1 APACHE 2.3 STATUS: -*-text-*-
2 Last modified at [$Date$]
4 The current version of this file can be found at:
6 * http://svn.apache.org/repos/asf/httpd/httpd/trunk/STATUS
8 Documentation status is maintained seperately and can be found at:
10 * docs/STATUS in this source tree, or
11 * http://svn.apache.org/repos/asf/httpd/httpd/trunk/docs/STATUS
13 Consult the following STATUS files for information on related projects:
15 * http://svn.apache.org/repos/asf/apr/apr/trunk/STATUS
16 * http://svn.apache.org/repos/asf/apr/apr-util/trunk/STATUS
18 Patches considered for backport are noted in their branches' STATUS:
20 * http://svn.apache.org/repos/asf/httpd/httpd/branches/1.3.x/STATUS
21 * http://svn.apache.org/repos/asf/httpd/httpd/branches/2.0.x/STATUS
22 * http://svn.apache.org/repos/asf/httpd/httpd/branches/2.2.x/STATUS
26 [NOTE that x.{odd}.z versions are strictly Alpha/Beta releases,
27 while x.{even}.z versions are Stable/GA releases.]
29 2.3.9 : In Development. Jim proposes T&R ~ November 14
30 2.3.8 : Tagged on August 24, 2010.
31 2.3.7 : Tagged on August 19, 2010, not released.
32 2.3.6 : Released on June 21, 2010.
33 2.3.5 : Released on January 26, 2010.
34 2.3.4 : Released on December 8, 2009.
35 2.3.3 : Tagged on November 11, 2009, not released.
36 2.3.2 : Tagged on March 23, 2009, not released.
37 2.3.1 : Tagged on January 2, 2009, not released.
38 2.3.0 : Tagged on December 6, 2008, not released.
40 Contributors looking for a mission:
42 * Just do an egrep on "TODO" or "XXX" in the source.
44 * Review the bug database at: http://issues.apache.org/bugzilla/
46 * Review the "PatchAvailable" bugs in the bug database:
48 https://issues.apache.org/bugzilla/buglist.cgi?bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&product=Apache+httpd-2&keywords=PatchAvailable
50 After testing, you can append a comment saying "Reviewed and tested".
52 * Open bugs in the bug database.
54 * See also the STATUS file in the docs/ directory, which lists documentation-specific TODO items.
57 CURRENT RELEASE NOTES:
64 * Modules that are not ready for production use should be marked as
65 experimental. Candidates:
69 * Review the example configuration. It should be based on current best
70 practices and not use deprecated features.
74 * Modules without documentation need to be moved to experimental or be
77 * Not all MPMs are updated to set conn_rec::current_thread correctly.
78 (Prefork, Worker, Event, Simple are updated).
79 jim sez: Then we just ship with those... mark any others as
82 * The grammar and interface of ap_expr should be something that we are
83 comfortable with keeping for the next few years.
88 * The mod_session* modules need to be checked that their hooks respect
89 the returning of int (HTTP status codes) and apr_status_t as appropriate,
90 and any anomolies fixed.
93 OLD ISSUES THAT WERE THOUGHT TO BE SHOWSTOPPERS FOR 2.2 BUT OBVIOUSLY WEREN'T:
95 * Handling of non-trailing / config by non-default handler is broken
96 http://marc.theaimsgroup.com/?l=apache-httpd-dev&m=105451701628081&w=2
97 jerenkrantz asks: Why should this block a release?
98 wsanchez agrees: this may be a change in behavior, but isn't
99 clearly wrong, and even if so, it doesn't seem like a
102 * the edge connection filter cannot be removed
103 http://marc.theaimsgroup.com/?l=apache-httpd-dev&m=105366252619530&w=2
105 jerenkrantz asks: Why should this block a release?
107 stas replies: because it requires a rewrite of the filters stack
108 implementation (you have suggested that) and once 2.2 is
109 released you can't do that anymore.
111 pgollucci: this affects mod_perl I'm pretty sure.
115 * Name the Server (version 2.4 or 3.0, depending on the final call)
116 Recent discussion indicates we should designate a (short name).
117 This is not yet a [Vote] - Your nominations please:
118 * Apache HTTP Server (httpd)
119 +1: sctemme (why mess with it?)
121 RELEASE NON-SHOWSTOPPERS BUT WOULD BE REAL NICE TO WRAP THESE UP:
123 * Add mod_allow_method or some other (usable) functionality to replace
126 * Sort out modules selections for most/all/reallyall. Maybe rename
127 all -> most, reallyall -> all, and remove the old 'most'.
131 * Patches submitted to the bug database:
132 http://issues.apache.org/bugzilla/buglist.cgi?bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&product=Apache+httpd-2&keywords=PatchAvailable
134 * Filter stacks and subrequests, redirects and fast redirects.
