1 APACHE 2.0 STATUS: -*-text-*-
2 Last modified at [$Date: 2001/10/31 07:18:47 $]
6 2.0.27 : tagged October 30th, 2001.
7 2.0.26 : tagged October 16th, 2001. not rolled.
8 2.0.25 : rolled August 29, 2001
9 2.0.24 : rolled August 18, 2001
10 2.0.23 : rolled August 9, 2001
11 2.0.22 : rolled July 29, 2001
12 2.0.21 : rolled July 20, 2001
13 2.0.20 : rolled July 8, 2001
14 2.0.19 : rolled June 27, 2001
15 2.0.18 : rolled May 18, 2001
16 2.0.17 : rolled April 17, 2001
17 2.0.16 : rolled April 4, 2001
18 2.0.15 : rolled March 21, 2001
19 2.0.14 : rolled March 7, 2001
20 2.0a9 : released December 12, 2000
21 2.0a8 : released November 20, 2000
22 2.0a7 : released October 8, 2000
23 2.0a6 : released August 18, 2000
24 2.0a5 : released August 4, 2000
25 2.0a4 : released June 7, 2000
26 2.0a3 : released April 28, 2000
27 2.0a2 : released March 31, 2000
28 2.0a1 : released March 10, 2000
30 Please consult the following STATUS files for information
34 * srclib/apr-util/STATUS
39 * If any request gets to the core handler, without a flag that this
40 r->filename was tested by dir/file_walk, we need to 500 at the very
41 end of the ap_process_request_internal() processing. This provides
42 authors of older modules better compatibility, while still improving
43 the security and robustness of 2.0.
44 Status: still need to decide where this goes, OtherBill comments...
45 Message-ID: <065701c14526$495203b0$96c0b0d0@roweclan.net>
46 we need to look at halting this in the 'default handler' case,
47 and that implies pushing the 'handler election' into the request
48 internal processing phase from the run request phase.
50 * There is a bug in how we sort some hooks, at least the pre-config
51 hook. The first time we call the hooks, they are in the correct
52 order, but the second time, we don't sort them correctly. Currently,
53 the modules/http/config.m4 file has been renamed to
54 modules/http/config2.m4 to work around this problem, it should moved
55 back when this is fixed. rbb
57 * The Add...Filter and Set...Filter directives do not allow the
58 administrator to order filters, beyond the order of filename (mime)
59 extensions. It isn't clear if Set...Filter(s) should be inserted
60 before or after the Add...Filter(s) which are ordered by sequence of
61 filename extensions. Add...FilterByType will add to this quandry.
62 Some sort of resolution needs to be proposed,
64 * mod_dir should normally redirect ALL directory requests which do
65 not include a trailing slash on the URI. However, if a "notes"
66 flag is set (say, via BrowserMatch), this behavior will be
67 disabled for non-GET requests.
68 Status: Greg volunteers
69 MsgId: <20010227104646.E2297@lyra.org>
70 MsgId: <3A9C0097.9C83F07C@Golux.Com>
72 * mod_negotiation needs a new option or directive, something like
73 ForceLanguagePriority, to fall back to the LanguagePriority
74 directive instead of returning a "no acceptable variant" error.
75 Status: Bill has some code in his tree that accomplishes
76 this, and will commit it Friday after it's tested.
78 * Usability: Sanitize the MPM config directives. MaxClients in
79 the threaded MPM is totally misleading now as it has little to
80 do with limiting the number of clients (it limits the number
81 of child processes). Bill proposed nomenclature change to
82 something like "StartWorkers, MaxWorkers, etc." that could
83 apply to most all the MPMs (with some notable exceptions).
84 Bill would be happy with changing MaxClients to MaxServers
85 to make it agree with the operation of the StartServers
88 * revamp the input filter semantics, per discussions since
89 February (and especially at the hackathon last
90 April). Specifically, ap_get_brigade will return a brigade with
91 *up to* a specific number of bytes, or a "line" of data. The
92 read may be blocking or nonblocking. ap_getline() will be
93 refactored into apr_brigade_getline(), and then DECHUNK can use
94 f->next (ap_getline will always read "top of input stack"). Also
95 fix the bug where request body content will end up closing the
96 connection (buggering up persistent conns).
