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83 * Clean up all the kruft and *extremely* outdated stuff below...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