135 There's at least one PR that suffers from the current unclean behaviour
136 (which lets the server send garbage): PR 17629
137 nd says: Every subrequest should get its own filter stack with the
138 subreq_core filter as bottom-most. That filter does two things:
139 - swallow EOS buckets
140 - redirect the data stream to the upper request's (rr->main)
141 filter chain directly after the subrequest's starting
143 Once we have a clean solution, we can try to optimize
144 it, so that the server won't be slow down too much.
146 * RFC 2616 violations.
147 Closed PRs: 15852, 15857, 15859, 15861, 15864, 15869, 15870, 16120,
148 16125, 16135, 16136, 16137, 16138, 16139, 16140, 16518,
150 Open PRs: 15865, 15866, 15868, 16126, 16133, 16142, 16521
151 jerenkrantz says: need to decide how many we need to backport and/or
152 if these rise to showstopper status.
153 wrowe suggests: it would be nice to see "MUST" v.s. "SHOULD" v.s. "MAY"
154 out of this list, without reviewing them individually.
156 * There is a bug in how we sort some hooks, at least the pre-config
157 hook. The first time we call the hooks, they are in the correct
158 order, but the second time, we don't sort them correctly. Currently,
159 the modules/http/config.m4 file has been renamed to
160 modules/http/config2.m4 to work around this problem, it should moved
161 back when this is fixed.
163 OtherBill offers that this is a SERIOUS problem. We do not sort
164 correctly by the ordering arguments passed to the register hook
165 functions. This was proven when I reordered the open_logs hook
166 to attempt to open the error logs prior to the access logs. Possibly
167 the entire sorting code needs to be refactored.
169 * pipes deadlock on all platforms with limited pipe buffers (e.g. both
170 Linux and Win32, as opposed to only Win32 on 1.3). The right solution
171 is either GStein's proposal for a "CGI Brigade", or OtherBill's proposal
172 for "Poll Buckets" for "Polling Filter Chains". Or maybe both :-)
174 * All handlers should always send content down even if r->header_only
175 is set. If not, it means that the HEAD requests don't generate the
176 same headers as a GET which is wrong.
178 * exec cmd and suexec arg-passing enhancements
179 Status: Patches proposed
180 Message-ID: <20020526041748.A29148@prodigy.Redbrick.DCU.IE>
181 (see the "proc.patch" and "suexec-shell.patch" links in this message)
183 * The 2.0.36 worker MPM graceless shutdown changes work but are
184 a bit clunky on some platforms; eg, on Linux, the loop to
185 join each worker thread seems to hang, and the parent ends up
186 killing off the child with SIGKILL. But at least it shuts down.
188 chrisd: Has this been fixed by the changes for PR 38737?
190 * We do not properly substitute the prefix-variables in the configuration
191 scripts or generated-configs. (i.e. if sysconfdir is etc,
192 httpd-std.conf points to conf.)
194 * If any request gets through ap_process_request_internal() and is
195 scheduled to be served by the core handler, without a flag that this
196 r->filename was tested by dir/file_walk, we need to 500 at the very
197 end of the ap_process_request_internal() processing so sub_req-esters
198 know this request cannot be run. This provides authors of older
199 modules better compatibility, while still improving the security and
202 Status: still need to decide where this goes, OtherBill comments...
203 Message-ID: <065701c14526$495203b0$96c0b0d0@roweclan.net>
204 [Deleted comments regarding the ap_run_handler phase, as irrelevant
205 as BillS points out that "common case will be caught in
206 default_handler already (with the r->finfo.filetype == 0 check)"
207 and the issue is detecting this -before- we try to run the req.]
209 gregames says: can this happen somehow without a broken module
210 being involved? If not, why waste cycles trying to defend against
211 potential broken modules? It seems futile.
212 wrowe counters: no, it shouldn't happen unless the module is broken.
213 But the right answer is to fail the request up-front in dir/file
214 walk if the path was entirely invalid; and we can't do that either
215 UNTIL 2.1 or we break modules that haven't hooked map_to_storage.
217 * Can a static httpd be built reliably?
218 Message-ID: <20020207142751.T31582@clove.org>
220 * Usage of APR_BRIGADE_NORMALIZE in core_input_filter should be
223 <Pine.LNX.4.33.0201202232430.318-100000@deepthought.cs.virginia.edu>
224 Jeff wonders if we still care about this. It is no longer an
225 API issue but simply an extra trip through the brigade.
227 * Try to get libtool inter-library dependency code working on AIX.
228 Message-ID: <cm3n10lx555.fsf@rdu163-40-092.nc.rr.com>
230 Justin says: If we get it working on AIX, we can enable this
231 on all platforms and clean up our build system somewhat.