97 Status: Justin is working on this as fast as he can.
98 The core input filters, HTTP-related filters, mod_ssl, and
99 mod_proxy are switched to the new logic.
101 - socket bucket and core input filter changes. see end of
102 message ID (Feb 27): <20010227075326.S2297@lyra.org>
104 - fix up ap_get_brigade() semantics, fix bug in DECHUNK /
105 ap_getline. many messages (plus their threads) (Apr/May):
106 Message-ID: <20010402101207.J27539@lyra.org>
107 Message-ID: <3AF7F921.D2EEC41A@algroup.co.uk>
108 Message-ID: <20010508190029.E18404@lyra.org>
110 - further work with combining/tweaking the builtin filters:
111 Message-ID: <20010509115445.D1374@lyra.org>
113 - thoughts on filter modes:
114 Message-ID: <021b01c14dee$09782af0$93c0b0d0@roweclan.net>
116 * Fold mod_auth_db features back into mod_auth_dbm, and depricate it.
117 This can't wait until we have a 2.0-gold release, if folks need
118 to move over to auth_dbm, we can't do that to them after 2.0 gold.
120 RELEASE NON-SHOWSTOPPERS BUT WOULD BE REAL NICE TO WRAP THESE UP:
122 * Allow the DocumentRoot directive within <Location > scopes? This
123 allows the beloved (crusty) Alias /foo/ /somepath/foo/ followed
124 by a <Directory /somepath/foo> to become simply
125 <Location /foo/> DocumentRoot /somefile/foo (IMHO a bit more legible
126 and in-your-face.) DocumentRoot unset would be accepted [and would
127 not permit content to be served, only virtual resources such as
128 server-info or server-status.
129 This proposed change would _not_ depricate Alias.
131 * add mod_gz to httpd-2.0 (in modules/experimental/)
132 +1: Greg, Justin, ben, Ken, Jeff, Bill
133 0: Cliff (there's now another candidate to be evaluated)
134 0: Jim (premature decision at present, IMO)
137 non-binding: Ian (+1)
139 * daedalus: mod_cgid and suexec have a problem co-existing. suexec
140 sees a null command string sometimes. The problem happens when
141 you access bugs.apache.org, then click on the "search the bug db"
144 * Win32: Rotatelogs sometimes is not terminated when Apache
145 goes down hard. FirstBill was looking at possibly tracking the
146 child's-child processes in the parent process.
147 OtherBill asks, wasn't this fixed?
149 * Win32: Add a simple hold console open patch (wait for close or
150 the ESC key, with a nice message) if the server died a bad
151 death (non-zero exit code) in console mode.
152 Resolution: bring forward same ugly hacks from 1.3.13-.20
154 * Port of mod_ssl to Apache 2.0:
156 The current porting state is summarized in modules/ssl/README. The
157 remaining work includes:
158 (1) stablizing/optimizing the SSL filter logic
159 (2) Enabling the various SSL caching mechanisms (shmcb, shmht)
160 (3) Enabling SSL extentions
161 (4) Trying to seperate the https filter logic from mod_ssl -
162 This is to facilitate other modules that wish to use the https
163 filter or the mod_ssl logic or both as required.
165 * Performance: Get the SINGLE_LISTEN_UNSERIALIZED_ACCEPT
166 optimization working in threaded. prefork's new design for how
167 to notice data on the pod should be sufficient.
169 * Performance & Debug: Eliminate most (and perhaps all) of the
170 malloc/free calls in the bucket brigade code. Need some
171 light weight memory management functions that allow freeing
172 memory (putting it back into a memory pool) when it is no
173 longer needed. Enabling simple debugging features like guard
174 bands, double free detection, etc. would be cool but certainly
175 not a hard requirement.
177 Status: Cliff started to implement this using SMS as has
178 been discussed at length for months, but since
179 SMS is not being used anywhere else in the server,
180 several people expressed the opinion that we should
181 get rid of it entirely, meaning that the buckets
182 need their own memory management (free list) functions.
183 Cliff will implement that this weekend so we at least
184 have something to look at/compare with.
