1 APACHE 2.1 STATUS: -*-text-*-
2 Last modified at [$Date$]
4 Release [NOTE that only Alpha/Beta releases occur in 2.1 development]:
6 2.1.1 : Proposed roll on 11/14/2004 (around/after Hackathon).
7 Justin volunteers as RM.
10 Please consult the following STATUS files for information
14 * srclib/apr-util/STATUS
17 Contributors looking for a mission:
19 * Just do an egrep on "TODO" or "XXX" in the source.
21 * Review the "PatchAvailable" bugs in the bug database.
22 Append a comment saying "Reviewed and tested".
24 * Open bugs in the bug database.
26 CURRENT RELEASE NOTES:
30 * Handling of non-trailing / config by non-default handler is broken
31 http://marc.theaimsgroup.com/?l=apache-httpd-dev&m=105451701628081&w=2
32 jerenkrantz asks: Why should this block a release?
34 * the edge connection filter cannot be removed
35 http://marc.theaimsgroup.com/?l=apache-httpd-dev&m=105366252619530&w=2
36 jerenkrantz asks: Why should this block a release?
40 * Promote mod_cache from experimental to non-experimental
41 status (keep issues noted below in EXPERIMENTAL MODULES as
42 items to be addressed as a supported module).
43 +1: jerenkrantz, pquerna
46 There are a couple of problems that need to be resolved
47 before this module is moved out of experimental.
48 1) We need to at least review and comment on the RFC violations
49 2) Resolve issue of how to cache page fragements (or perhaps -if- we
50 want to cache page fragements). Today, mod_cache/mod_mem_cache
51 will cache #include 'virtual' requests (but not #include 'file'
52 requests). This was accomplished by making CACHE_IN a
53 CONTENT_SET-1 filter to force it to run before the SUBREQ_CORE
54 filter. But now responses cannot be cached that include the
55 effects of having been run through CONTENT_SET filters
56 (mod_deflate, mod_expires, etc). We could rerun all the
57 CONTENT_SET filters on the cached response, but this will not
58 work in all cases. For example, mod_expires relies on installing
59 the EXPIRATION filter during fixups. Contents served out of
60 mod_cache (out of the quick_handler) bypass -all- the request
61 line server hooks (Ryan really hated this. It is great for
62 performance, but bad because of the complications listed above).
64 jerenkrantz: I think it's time. We've done a *lot* of work to it, and
65 we think most of the blatant RFC violations are now gone.
66 mod_cache just belongs in cache/. There may still be bugs,
67 but not likely to be major ones. Note that I'm not moving
68 *anything* until we switch to SVN.
70 * httpd-std.conf and friends
72 a) httpd-std.conf should be tailored by install (from src or
73 binbuild) even if user has existing httpd.conf
74 +1: trawick, slive, gregames, ianh, Ken, wrowe, jwoolley, jim, nd,
76 wrowe - prefer httpd.default.conf to avoid ambiguity with cvs
78 b) tailored httpd-std.conf should be copied by install to
82 c) tailored httpd-std.conf should be installed to
83 sysconfdir/examples or manualdir/exampleconf/
84 +1: slive, trawick, Ken, nd (prefer the latter), erikabele
86 d) Installing a set of default config files when upgrading a server
87 doesn't make ANY sense at all.
88 +1: ianh - medium/big sites don't use 'standard config' anyway, as it
89 usually needs major customizations
90 -1: Ken, wrowe, jwoolley, jim, nd, erikabele
91 wrowe - diff is wonderful when comparing old/new default configs,
92 even for customized sites that ianh mentions
93 jim - ... assuming that the default configs have been updated
94 with the required inline docs to explain the
97 * If the parent process dies, should the remaining child processes
98 "gracefully" self-terminate. Or maybe we should make it a runtime
99 option, or have a concept of 2 parent processes (one being a
101 See: Message-ID: <3C58232C.FE91F19F@Golux.Com>
103 Self-destruct: Ken, Martin, Lars
104 Not self-destruct: BrianP, Ian, Cliff, BillS
105 Make it runtime configurable: Aaron, jim, Justin, wrowe, rederpj, nd
107 /* The below was a concept on *how* to handle the problem */
108 Have 2 parents: +1: jim
109 -1: Justin, wrowe, rederpj, nd
110 +0: Lars, Martin (while standing by, could it do
113 * Make the worker MPM the default MPM for threaded Unix boxes.
