1 APACHE 2.5 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 separately 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
17 Patches considered for backport are noted in their branches' STATUS:
19 * http://svn.apache.org/repos/asf/httpd/httpd/branches/2.0.x/STATUS
20 * http://svn.apache.org/repos/asf/httpd/httpd/branches/2.2.x/STATUS
21 * http://svn.apache.org/repos/asf/httpd/httpd/branches/2.4.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 2.5.0 : In Development.
31 Contributors looking for a mission:
33 * Just do an egrep on "TODO" or "XXX" in the source.
35 * Review the bug database at: http://issues.apache.org/bugzilla/
37 * Review the "PatchAvailable" bugs in the bug database:
39 https://issues.apache.org/bugzilla/buglist.cgi?bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&product=Apache+httpd-2&keywords=PatchAvailable
41 After testing, you can append a comment saying "Reviewed and tested".
43 * Open bugs in the bug database.
45 * See also the STATUS file in the docs/ directory, which lists documentation-specific TODO items.
48 CURRENT RELEASE NOTES:
53 THINGS THAT SHOULD BE CONSIDERED EARLY IN THE 2.6/3.0 DEVELOPMENT CYCLE:
55 * Change default prefix from /usr/local/apache2 to something corresponding
56 to the project name. Rename apachectl.
58 * Change merge order of <Location> to be most specific match last. This
59 is more consistent with <Directory> and allows some optimizations for the
62 * Detect Lua 5.2.0 during configure and add LUA_COMPAT_ALL to CPPFLAGS.
63 Maybe it even suffices to add LUA_COMPAT_MODULE and individually
64 care about the two remaining incompatible code lines (one with lua_strlen,
67 OLD ISSUES THAT WERE THOUGHT TO BE SHOWSTOPPERS FOR 2.4 BUT OBVIOUSLY WEREN'T:
69 * Handling of non-trailing / config by non-default handler is broken
70 http://marc.theaimsgroup.com/?l=apache-httpd-dev&m=105451701628081&w=2
71 jerenkrantz asks: Why should this block a release?
72 wsanchez agrees: this may be a change in behavior, but isn't
73 clearly wrong, and even if so, it doesn't seem like a
76 * the edge connection filter cannot be removed
77 http://marc.theaimsgroup.com/?l=apache-httpd-dev&m=105366252619530&w=2
79 jerenkrantz asks: Why should this block a release?
81 stas replies: because it requires a rewrite of the filters stack
82 implementation (you have suggested that) and once 2.2 is
83 released you can't do that anymore.
85 pgollucci: this affects mod_perl I'm pretty sure.
89 * Name the Server (version 2.4 or 3.0, depending on the final call)
90 Recent discussion indicates we should designate a (short name).
91 This is not yet a [Vote] - Your nominations please:
92 * Apache HTTP Server (httpd)
93 +1: sctemme (why mess with it?), pgollucci
95 RELEASE NON-SHOWSTOPPERS BUT WOULD BE REAL NICE TO WRAP THESE UP:
97 * Clean up all the kruft and *extremely* outdated stuff below...
99 * Maybe remove Limit/LimitExcept or at least make it log warnings when
102 * Patches submitted to the bug database:
103 http://issues.apache.org/bugzilla/buglist.cgi?bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&product=Apache+httpd-2&keywords=PatchAvailable
105 * Filter stacks and subrequests, redirects and fast redirects.
106 There's at least one PR that suffers from the current unclean behaviour
107 (which lets the server send garbage): PR 17629
108 nd says: Every subrequest should get its own filter stack with the
109 subreq_core filter as bottom-most. That filter does two things:
110 - swallow EOS buckets
111 - redirect the data stream to the upper request's (rr->main)
112 filter chain directly after the subrequest's starting
114 Once we have a clean solution, we can try to optimize
115 it, so that the server won't be slow down too much.
117 * RFC 2616 violations.
118 Closed PRs: 15852, 15857, 15859, 15861, 15864, 15869, 15870, 16120,
119 16125, 16135, 16136, 16137, 16138, 16139, 16140, 16518,
121 Open PRs: 15865, 15866, 15868, 16126, 16133, 16142, 16521, 42978
122 jerenkrantz says: need to decide how many we need to backport and/or
123 if these rise to showstopper status.