186 * Eliminate unnecessary creation of pipes in mod_cgid
188 * the autoconf setup should be fixed to default to using the
189 "Apache" layout from config.layout, and each variable settable
190 in a layout should be overridable on the command line. Plus,
191 what we do right now just doesn't seem to fully fit into how autoconf
192 works, eg. AC_PREFIX_DEFAULT issues.
193 Message-ID: <Pine.BSF.4.20.0104031557420.20876-100000@alive.znep.com>
195 * Combine log_child and piped_log_spawn. Clean up http_log.c.
198 * Document mod_file_cache.
200 * OS/2: Make mod_status work for spmt_os2 MPM.
202 * Platforms that do not support fork (primarily Win32 and AS/400)
203 Architect start-up code that avoids initializing all the modules
204 in the parent process on platforms that do not support fork.
206 * Win32: Migrate the MPM over to use APR thread/process calls. This
207 would eliminate some code in the Win32 branch that essentially
208 duplicates what is in APR.
210 * There are still a number of places in the code where we are
211 losing error status (i.e. throwing away the error returned by a
212 system call and replacing it with a generic error code)
214 * Mass vhosting version of suEXEC.
216 * All DBMs suffer from confusion in support/dbmmanage (perl script) since
217 the dbmmanage employs the first-matched dbm format. This is not
218 necessarily the library that Apache was built with. Aught to
219 rewrite dbmmanage upon installation to bin/ with the proper library
220 for predictable mod_auth_db/dbm administration.
221 Status: Mladen Turk has posted several patches and ideas.
222 Key question, part of htpasswd, or a seperate utility?
224 prefer seperate: OtherBill
227 * use apu_dbm in mod_auth_dbm
228 Status: Greg +1 (low-priority volunteer)
229 Justin says: "Seems like this is already there, so should we just
230 remove the other DBM code in that file? If you want
231 to use gdbm, or dbm, etc, you should tell apr-util."
232 Will says: "bs - I may choose the fastest - most efficient native
233 dbm implementation, for shared proc caches, ssl session
234 caching, etc, but that has nothing to do with maintaining
235 a userlist via dbm, which has to remain readable between
236 builds/machines, etc. The use-multiple database schema
237 for apr-util would let us do this with just apr, though."
240 Some additional items remaining:
241 - case_preserved_filename stuff
242 (use the new canonical name stuff?)
243 - find a new home for ap_text(_header)
244 - is it possible to remove the DAV: namespace stuff from util_xml?
246 * ap_core_translate() and its use by mod_mmap_static and mod_file_cache
247 are a bit wonky. The function should probably be exposed as a utility
248 function (such as ap_translate_url2fs() or ap_validate_fs_url() or
249 something). Another approach would be a new hook phase after
250 "translate" which would allow the module to munge what the
251 translation has decided to do.
252 Status: Greg +1 (volunteers), Ryan +1
254 * Explore use of a post-config hook for the code in http_main.c which
255 calls ap_fixup_virutal_hosts(), ap_fini_vhost_config(), and
256 ap_sort_hooks() [to reduce the logic in main()]
258 * read the config tree just once, and process N times (as necessary)
260 * (possibly) use UUIDs in mod_unique_id and/or mod_usertrack
262 * (possibly) port the bug fix for PR 6942 (segv when LoadModule is put
263 into a VirtualHost container) to 2.0.
265 * shift stuff to mod_core.h
267 * APR-ize resolver stuff in mod_unique_id (Jeff volunteers)
269 * callers of ap_run_create_request() should check the return value
270 for failure (Doug volunteers)
272 * when prefork bails out due to an initialization error in the detached
273 process (e.g., mutex init failure), other children (cgid, at least)
274 are left hanging around
276 * Win32: Get Apache working on Windows 95/98. The following work
277 (at least) needs to be done:
278 - winnt MPM: Fix 95/98 code paths in the winnt MPM. There is some NT
279 specific code that is still not in NT only code paths
280 - IOL binds to APR sendfile, implemented with TransmitFile, which
281 is not available on 95/98.
282 - Document warning that OSR2 is required (for Crypt functions, in
283 rand.c, at least.) This could be resolved with an SSL library, or
284 randomization in APR itself.