114 +1: Justin, Ian, Cliff, BillS, striker, wrowe, nd
115 +0: BrianP, Aaron (mutex contention is looking better with the
116 latest code, let's continue tuning and testing), rederpj, jim
119 pquerna: Do we want to change this for 2.2?
121 RELEASE NON-SHOWSTOPPERS BUT WOULD BE REAL NICE TO WRAP THESE UP:
123 * Patches submitted to the bug database:
124 http://nagoya.apache.org/bugzilla/buglist.cgi?bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&product=Apache+httpd-2.0&keywords=PatchAvailable
126 * Filter stacks and subrequests, redirects and fast redirects.
127 There's at least one PR that suffers from the current unclean behaviour
128 (which lets the server send garbage): PR 17629
129 nd says: Every subrequest should get its own filter stack with the
130 subreq_core filter as bottom-most. That filter does two things:
131 - swallow EOS buckets
132 - redirect the data stream to the upper request's (rr->main)
133 filter chain directly after the subrequest's starting
135 Once we have a clean solution, we can try to optimize
136 it, so that the server won't be slow down too much.
138 * RFC 2616 violations.
140 Open PRs: 15852, 15859, 15861, 15864, 15865, 15866, 15868, 15869,
141 15870, 16120, 16125, 16126, 16133, 16135, 16136, 16137,
142 16138, 16139, 16140, 16142, 16518, 16520, 16521,
143 jerenkrantz says: need to decide how many we need to backport and/or
144 if these rise to showstopper status.
145 wrowe suggests: it would be nice to see "MUST" v.s. "SHOULD" v.s. "MAY"
146 out of this list, without reviewing them individually.
148 * There is a bug in how we sort some hooks, at least the pre-config
149 hook. The first time we call the hooks, they are in the correct
150 order, but the second time, we don't sort them correctly. Currently,
151 the modules/http/config.m4 file has been renamed to
152 modules/http/config2.m4 to work around this problem, it should moved
153 back when this is fixed.
155 OtherBill offers that this is a SERIOUS problem. We do not sort
156 correctly by the ordering arguments passed to the register hook
157 functions. This was proven when I reordered the open_logs hook
158 to attempt to open the error logs prior to the access logs. Possibly
159 the entire sorting code needs to be refactored.
161 * pipes deadlock on all platforms with limited pipe buffers (e.g. both
162 Linux and Win32, as opposed to only Win32 on 1.3). The right solution
163 is either GStein's proposal for a "CGI Brigade", or OtherBill's proposal
164 for "Poll Buckets" for "Polling Filter Chains". Or maybe both :-)
166 * All handlers should always send content down even if r->header_only
167 is set. If not, it means that the HEAD requests don't generate the
168 same headers as a GET which is wrong.
170 * HP/UX 10.20: compile breakage in APR. Looks like it should be easy
171 to fix, probably just some extraneous #include's that are fouling
174 Jeff: See my reply and patch in the PR (and previous commit to
175 stop using "pipe" as a field name). If patch is committed, we
176 should be okay. I'll wait to see if the user tests the patch.
177 Update by Jeff 20020722: I got an account on HP 10.20. It looks
178 like some of the APR thread detection is screwed up. If we find
179 pthread.h but we can't compile the pthread test program we still
180 think we can use threads. For that reason, the patch I posted
181 to the PR won't work as-is since a failed compile of the test
182 program means nothing.
184 * exec cmd and suexec arg-passing enhancements
185 Status: Patches proposed
186 Message-ID: <20020526041748.A29148@prodigy.Redbrick.DCU.IE>
187 (see the "proc.patch" and "suexec-shell.patch" links in this message)
189 * The 2.0.36 worker MPM graceless shutdown changes work but are
190 a bit clunky on some platforms; eg, on Linux, the loop to
191 join each worker thread seems to hang, and the parent ends up
192 killing off the child with SIGKILL. But at least it shuts down.
194 * --enable-mods-shared="foo1 foo2" is busted on Darwin. Pier
195 posted a patch (Message-ID: <B8DBBE8D.575A%pier@betaversion.org>).
197 * We do not properly substitute the prefix-variables in the configuration
198 scripts or generated-configs. (i.e. if sysconfdir is etc,
199 httpd-std.conf points to conf.)
