: mod_rewrite
1N/A mod_rewrite forms REQUEST_URI different than mod_cgi does
1N/A * PR#1677: mod_headers
1N/A mod_headers should allow mod_log_config-style formats in
1N/A * PR#1702: mod_proxy
1N/A mod_proxy to support persistent conns?
1N/A * PR#1803: mod_include
1N/A patches to mod_include to allow for file tests
1N/A * PR#1809: mod_auth-any
1N/A Suggestion for improving authentication modules and core source
1N/A code, problem with 401 and ErrorDocument
1N/A * PR#1878: mod_proxy
1N/A listing of proxy cache content
1N/A Allow modules to set user:group for execution.
1N/A * PR#2024: apache-api
1N/A adding auth_why to conn_rec
1N/A * PR#2073: mod_log-any
1N/A pipelined connections are not logged correctly
1N/A * PR#2074: mod_rewrite
1N/A mod_rewrite doesn't pass Proxy Throughput on internal subrequests
1N/A HTTP Server Rebuild Line Needs Changing for the better
1N/A * PR#2138: mod_status
1N/A mod_status always displays 256 possible connection slots
1N/A * PR#2221: documentation
1N/A Make online documentation search link back to my installation
Can not POST to ErrorDocument - Apache/1.3b6
Status module averages are for entire uptime
suexec for general access of user content?
Proposal for TimeZone directive
/server-info doesn't check for the virtual host to list the info
problem specifying ndbm library for build ?with autoconfigure
A small addition to
rotatelogs.c to improve program functionality.
AllowOverride FileInfo is too coarse
TimeOut applies to output of CGI scripts
<IfDenied> directive wanted
CGI's for general use still have to be run as another user
Cache file names in Proxy module
mailto tags and bundling bug report script
Support for System Resource Controller
When will Apache support P3P? Any Plans?
Propose that Apache recommend $UNIQUE_ID for all "session id"
suggestion: power up your Include directive :)
cannot limit some HTTP methods
No module specific data hook for per-connection data
* PR#3191: mod_negotiation
no way to set global quality-of-source (qs) coneg values
Accessing URL through proxy server corrupts data.
Some anonymous FTP URLs ask for authentication
New ErrorDocumentMatch directive
Need to be able to override shebang line to make CGI scripts
"Files" and "FilesMatch" regexp does not recognize bang as
Please allow CGI env variables (QUERY_STRING, ...) to be logged
Suggestion for better handling of Last-modified headers
mod_cgi prevents handling of OPTIONS requests
[PATCH] install as win32 service with domain account
AllowOverride should have a 'CheckNone' and 'AllowNone' argument
Other bugs that need fixing:
* MaxRequestsPerChild measures connections, not requests.
Until someone has a better way, we'll probably just rename it
"MaxConnectionsPerChild".
* Regex containers don't work in an intutive way
Status: No one has come up with an efficient way to fix this
behavior. Dean has suggested getting rid of regex containers
* SIGSEGV on Linux (glibc 2.1.2) isn't caught properly by a
sigwaiting thread. We need to work around this, perhaps unless
there is hope soon for a fixed glibc.
needed. Apache 1.3 just never stashed "multipart" into
r->content_type. We should probably follow suit since the
byterange stuff doesn't want the rest of the code to see the
multipart content-type; the other code should still think it is
dealing with the <orig_ct> stuff.
Status: Greg volunteers to investigate (esp. since he was most
likely the one to break it :-)
Other features that need writing:
* Finish infrastructure in core for async MPMs
* TODO in source -- just do an egrep on "TODO" and see what's there
* Jon Travis's <jtravis@covalent.net> patch to deal with thread-safe
issues with inet_ntoa. See message <20001201163220.A12827@covalent.net>
Status: This is being set aside until the IPv6 work is finished
so that we know exactly what is required.
* Martin Sojka <msojka@gmx.de>'s patch to add error reporting for failed
htpasswd actions due to a full /tmp volume (other programs may have
* Mike Abbott's <mja@trudge.engr.sgi.com> patches to improve
Status: These were written for 1.3, and are awaiting a port to
* Jim Winstead's <jimw@trainedmonkey.com> patch to add CookieDomain and
other small mod_usertrack features
* Dan Rench's <drench@xnet.com> patch to add allow the errmsg and timefmt
of SSI's to be modified in the config file. Patch is available in
* Which MPMs will be included with Apache 2.0?