232 Jeff says: I thought I tested a patch for you sometime in
233 January that you were going to commit within a few days.
235 * Handling of %2f in URIs. Currently both 1.3 and 2.0
236 completely disallow %2f in the request URI path (see
237 ap_unescape_url() in util.c). It's permitted and passed
238 through in the query string, however. Roy says the
239 original reason for disallowing it, from five years ago,
240 was to protect CGI scripts that applied PATH_INFO to
241 a filesystem location and which might be tricked by
242 ..%2f..%2f(...). We *should* allow path-info of the
243 form 'http://foo.com/index.cgi/path/to/path%2finfo'.
244 Since we've revamped a lot of our processing of path
245 segments, it would be nice to allow this, or at least
246 allow it conditionally with a directive.
248 OtherBill adds that %2f as the SECOND character of a multibyte
249 sequence causes the request to fail! This happens notably in
252 * There is increasing demand from module writers for an API
253 that will allow them to control the server à la apachectl.
254 Reasons include sole-function servers that need to die if
255 an external dependency (e.g., a database) fails, et cetera.
256 Perhaps something in the (ever more abused) scoreboard?
258 On the other hand, we already have a pipe that goes between parent
259 and child for graceful shutdown events, along with an API that
260 can be used to send a message down that pipe. In threaded MPMs,
261 it is easy enough to make that one pipe be used for graceful
262 and graceless events, and it is also easy to open that pipe
263 to both parent and child for writing. Then we just need to
264 figure out how to do graceless on non-threaded MPMs.
266 * Allow the DocumentRoot directive within <Location > scopes? This
267 allows the beloved (crusty) Alias /foo/ /somepath/foo/ followed
268 by a <Directory /somepath/foo> to become simply
269 <Location /foo/> DocumentRoot /somefile/foo (IMHO a bit more legible
270 and in-your-face.) DocumentRoot unset would be accepted [and would
271 not permit content to be served, only virtual resources such as
272 server-info or server-status.
273 This proposed change would _not_ depricate Alias.
274 striker: See the thread starting with Message-ID:
275 JLEGKKNELMHCJPNMOKHOGEEJFBAA.striker@apache.org.
277 * Win32: Rotatelogs sometimes is not terminated when Apache
278 goes down hard. FirstBill was looking at possibly tracking the
279 child's-child processes in the parent process.
280 stoddard: Shared scoreboard might offer a good way for the parent
281 to keep track of 'other child' processes and whack them if the child
283 Other thoughts on walking the process chain using the NT kernel
284 have also been proposed on APR.
286 * Eliminate unnecessary creation of pipes in mod_cgid
288 * Combine log_child and piped_log_spawn. Clean up http_log.c.
291 * Platforms that do not support fork (primarily Win32 and AS/400)
292 Architect start-up code that avoids initializing all the modules
293 in the parent process on platforms that do not support fork.
295 * There are still a number of places in the code where we are
296 losing error status (i.e. throwing away the error returned by a
297 system call and replacing it with a generic error code)
299 * Mass vhosting version of suEXEC.
301 * All DBMs suffer from confusion in support/dbmmanage (perl script) since
302 the dbmmanage employs the first-matched dbm format. This is not
303 necessarily the library that Apache was built with. Aught to
304 rewrite dbmmanage upon installation to bin/ with the proper library
305 for predictable mod_auth_dbm administration.
306 Questions; htdbm exists, time to kill dbmmanage, or does it remain
307 useful as a perl dbm management example? If we keep it,
308 do we address the issue above?
311 Some additional items remaining:
312 - case_preserved_filename stuff
313 (use the new canonical name stuff?)
314 - find a new home for ap_text(_header)
315 - is it possible to remove the DAV: namespace stuff from util_xml?
317 * ap_core_translate() and its use by mod_mmap_static and mod_file_cache
318 are a bit wonky. The function should probably be exposed as a utility
319 function (such as ap_translate_url2fs() or ap_validate_fs_url() or
320 something). Another approach would be a new hook phase after
321 "translate" which would allow the module to munge what the
322 translation has decided to do.
323 Status: Greg +1 (volunteers)
325 * Explore use of a post-config hook for the code in http_main.c which
326 calls ap_fixup_virutal_hosts(), ap_fini_vhost_config(), and
327 ap_sort_hooks() [to reduce the logic in main()]
329 * read the config tree just once, and process N times (as necessary)
331 * (possibly) use UUIDs in mod_unique_id and/or mod_usertrack
333 * (possibly) port the bug fix for PR 6942 (segv when LoadModule is put
334 into a VirtualHost container) to 2.0.