201 * If any request gets through ap_process_request_internal() and is
202 scheduled to be served by the core handler, without a flag that this
203 r->filename was tested by dir/file_walk, we need to 500 at the very
204 end of the ap_process_request_internal() processing so sub_req-esters
205 know this request cannot be run. This provides authors of older
206 modules better compatibility, while still improving the security and
209 Status: still need to decide where this goes, OtherBill comments...
210 Message-ID: <065701c14526$495203b0$96c0b0d0@roweclan.net>
211 [Deleted comments regarding the ap_run_handler phase, as irrelevant
212 as BillS points out that "common case will be caught in
213 default_handler already (with the r->finfo.filetype == 0 check)"
214 and the issue is detecting this -before- we try to run the req.]
216 gregames says: can this happen somehow without a broken module
217 being involved? If not, why waste cycles trying to defend against
218 potential broken modules? It seems futile.
219 wrowe counters: no, it shouldn't happen unless the module is broken.
220 But the right answer is to fail the request up-front in dir/file
221 walk if the path was entirely invalid; and we can't do that either
222 UNTIL 2.1 or we break modules that haven't hooked map_to_storage.
224 * With AP_MODE_EXHAUSTIVE in the core, it is finally clear to me
225 how the Perchild MPM should be re-written. It hasn't worked
226 correctly since filters were added because it wasn't possible to
227 get the content that had already been written and the socket at
228 the same time. This mode lets us do that, so the MPM can be
231 * Can a static httpd be built reliably?
232 Message-ID: <20020207142751.T31582@clove.org>
234 * [Ken] Test suite failures:
235 o worker is also failing some of the 'cgi' subtests
236 (see <URL:http://Source-Zone.Org/Apache/regression/>):
237 Justin says: "Worker should be fine and passes httpd-test here.
238 I think it's a perl or a httpd-test problem."
240 * Usage of APR_BRIGADE_NORMALIZE in core_input_filter should be
242 Message-ID: <Pine.LNX.4.33.0201202232430.318-100000@deepthought.cs.virginia.edu>
243 Jeff wonders if we still care about this. It is no longer an
244 API issue but simply an extra trip through the brigade.
246 * The Add...Filter and Set...Filter directives do not allow the
247 administrator to order filters, beyond the order of filename (mime)
248 extensions. It isn't clear if Set...Filter(s) should be inserted
249 before or after the Add...Filter(s) which are ordered by sequence of
250 filename extensions. At minimum, some sort of +-[0-10] syntax seems
251 like a nice solution. See ROADMAP.
253 * Get perchild to work on platforms other than Linux. This
254 will require a portable mechanism to pass data and file/socket
255 descriptors between vhost child groups. An API was proposed
257 Message-ID: <20020111115006.K1529@clove.org>
259 * Try to get libtool inter-library dependency code working on AIX.
260 Message-ID: <cm3n10lx555.fsf@rdu163-40-092.nc.rr.com>
262 Justin says: If we get it working on AIX, we can enable this
263 on all platforms and clean up our build system
265 Jeff says: I thought I tested a patch for you sometime in
266 January that you were going to commit within a few
269 * Handling of %2f in URIs. Currently both 1.3 and 2.0
270 completely disallow %2f in the request URI path (see
271 ap_unescape_url() in util.c). It's permitted and passed
272 through in the query string, however. Roy says the
273 original reason for disallowing it, from five years ago,
274 was to protect CGI scripts that applied PATH_INFO to
275 a filesystem location and which might be tricked by
276 ..%2f..%2f(...). We *should* allow path-info of the
277 form 'http://foo.com/index.cgi/path/to/path%2finfo'.
278 Since we've revamped a lot of our processing of path
279 segments, it would be nice to allow this, or at least
280 allow it conditionally with a directive.
282 OtherBill adds that %2f as the SECOND character of a multibyte
283 sequence causes the request to fail! This happens notably in
286 * FreeBSD, threads, and worker MPM. All seems to work fine
287 if you only have one worker process with many threads. Add
288 a second worker process and the accept lock seems to be
289 lost. This might be an APR issue with how it deals with
290 the child_init hook (i.e. the fcntl lock needs to be resynced).
291 More examination and analysis is required.
292 Status: This has also been reported on Cygwin.
293 FreeBSD 4.7 was reputed to have 'fixed' threads. Not.
294 FreeBSD 5.2-RC is a confirmed fix w/either libkse or libthr.
295 [libc_r, still the default, does not serve any pages w/worker;
296 so on FreeBSD 5.2, you must use libmap.conf (see man page).]
