2 Last modified at [$Date: 2000/04/13 00:38:55 $]
6 2.0a3 : vague plans to release in April sometime.
7 2.0a2 : released March 31, 2000
8 2.0a1 : released March 10, 2000
11 * Re-work configuration for top level Apache. Work should start with
12 trying to clean the autoconf stuff. If and only if this proves
13 impossible to do (very unlikely), autoconf should be removed and we
14 will roll our own config implementation.
15 Status: Jim Jagielski is looking into this.
17 * Put buffered file I/O back into APR. This should be done using the
18 buffering logic from the OS/2 directory. stdio FILE *'s should not be
19 used on Unix because of the limitations these impose on some platforms,
20 for example the 256 File * limit on Solaris.
23 see <Pine.LNX.4.21.0004101511380.7822-100000@twinlark.arctic.org>
25 * MPMs should have responsibility for closing sockets and
26 registering the socket cleanup. Stop registering multiple
27 cleanups for each socket. Move socket close code out of
28 http_connection.c and into the MPMs.
31 * Reliable piped logs look broken everywhere. Each MPM includes essentially
32 identical code to ap_register_other_child(), etc. Most of this code can
33 be moved out of the MPMs and into some common file (http_core.c?).
34 Dean says presumably you mean an os-specific file?
36 * Pipes to CGI scripts are not being timed out
37 Status: code has been added to APR to support timing out pipes.
38 This needs to be used in Apache now.
40 * Put back resource limit code
43 Status: Manoj has posted an patch to fix this.
44 <19991103003605.A20612@samosa.mindspring.com>
46 * Win32: Enable the Windows MPM to honor max_requests_per_child
47 Status: Bill will fix this.
49 * Win32: Get Apache working on Windows 95/98. The following work
50 (at least) needs to be done:
51 - winnt MPM: Fix 95/98 code paths in the winnt MPM. There is some NT
52 specific code that is still not in NT only code paths
53 - APR sendfile uses TransmitFile which is not available on
55 - ap_stat uses GetFileAttributeEx which is not available on
60 * Win32: Test access logging with multiple threads. Will the
61 native file I/O calls serialize automagically like the
62 CRT calls or do we need to add region locking each time
66 * Win32: Smoke test all ported modules to make sure they work
67 correctly under Windows.
70 * Current 2.0 code is not tested on many Unix platforms. Make 2.0
71 work on most, if not all the systems 1.3 did
72 Status: Autoconf and APR will get us most of the way there.
74 * The handler hook needs to use the new registration system
76 * Modules. Which modules do not work yet, can we get a list?
78 * We need a thread-safe resolver, at least on Unix.
79 Status: The best known candidate would be something from
82 RELEASE NON-SHOWSTOPPERS BUT WOULD BE REAL NICE TO WRAP THESE UP:
83 * Configuration Modules. Allow a new module type that reads
84 a config file and sets up an internal tree structure for
87 * "make distclean" leaves (at least these) turds:
90 lib/apr/shmem/unix/mm/config.log
91 lib/apr/shmem/unix/mm/mm_conf.h
92 lib/apr/shmem/unix/mm/libtool
93 lib/apr/shmem/unix/mm/config.status
94 lib/apr/shmem/unix/mm/Makefile
95 lib/apr/shmem/unix/mm/mm-config
101 * the top-level configure.in appears to use $USE_MAINTAINER_MODE = yes
102 to set -Wall/etc. flags, and the APR configure.in appears to use
103 --with-debug. both seem like more of a hardwired pain compaired
104 to the pre-2.0 method of setting EXTRA_CFLAGS.
106 * Win32: Reuse accept socket after transmitfile/close
107 This is not a bug, but would be nice to get this feature in
110 * Win32: Enable the winnt MPM to use the new scoreboard API
112 * Win32: Non-blocking CGI pipes
114 * Win32: Graceful restart of a service is broken (code does shutdown then start)
116 * Dynamically loadable MPMs
117 Enable users of the binary distributions of Apache to be able to
118 dynamically load the MPM.