285 - Bring the Win9xConHook.dll from 1.3 into 2.0 (no sense till it
286 actually works) and add in a splash of Win9x service code.
288 * When --enable-modules=most or --enable-mods-shared=most is passed
289 to the configure script it is still possible to enable many of the
290 modules while mod_so itself is not buildable (for whatever reason,
291 like not having APR_HAS_DSO). We need to emit an error message
292 and fail the configure script at this point.
294 PRs that have been suspended forever waiting for someone to
295 put them into 'the next release':
298 missing call to "setlocale();"
302 Additional status for XBitHack directive
306 Mod_proxy doesn't allow change of error pages
310 Modified PATH environemnt variable is not passed, instead
315 Proxy doesn't deliver documents if not connected
319 proxy converts ~name to %7Ename when name starts with a dot (.)
323 mod_access syntax allows hosts that should be restricted
326 * PR#557: mod_auth-any
327 ~UserHome directories are not honored in absolute pathname
332 Proxy FTP Authentication Fails
335 * PR#623: mod_include
336 A smarter "Last Modified" value for SSI documents (see PR number 600)
340 Request of "Options SymLinksIfGroupMatch"
344 Proxy doesn't do links right for OpenVMS files through ftp:
348 imap should read <MAP><AREA>*</MAP> too!
352 RLimitCPU and RLimitMEM don't apply to all children like they should
356 Uses cwd before filling it in, doesn't use syslog
360 it is useful to allow specifiction that root-owned symlinks
361 should always be followed
365 Controlling Access to Remote Proxies would be nice...
369 Adding authentication "on the fly" through the proxy module
372 * PR#1004: apache-api
373 request_config field in request_rec is moderately bogus
377 DoS attacks involving memory consumption
380 * PR#1050: mod_log-any
381 Logging of virtual server to error_log as well
385 ProxyRemote make a dead cycle.
388 * PR#1117: mod_auth-any
389 Using NIS passwd.byname dbm files with AuthDBMUserFile
393 suexec does not parse arguments to #exec cmd
396 * PR#1145: mod_include
397 Allow for Last-Modified: without resorting to XBitHack
400 * PR#1158: apache-api
401 improvements to child spawning API
405 ``nph-'' not honored (no buffering) for ProxyRemote mapping
409 Apache cannot handle continuation line in headers
413 setlogin() is not called, causing problems with e.g. identd
417 regerror() exists, use it
420 * PR#1233: apache-api
421 there is no way to keep per-connection per-module state
424 * PR#1263: mod_autoexec
425 Add frame-safe anchor attribute to mod_autoindex links
429 CGI scripts running as Apache user: security (suexec etc.)
433 Error messages could be easier to spot in cgi.log file for suexec.c
436 * PR#1287: mod_access
437 add allow,deny/deny,allow warning to mod_access
441 Need to know "hit-rate" on proxy cache
444 * PR#1358: mod_log-any
445 Selective url-encode of log fields (or maybe a pseudo
449 * PR#1383: mod_headers
450 I make mod_headers to modify request headers as well as
455 Proxy transfer logging
459 No HTTP_X_FORWARDED_FOR set...
463 ProxyRemote proxy requests fail authentication by firewall
466 * PR#1582: mod_rewrite
467 mod_rewrite forms REQUEST_URI different than mod_cgi does
470 * PR#1677: mod_headers
471 mod_headers should allow mod_log_config-style formats in
476 mod_proxy to support persistent conns?
479 * PR#1803: mod_include
480 patches to mod_include to allow for file tests
483 * PR#1809: mod_auth-any
484 Suggestion for improving authentication modules and core source
485 code, problem with 401 and ErrorDocument
489 listing of proxy cache content
493 Allow modules to set user:group for execution.
496 * PR#2024: apache-api
497 adding auth_why to conn_rec
500 * PR#2073: mod_log-any
501 pipelined connections are not logged correctly
504 * PR#2074: mod_rewrite
505 mod_rewrite doesn't pass Proxy Throughput on internal subrequests
509 HTTP Server Rebuild Line Needs Changing for the better
512 * PR#2138: mod_status
513 mod_status always displays 256 possible connection slots
516 * PR#2221: documentation
517 Make online documentation search link back to my installation
521 Can not POST to ErrorDocument - Apache/1.3b6
525 patterns in ProxyRemote
528 * PR#2343: mod_status
529 Status module averages are for entire uptime
533 suexec for general access of user content?