297 Work needs to be done to get APR to try to be knowledgable that
298 libkse/libthr are acceptable. Still not recommended for the
299 default since libc_r is still broken.
300 Message-ID: <3C2CC514.8EF3BED1@wapme-systems.de> (cygnus)
302 * There is increasing demand from module writers for an API
303 that will allow them to control the server à la apachectl.
304 Reasons include sole-function servers that need to die if
305 an external dependency (e.g., a database) fails, et cetera.
306 Perhaps something in the (ever more abused) scoreboard?
308 On the other hand, we already have a pipe that goes between parent
309 and child for graceful shutdown events, along with an API that
310 can be used to send a message down that pipe. In threaded MPMs,
311 it is easy enough to make that one pipe be used for graceful
312 and graceless events, and it is also easy to open that pipe
313 to both parent and child for writing. Then we just need to
314 figure out how to do graceless on non-threaded MPMs.
316 * Allow the DocumentRoot directive within <Location > scopes? This
317 allows the beloved (crusty) Alias /foo/ /somepath/foo/ followed
318 by a <Directory /somepath/foo> to become simply
319 <Location /foo/> DocumentRoot /somefile/foo (IMHO a bit more legible
320 and in-your-face.) DocumentRoot unset would be accepted [and would
321 not permit content to be served, only virtual resources such as
322 server-info or server-status.
323 This proposed change would _not_ depricate Alias.
324 striker: See the thread starting with Message-ID:
325 JLEGKKNELMHCJPNMOKHOGEEJFBAA.striker@apache.org.
327 * Win32: Rotatelogs sometimes is not terminated when Apache
328 goes down hard. FirstBill was looking at possibly tracking the
329 child's-child processes in the parent process.
330 stoddard: Shared scoreboard might offer a good way for the parent
331 to keep track of 'other child' processes and whack them if the child
333 Other thoughts on walking the process chain using the NT kernel
334 have also been proposed on APR.
336 * Eliminate unnecessary creation of pipes in mod_cgid
338 * Combine log_child and piped_log_spawn. Clean up http_log.c.
341 * Platforms that do not support fork (primarily Win32 and AS/400)
342 Architect start-up code that avoids initializing all the modules
343 in the parent process on platforms that do not support fork.
345 * There are still a number of places in the code where we are
346 losing error status (i.e. throwing away the error returned by a
347 system call and replacing it with a generic error code)
349 * Mass vhosting version of suEXEC.
351 * All DBMs suffer from confusion in support/dbmmanage (perl script) since
352 the dbmmanage employs the first-matched dbm format. This is not
353 necessarily the library that Apache was built with. Aught to
354 rewrite dbmmanage upon installation to bin/ with the proper library
355 for predictable mod_auth_dbm administration.
356 Questions; htdbm exists, time to kill dbmmanage, or does it remain
357 useful as a perl dbm management example? If we keep it,
358 do we address the issue above?
361 Some additional items remaining:
362 - case_preserved_filename stuff
363 (use the new canonical name stuff?)
364 - find a new home for ap_text(_header)
365 - is it possible to remove the DAV: namespace stuff from util_xml?
367 * ap_core_translate() and its use by mod_mmap_static and mod_file_cache
368 are a bit wonky. The function should probably be exposed as a utility
369 function (such as ap_translate_url2fs() or ap_validate_fs_url() or
370 something). Another approach would be a new hook phase after
371 "translate" which would allow the module to munge what the
372 translation has decided to do.
373 Status: Greg +1 (volunteers)
375 * Explore use of a post-config hook for the code in http_main.c which
376 calls ap_fixup_virutal_hosts(), ap_fini_vhost_config(), and
377 ap_sort_hooks() [to reduce the logic in main()]
379 * read the config tree just once, and process N times (as necessary)
381 * (possibly) use UUIDs in mod_unique_id and/or mod_usertrack
383 * (possibly) port the bug fix for PR 6942 (segv when LoadModule is put
384 into a VirtualHost container) to 2.0.
386 * shift stuff to mod_core.h
388 * callers of ap_run_create_request() should check the return value
389 for failure (Doug volunteers)
391 * Win32: Get Apache working on Windows 95/98. The following work
392 (at least) needs to be done:
393 - Document warning that OSR2 is required (for Crypt functions, in
394 rand.c, at least.) This could be resolved with an SSL library, or
395 randomization in APR itself.