336 * shift stuff to mod_core.h
338 * callers of ap_run_create_request() should check the return value
339 for failure (Doug volunteers)
341 * Fix the worker MPM to use POD to kill child processes instead
342 of ap_os_killpg, regardless of how they should die.
344 chrisd: Is this done, by any chance? See r92598 and r93358.
346 * Scoreboard structures could be changed in the future such that
347 proper alignment is not maintained, leading to segfaults on
348 some systems. Cliff posted a patch to deal with this issue but
349 later recanted. See this message to dev@apr.apache.org:
351 <Pine.LNX.4.44.0203011354090.16457-200000@deepthought.cs.virginia.edu>
353 * APXS either needs to be fixed completely for use when apr is out of tree,
354 or it should drop query mode altogether, and we just grow an
355 httpd-config or similar arrangement.
356 To quote a discussion in STATUS earlier:
358 thommay: this doesn't fix all the problems with apxs and out of
359 tree apr/apr-util, but it's a good start. There's still the
360 query cases; but I'm beginning to think that in these cases
361 the app should be querying ap{r,u}-config directly
362 deprecate -q: add htpd-config: gstein, pquerna, minfrin, pgollucci
365 TODO ISSUES REMAINING IN MOD_SSL:
367 * Do we need SSL_set_read_ahead()?
369 * SSLRequire directive (parsing of) leaks memory
371 * Diffie-Hellman-Parameters for temporary keys are hardcoded in
372 ssl_engine_dh.c, while the comment in ssl_engine_kernel.c says:
373 "it is suggested that keys be changed daily or every 500
374 transactions, and more often if possible."
376 * ssl_var_lookup could be rewritten to be MUCH faster
378 * CRL callback should be pluggable
380 * session cache store should be pluggable
382 * init functions should return status code rather than ssl_die()
384 * ssl_engine_pphrase.c needs to be reworked so it is generic enough
385 to also decrypt proxy keys
387 * output warning when allowing SSL v2.0 ? its so old
390 * mod_proxy: Ability to run SSL over proxy gateway connections,
391 encrypting (or reencrypting) at the proxy.
393 * mod_cache: Handle ESI tags.
395 * mod_cache: Resolve issue of how to cache page fragments (or perhaps
396 -if- we want to cache page fragments). Today, mod_cache/mod_mem_cache
397 will cache #include 'virtual' requests (but not #include 'file'
398 requests). This was accomplished by making CACHE_IN a
399 CONTENT_SET-1 filter to force it to run before the SUBREQ_CORE
400 filter. But now responses cannot be cached that include the
401 effects of having been run through CONTENT_SET filters
402 (mod_deflate, mod_expires, etc). We could rerun all the
403 CONTENT_SET filters on the cached response, but this will not
404 work in all cases. For example, mod_expires relies on installing
405 the EXPIRATION filter during fixups. Contents served out of
406 mod_cache (out of the quick_handler) bypass -all- the request
407 line server hooks (Ryan really hated this. It is great for
408 performance, but bad because of the complications listed above).
410 mod_cache/mod_mem_cache/mod_disk_cache:
412 * mod_mem_cache: Consider adding a RevalidateTimeout directive to
413 specify time at which local cached content is to be revalidated
414 (ie, underlying file stat'ed to see if it has changed).
416 * mod_cache: CacheEnable/CacheDisable should accept regular expressions.
417 jerenkrantz says: Too slow. Get regexs away from speedy caches by
418 default. Introduce a new CacheEnableRegex if you want.
420 * mod_mem_cache/mod_disk_cache: Need to be able to query cache
421 status (num of entries, cache object properties, etc.).
422 mod_status could be extended to query optional hooks defined
423 by modules for the purpose of reporting module status.
424 mod_cache (et. al.) could define optional hooks that are called
425 to collect status. Status should be queryable by
427 jerenkrantz says: Yawn. Who cares.
429 * Regex containers don't work in an intutive way
430 Status: No one has come up with an efficient way to fix this
431 behavior. Dean has suggested getting rid of regex containers
433 OtherBill suggests: We at least seem to agree on eliminating
434 the <Container ~ foo> forms, and using only
435 <ContainerMatch foo> semantics.
437 * orig_ct in the byterange/multipart handling may not be
438 needed. Apache 1.3 just never stashed "multipart" into
439 r->content_type. We should probably follow suit since the
440 byterange stuff doesn't want the rest of the code to see the
441 multipart content-type; the other code should still think it is
442 dealing with the <orig_ct> stuff.
443 Status: Greg volunteers to investigate (esp. since he was most
444 likely the one to break it :-)
446 EXPERIMENTAL MODULES:
448 Experimental modules should eventually be be promoted to fully supported
449 status or removed from the repository entirely (ie, the
450 'experiment' failed). This section tracks what needs to happen to
451 get the modules promoted to fully supported status.