124 wrowe suggests: it would be nice to see "MUST" v.s. "SHOULD" v.s. "MAY"
125 out of this list, without reviewing them individually.
127 * pipes deadlock on all platforms with limited pipe buffers (e.g. both
128 Linux and Win32, as opposed to only Win32 on 1.3). The right solution
129 is either GStein's proposal for a "CGI Brigade", or OtherBill's proposal
130 for "Poll Buckets" for "Polling Filter Chains". Or maybe both :-)
132 * All handlers should always send content down even if r->header_only
133 is set. If not, it means that the HEAD requests don't generate the
134 same headers as a GET which is wrong.
136 * exec cmd and suexec arg-passing enhancements
137 Status: Patches proposed
138 Message-ID: <20020526041748.A29148@prodigy.Redbrick.DCU.IE>
139 (see the "proc.patch" and "suexec-shell.patch" links in this message)
141 * The 2.0.36 worker MPM graceless shutdown changes work but are
142 a bit clunky on some platforms; eg, on Linux, the loop to
143 join each worker thread seems to hang, and the parent ends up
144 killing off the child with SIGKILL. But at least it shuts down.
146 chrisd: Has this been fixed by the changes for PR 38737?
148 * We do not properly substitute the prefix-variables in the configuration
149 scripts or generated-configs. (i.e. if sysconfdir is etc,
150 httpd-std.conf points to conf.)
152 * If any request gets through ap_process_request_internal() and is
153 scheduled to be served by the core handler, without a flag that this
154 r->filename was tested by dir/file_walk, we need to 500 at the very
155 end of the ap_process_request_internal() processing so sub_req-esters
156 know this request cannot be run. This provides authors of older
157 modules better compatibility, while still improving the security and
160 Status: still need to decide where this goes, OtherBill comments...
161 Message-ID: <065701c14526$495203b0$96c0b0d0@roweclan.net>
162 [Deleted comments regarding the ap_run_handler phase, as irrelevant
163 as BillS points out that "common case will be caught in
164 default_handler already (with the r->finfo.filetype == 0 check)"
165 and the issue is detecting this -before- we try to run the req.]
167 gregames says: can this happen somehow without a broken module
168 being involved? If not, why waste cycles trying to defend against
169 potential broken modules? It seems futile.
170 wrowe counters: no, it shouldn't happen unless the module is broken.
171 But the right answer is to fail the request up-front in dir/file
172 walk if the path was entirely invalid; and we can't do that either
173 UNTIL 2.1 or we break modules that haven't hooked map_to_storage.
175 * Can a static httpd be built reliably?
176 Message-ID: <20020207142751.T31582@clove.org>
178 * Usage of APR_BRIGADE_NORMALIZE in core_input_filter should be
181 <Pine.LNX.4.33.0201202232430.318-100000@deepthought.cs.virginia.edu>
182 Jeff wonders if we still care about this. It is no longer an
183 API issue but simply an extra trip through the brigade.
185 * Try to get libtool inter-library dependency code working on AIX.
186 Message-ID: <cm3n10lx555.fsf@rdu163-40-092.nc.rr.com>
188 Justin says: If we get it working on AIX, we can enable this
189 on all platforms and clean up our build system somewhat.
190 Jeff says: I thought I tested a patch for you sometime in
191 January that you were going to commit within a few days.
193 * Handling of %2f in URIs. Currently both 1.3 and 2.0
194 completely disallow %2f in the request URI path (see
195 ap_unescape_url() in util.c). It's permitted and passed
196 through in the query string, however. Roy says the
197 original reason for disallowing it, from five years ago,
198 was to protect CGI scripts that applied PATH_INFO to
199 a filesystem location and which might be tricked by
200 ..%2f..%2f(...). We *should* allow path-info of the
201 form 'http://foo.com/index.cgi/path/to/path%2finfo'.
202 Since we've revamped a lot of our processing of path
203 segments, it would be nice to allow this, or at least
204 allow it conditionally with a directive.