396 - Bring the Win9xConHook.dll from 1.3 into 2.0 (no sense till it
397 actually works) and add in a splash of Win9x service code.
399 * Fix the worker MPM to use POD to kill child processes instead
400 of ap_os_killpg, regardless of how they should die.
402 * Scoreboard structures could be changed in the future such that
403 proper alignment is not maintained, leading to segfaults on
404 some systems. Cliff posted a patch to deal with this issue but
405 later recanted. See this message to dev@apr.apache.org:
406 Message-ID: <Pine.LNX.4.44.0203011354090.16457-200000@deepthought
409 * When sufficiently tested, the AllowEncodedSlashes/%2f patch
410 needs to be backported to 2.0 and 1.3.
412 * APXS either needs to be fixed completely for use when apr is out of tree,
413 or it should drop query mode altogether, and we just grow an
414 httpd-config or similar arrangement.
415 To quote a discussion in STATUS earlier:
417 thommay: this doesn't fix all the problems with apxs and out of
418 tree apr/apr-util, but it's a good start. There's still the
419 query cases; but I'm beginning to think that in these cases
420 the app should be querying ap{r,u}-config directly
421 gstein: agreed. apxs should deprecate the -q flag
423 TODO ISSUES REMAINING IN MOD_SSL:
425 * In order to use a DSO version of mod_ssl we have to link with
426 -lssl and -lcrypto. A workaround is in place right now where the
427 entire EXTRA_LIBS macro is being appended to the objects list, but
428 this is a hack. We should either revamp the APACHE_CHECK_SSL_TOOLKIT
429 autoconf function or come up with some other autoconf checks to
430 search for libssl and libcrypto and properly add them to mod_ssl's
433 * SSL renegotiations in combination with POST request
435 * Port or dispose all code inside #if 0...#endif blocks that remain
436 from the porting effort.
438 * Do we need SSL_set_read_ahead()?
440 * the ssl_expr api is NOT THREAD SAFE. race conditions exist:
441 -in ssl_expr_comp() if SSLRequire is used in .htaccess
442 (ssl_expr_info is global)
443 -is ssl_expr_eval() if there is an error
444 (ssl_expr_error is global)
446 * SSLRequire directive (parsing of) leaks memory
448 * Diffie-Hellman-Parameters for temporary keys are hardcoded in
449 ssl_engine_dh.c, while the comment in ssl_engine_kernel.c says:
450 "it is suggested that keys be changed daily or every 500
451 transactions, and more often if possible."
453 * ssl_var_lookup could be rewritten to be MUCH faster
455 * CRL callback should be pluggable
457 * session cache store should be pluggable
459 * init functions should return status code rather than ssl_die()
461 * ssl_engine_pphrase.c needs to be reworked so it is generic enough
462 to also decrypt proxy keys
464 * the shmcb code should just align its memory segment rather than
465 jumping through all the "safe" memcpy and memset hoops
468 * mod_proxy performance: when mod_proxy is configured to do proxy
469 gateway (aka reverse proxy), it would be nice to be able to reuse
470 connections to the backend servers. Now, connections to the
471 backend servers are taken down when the corresponding frontend
472 connection is taken down.
474 * mod_proxy: Ability to run SSL over proxy gateway connections,
475 encrypting (or reencrypting) at the proxy.
477 * mod_proxy: Add capability of mod_proxy to load balance across
478 a farm of backend servers.
480 * mod_cache: Handle ESI tags.
482 EXPERIMENTAL MODULES:
484 Experimental modules should eventually be be promoted to fully supported
485 status or removed from the repository entirely (ie, the
486 'experiment' failed). This section tracks what needs to happen to
487 get the modules promoted to fully supported status.
489 mod_cache/mod_mem_cache/mod_disk_cache:
491 * mod_mem_cache: Consider adding a RevalidateTimeout directive to
492 specify time at which local cached content is to be revalidated
493 (ie, underlying file stat'ed to see if it has changed).
495 * mod_cache: CacheEnable/CacheDisable should accept regular expressions.
496 jerenkrantz says: Too slow. Get regexs away from speedy caches by
497 default. Introduce a new CacheEnableRegex if you want.
499 * mod_cache: Fix dependency on ATOMIC operators. Need
500 APR_HAS_ATOMIC_* feature macros.
501 jerenkrantz says: APR 1.0+ has some guaranteed form of atomics.