537 Proposal for TimeZone directive
541 /server-info doesn't check for the virtual host to list the info
545 problem specifying ndbm library for build ?with autoconfigure
549 A small addition to rotatelogs.c to improve program functionality.
553 AllowOverride FileInfo is too coarse
557 TimeOut applies to output of CGI scripts
560 * PR#2512: mod_access
561 <IfDenied> directive wanted
565 CGI's for general use still have to be run as another user
570 Cache file names in Proxy module
574 [PATCH] User/Group for <Directory> and <Location> i.e. not only
575 in global and <Virtual>.
579 mailto tags and bundling bug report script
583 Support for System Resource Controller
587 When will Apache support P3P? Any Plans?
591 Feedback/Comment on APACI
595 Inclusion of RPM spec file in CVS/distributions
599 Propose that Apache recommend $UNIQUE_ID for all "session id"
604 suggestion: power up your Include directive :)
608 cannot limit some HTTP methods
611 * PR#3143: apache-api
612 No module specific data hook for per-connection data
615 * PR#3191: mod_negotiation
616 no way to set global quality-of-source (qs) coneg values
621 Accessing URL through proxy server corrupts data.
625 Some anonymous FTP URLs ask for authentication
629 New ErrorDocumentMatch directive
633 Need to be able to override shebang line to make CGI scripts
638 "Files" and "FilesMatch" regexp does not recognize bang as
642 * PR#4448: mod_log-any
643 Please allow CGI env variables (QUERY_STRING, ...) to be logged
647 * PR#4459: mod_include
648 Suggestion for better handling of Last-modified headers
652 mod_cgi prevents handling of OPTIONS requests
655 * PR#5713: os-windows
656 [PATCH] install as win32 service with domain account
660 AllowOverride should have a 'CheckNone' and 'AllowNone' argument
661 instead of only 'None'
664 Other bugs that need fixing:
666 * MaxRequestsPerChild measures connections, not requests.
667 Until someone has a better way, we'll probably just rename it
668 "MaxConnectionsPerChild".
670 * Regex containers don't work in an intutive way
671 Status: No one has come up with an efficient way to fix this
672 behavior. Dean has suggested getting rid of regex containers
675 * SIGSEGV on Linux (glibc 2.1.2) isn't caught properly by a
676 sigwaiting thread. We need to work around this, perhaps unless
677 there is hope soon for a fixed glibc.
679 * orig_ct in the byterange/multipart handling may not be
680 needed. Apache 1.3 just never stashed "multipart" into
681 r->content_type. We should probably follow suit since the
682 byterange stuff doesn't want the rest of the code to see the
683 multipart content-type; the other code should still think it is
684 dealing with the <orig_ct> stuff.
685 Status: Greg volunteers to investigate (esp. since he was most
686 likely the one to break it :-)
688 Other features that need writing:
690 * Finish infrastructure in core for async MPMs
693 * TODO in source -- just do an egrep on "TODO" and see what's there
697 * Jon Travis's <jtravis@covalent.net> patch to deal with thread-safe
698 issues with inet_ntoa. See message <20001201163220.A12827@covalent.net>
699 Status: This is being set aside until the IPv6 work is finished
700 so that we know exactly what is required.
702 * Martin Sojka <msojka@gmx.de>'s patch to add error reporting for failed
703 htpasswd actions due to a full /tmp volume (other programs may have
708 * Mike Abbott's <mja@trudge.engr.sgi.com> patches to improve
710 Status: These were written for 1.3, and are awaiting a port to
713 * Jim Winstead's <jimw@trainedmonkey.com> patch to add CookieDomain and
714 other small mod_usertrack features
716 * Dan Rench's <drench@xnet.com> patch to add allow the errmsg and timefmt
717 of SSI's to be modified in the config file. Patch is available in
722 * Which MPMs will be included with Apache 2.0?