206 OtherBill adds that %2f as the SECOND character of a multibyte
207 sequence causes the request to fail! This happens notably in
210 * There is increasing demand from module writers for an API
211 that will allow them to control the server à la apachectl.
212 Reasons include sole-function servers that need to die if
213 an external dependency (e.g., a database) fails, et cetera.
214 Perhaps something in the (ever more abused) scoreboard?
216 On the other hand, we already have a pipe that goes between parent
217 and child for graceful shutdown events, along with an API that
218 can be used to send a message down that pipe. In threaded MPMs,
219 it is easy enough to make that one pipe be used for graceful
220 and graceless events, and it is also easy to open that pipe
221 to both parent and child for writing. Then we just need to
222 figure out how to do graceless on non-threaded MPMs.
224 * Allow the DocumentRoot directive within <Location > scopes? This
225 allows the beloved (crusty) Alias /foo/ /somepath/foo/ followed
226 by a <Directory /somepath/foo> to become simply
227 <Location /foo/> DocumentRoot /somefile/foo (IMHO a bit more legible
228 and in-your-face.) DocumentRoot unset would be accepted [and would
229 not permit content to be served, only virtual resources such as
230 server-info or server-status.
231 This proposed change would _not_ depricate Alias.
232 striker: See the thread starting with Message-ID:
233 JLEGKKNELMHCJPNMOKHOGEEJFBAA.striker@apache.org.
235 * Win32: Rotatelogs sometimes is not terminated when Apache
236 goes down hard. FirstBill was looking at possibly tracking the
237 child's-child processes in the parent process.
238 stoddard: Shared scoreboard might offer a good way for the parent
239 to keep track of 'other child' processes and whack them if the child
241 Other thoughts on walking the process chain using the NT kernel
242 have also been proposed on APR.
244 * Eliminate unnecessary creation of pipes in mod_cgid
246 * Combine log_child and piped_log_spawn. Clean up http_log.c.
249 * Platforms that do not support fork (primarily Win32 and AS/400)
250 Architect start-up code that avoids initializing all the modules
251 in the parent process on platforms that do not support fork.
253 * There are still a number of places in the code where we are
254 losing error status (i.e. throwing away the error returned by a
255 system call and replacing it with a generic error code)
257 * Mass vhosting version of suEXEC.
259 * All DBMs suffer from confusion in support/dbmmanage (perl script) since
260 the dbmmanage employs the first-matched dbm format. This is not
261 necessarily the library that Apache was built with. Aught to
262 rewrite dbmmanage upon installation to bin/ with the proper library
263 for predictable mod_auth_dbm administration.
264 Questions; htdbm exists, time to kill dbmmanage, or does it remain
265 useful as a perl dbm management example? If we keep it,
266 do we address the issue above?
269 Some additional items remaining:
270 - case_preserved_filename stuff
271 (use the new canonical name stuff?)
272 - find a new home for ap_text(_header)
273 - is it possible to remove the DAV: namespace stuff from util_xml?
275 * ap_core_translate() and its use by mod_mmap_static and mod_file_cache
276 are a bit wonky. The function should probably be exposed as a utility
277 function (such as ap_translate_url2fs() or ap_validate_fs_url() or
278 something). Another approach would be a new hook phase after
279 "translate" which would allow the module to munge what the
280 translation has decided to do.
281 Status: Greg +1 (volunteers)
283 * Explore use of a post-config hook for the code in http_main.c which
284 calls ap_fixup_virutal_hosts(), ap_fini_vhost_config(), and
285 ap_sort_hooks() [to reduce the logic in main()]
287 * read the config tree just once, and process N times (as necessary)
289 * (possibly) use UUIDs in mod_unique_id and/or mod_usertrack
291 * (possibly) port the bug fix for PR 6942 (segv when LoadModule is put
292 into a VirtualHost container) to 2.0.
294 * shift stuff to mod_core.h
296 * callers of ap_run_create_request() should check the return value
297 for failure (Doug volunteers)
299 * Fix the worker MPM to use POD to kill child processes instead
300 of ap_os_killpg, regardless of how they should die.
302 chrisd: Is this done, by any chance? See r92598 and r93358.