503 * mod_mem_cache/mod_disk_cache: Need to be able to query cache
504 status (num of entries, cache object properties, etc.).
505 mod_status could be extended to query optional hooks defined
506 by modules for the purpose of reporting module status.
507 mod_cache (et. al.) could define optional hooks that are called
508 to collect status. Status should be queryable by
510 jerenkrantz says: Yawn. Who cares.
512 PRs that have been suspended forever waiting for someone to
513 put them into 'the next release':
515 * documentation and Q&A
517 PR#2221: Make online documentation search link back to my installation
520 PR#2906: Propose that Apache recommend $UNIQUE_ID for all "session id"
524 PR#2793: When will Apache support P3P? Any Plans?
529 PR#2113: HTTP Server Rebuild Line Needs Changing for the better
532 PR#2421: problem specifying ndbm library for build ?with autoconfigure
537 PR#76: missing call to "setlocale();"
540 PR#628: Request of "Options SymLinksIfGroupMatch"
543 PR#793: RLimitCPU and RLimitMEM don't apply to all children like they should
546 PR#922: it is useful to allow specifiction that root-owned symlinks
547 should always be followed
550 PR#1028: DoS attacks involving memory consumption
553 PR#1191: setlogin() is not called, causing problems with e.g. identd
556 PR#1204: regerror() exists, use it
559 PR#2284: Can not POST to ErrorDocument - Apache/1.3b6
562 PR#2396: Proposal for TimeZone directive
565 PR#2446: AllowOverride FileInfo is too coarse
568 PR#2760: [PATCH] User/Group for <Directory> and <Location> i.e. not only
569 in global and <Virtual>.
572 PR#2907: suggestion: power up your Include directive :)
575 PR#3018: cannot limit some HTTP methods
578 PR#3677: New ErrorDocumentMatch directive
581 PR#4244: "Files" and "FilesMatch" regexp does not recognize bang as
585 PR#5993: AllowOverride should have a 'CheckNone' and 'AllowNone' argument
586 instead of only 'None'
591 PR#537: mod_access syntax allows hosts that should be restricted
594 PR#1287: add allow,deny/deny,allow warning to mod_access
597 PR#2512: <IfDenied> directive wanted
602 PR#557: ~UserHome directories are not honored in absolute pathname
606 PR#1117: Using NIS passwd.byname dbm files with AuthDBMUserFile
609 PR#1809: Suggestion for improving authentication modules and core source
610 code, problem with 401 and ErrorDocument
615 PR#1263: Add frame-safe anchor attribute to mod_autoindex links
618 * mod_cgi (and suexec)