120 Dean says this probably isn't a good idea. For example consider
121 -lpthread, you couldn't have the same core run with pthreads,
122 prefork, and gnu pth without a lot of headaches is my guess.
124 * Platforms that do not support fork (primarily Win32 and AS/400)
125 Consider introducing HAVE_FORK feature macro. Architect start-up code
126 that avoids initializing all the modules in the parent process on
127 platforms that do not support fork.
129 * Clean the code. There are a lot of places we used APR but didn't
130 remove the hacks that were required for the cross-platform code in
131 1.3. We need to make the code look like APR was supposed to be there.
133 * Go throught the 1.3 Bug DB and research the bugs marked "suspended".
134 People were told these would be considered for inclusion in Apache 2.0,
135 it would be nice to actually do so.
137 * Win32: Migrate the MPM over to use APR thread/process calls. This
138 would eliminate some code in the Win32 branch that essentially
139 duplicates what is in APR.
141 Bill says we need a new procattr, APR_CREATE_SUSPENDED (or
142 something similar) to direct ap_create_process to create the
143 process suspended. We also need a call to wake up the suspended
144 process This may not be able to be implemented everywhere though.
148 * Move I/O layering into APR.
150 * There are still a number of places in the code where we are
151 loosing error status (i.e. throwing away the error returned by a
152 system call and replacing it with a generic error code)
154 * Win32: Implement reliable piped logs on Windows
157 * Switch to autoconf, et al. for configuration.
158 Status: Manoj has placed an initial version into the 2.0
159 repository. Todos include:
161 - "make install" should work
162 - a tool to simplfy third-party module building should be
163 written. Something like apxs
164 - Add a lot more checks to satisfy the various platforms, e.g.
167 * Use APR to get rid of more platform dependancies.
168 Status: Ryan Bloom <rbb@covalenet.net> is working on this.
170 * The connection status table is not very efficient. Also, very few stats
171 are exported to the connection status table (easy to fix), and mod_status
174 Other bugs that need fixing:
176 * MaxRequestsPerChild measures connections, not requests.
177 Until someone has a better way, we'll probably just rename it
178 "MaxConnectionsPerChild".
180 * Regex containers don't work in an intutive way
181 Status: No one has come up with an efficient way to fix this
182 behavior. Dean has suggested getting rid of regex containers
185 * SIGSEGV on Linux (glibc 2.1.2) isn't caught properly by a
186 sigwaiting thread. We need to work around this, perhaps unless
187 there is hope soon for a fixed glibc.
189 Other features that need writing:
191 * Finish infrastructure in core for async MPMs
194 * TODO in source -- just do an egrep on "TODO" and see what's there
196 Documentation that needs writing:
197 * Mod_status docs are needed.
199 * The concept of MPMs, especially if we ship more than one MPM for a
202 * New directives in the various MPMs and appropriate links from
203 obsolete directives in core.html to the MPM documentation.
205 * Revise manual/stopping.html and the last part of
206 manual/misc/perf-tuning.html to take account of the MPMs.
209 Status: Ben Laurie has written some hooks documentation
210 (apache-2.0/htdocs/hooks.html)
212 * Changes since 1.3.9 can be more easily seen in the commitlog file
213 dev.apache.org:/home/cvs/CVSROOT/commitlogs/apache-2.0
214 which includes some of Roy's comments when the changes were
215 committed in rough change-sets by purpose. Note that the commitlog
216 does not show the contents of new files until later.
220 * Mike Abbott's <mja@trudge.engr.sgi.com> patches to improve
222 Status: These were written for 1.3, and are awaiting a port to
227 * What do we do about mod_proxy?
229 * Which MPMs will be included with Apache 2.0?
231 * Is conf/highperformance.conf-dist obsolete? It looks obsolete.