304 * Scoreboard structures could be changed in the future such that
305 proper alignment is not maintained, leading to segfaults on
306 some systems. Cliff posted a patch to deal with this issue but
307 later recanted. See this message to dev@apr.apache.org:
309 <Pine.LNX.4.44.0203011354090.16457-200000@deepthought.cs.virginia.edu>
311 * APXS either needs to be fixed completely for use when apr is out of tree,
312 or it should drop query mode altogether, and we just grow an
313 httpd-config or similar arrangement.
314 To quote a discussion in STATUS earlier:
316 thommay: this doesn't fix all the problems with apxs and out of
317 tree apr/apr-util, but it's a good start. There's still the
318 query cases; but I'm beginning to think that in these cases
319 the app should be querying ap{r,u}-config directly
320 deprecate -q: add htpd-config: gstein, pquerna, minfrin, pgollucci
323 TODO ISSUES REMAINING IN MOD_SSL:
325 * Do we need SSL_set_read_ahead()?
327 * SSLRequire directive (parsing of) leaks memory
329 * Diffie-Hellman-Parameters for temporary keys are hardcoded in
330 ssl_engine_dh.c, while the comment in ssl_engine_kernel.c says:
331 "it is suggested that keys be changed daily or every 500
332 transactions, and more often if possible."
334 * ssl_var_lookup could be rewritten to be MUCH faster
336 * CRL callback should be pluggable
338 * init functions should return status code rather than ssl_die()
340 * ssl_engine_pphrase.c needs to be reworked so it is generic enough
341 to also decrypt proxy keys
344 * mod_proxy: Ability to run SSL over proxy gateway connections,
345 encrypting (or reencrypting) at the proxy.
347 * mod_cache: Handle ESI tags.
349 * mod_cache: Resolve issue of how to cache page fragments (or perhaps
350 -if- we want to cache page fragments). Today, mod_cache/mod_mem_cache
351 will cache #include 'virtual' requests (but not #include 'file'
352 requests). This was accomplished by making CACHE_IN a
353 CONTENT_SET-1 filter to force it to run before the SUBREQ_CORE
354 filter. But now responses cannot be cached that include the
355 effects of having been run through CONTENT_SET filters
356 (mod_deflate, mod_expires, etc). We could rerun all the
357 CONTENT_SET filters on the cached response, but this will not
358 work in all cases. For example, mod_expires relies on installing
359 the EXPIRATION filter during fixups. Contents served out of
360 mod_cache (out of the quick_handler) bypass -all- the request
361 line server hooks (Ryan really hated this. It is great for
362 performance, but bad because of the complications listed above).
364 mod_cache/mod_mem_cache/mod_cache_disk:
366 * mod_mem_cache: Consider adding a RevalidateTimeout directive to
367 specify time at which local cached content is to be revalidated
368 (ie, underlying file stat'ed to see if it has changed).
370 * mod_mem_cache/mod_cache_disk: Need to be able to query cache
371 status (num of entries, cache object properties, etc.).
372 mod_status could be extended to query optional hooks defined
373 by modules for the purpose of reporting module status.
374 mod_cache (et. al.) could define optional hooks that are called
375 to collect status. Status should be queryable by
377 jerenkrantz says: Yawn. Who cares.
379 * Regex containers don't work in an intutive way
380 Status: No one has come up with an efficient way to fix this
381 behavior. Dean has suggested getting rid of regex containers
383 OtherBill suggests: We at least seem to agree on eliminating
384 the <Container ~ foo> forms, and using only
385 <ContainerMatch foo> semantics.
387 * orig_ct in the byterange/multipart handling may not be
388 needed. Apache 1.3 just never stashed "multipart" into
389 r->content_type. We should probably follow suit since the
390 byterange stuff doesn't want the rest of the code to see the
391 multipart content-type; the other code should still think it is
392 dealing with the <orig_ct> stuff.
393 Status: Greg volunteers to investigate (esp. since he was most
394 likely the one to break it :-)
396 EXPERIMENTAL MODULES:
398 Experimental modules should eventually be be promoted to fully supported
399 status or removed from the repository entirely (ie, the
400 'experiment' failed). This section tracks what needs to happen to
401 get the modules promoted to fully supported status.