620 PR#921: suexec Uses cwd before filling it in, doesn't use syslog
623 PR#1176: Apache cannot handle continuation line in headers
626 PR#1120: suexec does not parse arguments to #exec cmd
629 PR#1268: CGI scripts running as Apache user: security (suexec etc.)
632 PR#1285: Error messages could be easier to spot in cgi.log file for suexec.c
635 PR#1905: suexec - Allow modules to set user:group for execution.
638 PR#2360: suexec for general access of user content?
641 PR#2460: TimeOut applies to output of CGI scripts
644 PR#2573: CGI's for general use still have to be run as another user
648 PR#4241: Need to be able to override shebang line to make CGI scripts
652 PR#4490: mod_cgi prevents handling of OPTIONS requests
657 PR#370: Modified PATH environemnt variable is not passed, instead
663 PR#1383: I make mod_headers to modify request headers as well as
667 PR#1677: mod_headers should allow mod_log_config-style formats in
673 PR#759: imap should read <MAP><AREA>*</MAP> too!
678 PR#78: Additional status for XBitHack directive
681 PR#623: A smarter "Last Modified" value for SSI documents (see PR number 600)
685 Allow for Last-Modified: without resorting to XBitHack
688 PR#1803: patches to mod_include to allow for file tests
691 PR#4459: Suggestion for better handling of Last-modified headers
696 PR#1050: Logging of virtual server to error_log as well
699 PR#1358: Selective url-encode of log fields (or maybe a pseudo
703 PR#2073: pipelined connections are not logged correctly
706 PR#4448: Please allow CGI env variables (QUERY_STRING, ...) to be logged
712 PR#3191: no way to set global quality-of-source (qs) coneg values
718 PR#362: Mod_proxy doesn't allow change of error pages
721 PR#440: Proxy doesn't deliver documents if not connected
724 PR#534: proxy converts ~name to %7Ename when name starts with a dot (.)
727 PR#612: Proxy FTP Authentication Fails
730 PR#700: Proxy doesn't do links right for OpenVMS files through ftp:
733 PR#980: Controlling Access to Remote Proxies would be nice...
736 PR#994: Adding authentication "on the fly" through the proxy module
739 PR#1085: ProxyRemote make a dead cycle.
742 PR#1166: ``nph-'' not honored (no buffering) for ProxyRemote mapping
745 PR#1290: Need to know "hit-rate" on proxy cache
748 PR#1532: Proxy transfer logging
751 PR#1547: No HTTP_X_FORWARDED_FOR set...
754 PR#1567: ProxyRemote proxy requests fail authentication by firewall
757 PR#1702: mod_proxy to support persistent conns?
760 PR#1878: listing of proxy cache content
763 PR#2314: patterns in ProxyRemote
766 PR#2648: Cache file names in Proxy module
769 PR#3568: Accessing URL through proxy server corrupts data.
772 PR#3605: Some anonymous FTP URLs ask for authentication
777 PR#1582: mod_rewrite forms REQUEST_URI different than mod_cgi does
780 PR#2074: mod_rewrite doesn't pass Proxy Throughput on internal subrequests
785 PR#2138: mod_status always displays 256 possible connection slots
788 PR#2343: Status module averages are for entire uptime
793 PR#1004: request_config field in request_rec is moderately bogus
796 PR#1158: improvements to child spawning API
799 PR#1233: there is no way to keep per-connection per-module state
802 PR#2024: adding auth_why to conn_rec
805 PR#2873: Feedback/Comment on APACI
808 PR#3143: No module specific data hook for per-connection data
811 * generally odds and ends
813 PR#2431: A small addition to rotatelogs.c to improve program functionality.
816 PR#2763: mailto tags and bundling bug report script
819 PR#2785: os-aix Support for System Resource Controller
822 PR#2889: Inclusion of RPM spec file in CVS/distributions
825 PR#5713: os-windows [PATCH] install as win32 service with domain account
826 Status: Cannot accept password-as-arg, we should prompt the
827 user when -k install/-k config with a user argument.
831 Other bugs that need fixing:
833 * ap_discard_request should be converted to use the bucket API
834 directly rather than waste cycles copying buffers with the old API.
836 * MaxRequestsPerChild measures connections, not requests.
837 Until someone has a better way, we'll probably just rename it
838 "MaxConnectionsPerChild".
840 * Regex containers don't work in an intutive way
841 Status: No one has come up with an efficient way to fix this
842 behavior. Dean has suggested getting rid of regex containers
844 OtherBill suggests: We at least seem to agree on eliminating
845 the <Container ~ foo> forms, and using only
846 <ContainerMatch foo> semantics.
848 * SIGSEGV on Linux (glibc 2.1.2) isn't caught properly by a
849 sigwaiting thread. We need to work around this, perhaps unless
850 there is hope soon for a fixed glibc.
852 * orig_ct in the byterange/multipart handling may not be
853 needed. Apache 1.3 just never stashed "multipart" into
854 r->content_type. We should probably follow suit since the
855 byterange stuff doesn't want the rest of the code to see the
856 multipart content-type; the other code should still think it is
857 dealing with the <orig_ct> stuff.
858 Status: Greg volunteers to investigate (esp. since he was most
859 likely the one to break it :-)
861 Binaries (probably not till beta):
863 Platform Avail. Volunteer
864 ------------------------------------------------------------------
865 AIX 4.3.3 no Bill Stoddard
868 hppa2.0w-hp-hpux11.00 no Cliff Woolley
869 i386-pc-solaris2.8 no Aaron Bannert
870 i386-unknown-freebsd4.5 no
871 i386-unknown-freebsd4.6 no Cliff Woolley
872 i686-pc-linux-gnu-slackware81 no Cliff Woolley
873 i686-pc-linux-gnu-rh70 no Aaron Bannert
874 i686-pc-linux-gnu-rh73 no Cliff Woolley
876 powerpc-apple-darwin5.5 no Aaron Bannert
877 powerpc-unknown-linux-gnu no Graham Leggett
878 s390-ibm-linux no Greg Ames
879 sparc-sun-solaris2.8 no Jim Jagielski
880 NetWare no Brad Nicholes
883 Win32-x86 no William Rowe