STATUS revision 70f6f32765cfaadd6da8de6f0fea97ddd72d8fad
2a6c49cfaef5979a5a06098f3ce987cd76769409manojApache 2.0 STATUS:
70f6f32765cfaadd6da8de6f0fea97ddd72d8fadmanojLast modified at [$Date: 2000/03/11 05:42:50 $]
3304cbd819df02e7548e9338dc0afa8d3ba29358manoj Vague plan, assuming more features aren't demanded for the 2.0 release
70f6f32765cfaadd6da8de6f0fea97ddd72d8fadmanoj 2.0a1 : released March 10, 2000
3304cbd819df02e7548e9338dc0afa8d3ba29358manoj 2.0 : later in 2000
2a6c49cfaef5979a5a06098f3ce987cd76769409manojRELEASE SHOWSTOPPERS:
40a5b7189dbbb28e107bf008ee625f2f0142c2ccdgaudet * Fix SAFEREAD
85cbdc16ac57fa68ce1358a308269abcd417f4d9stoddard * Fix lingering close
85cbdc16ac57fa68ce1358a308269abcd417f4d9stoddard * MPMs should have responsibility for closing sockets and
85cbdc16ac57fa68ce1358a308269abcd417f4d9stoddard registering the socket cleanup. Stop registering multiple
85cbdc16ac57fa68ce1358a308269abcd417f4d9stoddard cleanups for each socket. Move socket close code out of
85cbdc16ac57fa68ce1358a308269abcd417f4d9stoddard http_connection.c and into the MPMs.
2aae6faee508221efbeaba5547ca79b7a20ef047stoddard * Reliable piped logs look broken everywhere. Each MPM includes essentially
2aae6faee508221efbeaba5547ca79b7a20ef047stoddard identical code to ap_register_other_child(), etc. Most of this code can
2aae6faee508221efbeaba5547ca79b7a20ef047stoddard be moved out of the MPMs and into some common file (http_core.c?).
40a5b7189dbbb28e107bf008ee625f2f0142c2ccdgaudet Dean says presumably you mean an os-specific file?
75960f20f88dad6bc67892c711c429946063d133stoddard * DSO hooks registered during the preflight call to ap_read_config
75960f20f88dad6bc67892c711c429946063d133stoddard are not 'unregistered' during the preflight clean-up. Thus, when
75960f20f88dad6bc67892c711c429946063d133stoddard Apache comes up, the hook tables contain two entries for each DSO hook,
75960f20f88dad6bc67892c711c429946063d133stoddard one of which is bogus, leftover from the preflight.
4065b438067f3f08d0bd98b31ac4085b581b931dstoddard * Pipes to CGI scripts are not being timed out
2739add8add66e21526ce27f83c2ae133d089070rbb Status: code has been added to APR to support timing out pipes.
2739add8add66e21526ce27f83c2ae133d089070rbb This needs to be used in Apache now.
211bf1d44f4653bf753a15740cd5ebbf330b6e93manoj * Put back resource limit code
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj * suEXEC doesn't work
369edcdd0a9c5516c61e736ec2a6fc8fb0d92fe2manoj Status: Manoj has posted an patch to fix this.
369edcdd0a9c5516c61e736ec2a6fc8fb0d92fe2manoj <19991103003605.A20612@samosa.mindspring.com>
85cbdc16ac57fa68ce1358a308269abcd417f4d9stoddard * Win32: Enable the Windows MPM to honor max_requests_per_child
85cbdc16ac57fa68ce1358a308269abcd417f4d9stoddard Status: Bill will fix this.
75960f20f88dad6bc67892c711c429946063d133stoddard * Win32: Fix Win9* specific code in the winnt MPM
75960f20f88dad6bc67892c711c429946063d133stoddard * Win32: Test access logging with multiple threads. Will the
75960f20f88dad6bc67892c711c429946063d133stoddard native file I/O calls serialize automagically like the
75960f20f88dad6bc67892c711c429946063d133stoddard CRT calls or do we need to add region locking each time
75960f20f88dad6bc67892c711c429946063d133stoddard we access the logs?
75960f20f88dad6bc67892c711c429946063d133stoddard * Win32: Smoke test all ported modules to make sure they work
75960f20f88dad6bc67892c711c429946063d133stoddard correctly under Windows.
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj * Current 2.0 code is not tested on many Unix platforms. Make 2.0
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj work on most, if not all the systems 1.3 did
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj Status: Autoconf and APR will get us most of the way there.
70f6f32765cfaadd6da8de6f0fea97ddd72d8fadmanoj * The handler hook needs to use the new registration system
56ca30c968906053ae61acb218420667bb58d996rbb * Modules. Which modules do not work yet, can we get a list?
a9e07e4f90adcc7bc768db3055431c3dcd560cd1manoj * We need a thread-safe resolver, at least on Unix.
a9e07e4f90adcc7bc768db3055431c3dcd560cd1manoj Status: The best known candidate would be something from BIND
a9e07e4f90adcc7bc768db3055431c3dcd560cd1manoj (v8 or v9?) The only other option would be to mutex all the
a9e07e4f90adcc7bc768db3055431c3dcd560cd1manoj resolver calls,
2a6c49cfaef5979a5a06098f3ce987cd76769409manojRELEASE NON-SHOWSTOPPERS BUT WOULD BE REAL NICE TO WRAP THESE UP:
8de99b4c89d4ed4292a7dca42dd8a96b9a7c456fdgaudet * "make distclean" leaves (at least these) turds:
8de99b4c89d4ed4292a7dca42dd8a96b9a7c456fdgaudet modules/.deps
6686c651e8eb6ad762b842ad5139515c8038dd44dgaudet * apaci permissions problem:
6686c651e8eb6ad762b842ad5139515c8038dd44dgaudet cvs checkout apache-2.0
6686c651e8eb6ad762b842ad5139515c8038dd44dgaudet ./configure
6686c651e8eb6ad762b842ad5139515c8038dd44dgaudet now apaci is no longer executable and future makes fail
0b0a5225c5ed94b9f689839a14842ad4a24215e9dgaudet * the top-level configure.in appears to use $USE_MAINTAINER_MODE = yes
0b0a5225c5ed94b9f689839a14842ad4a24215e9dgaudet to set -Wall/etc. flags, and the APR configure.in appears to use
0b0a5225c5ed94b9f689839a14842ad4a24215e9dgaudet --with-debug. both seem like more of a hardwired pain compaired
0b0a5225c5ed94b9f689839a14842ad4a24215e9dgaudet to the pre-2.0 method of setting EXTRA_CFLAGS.
f824925ac58ff729289c017235eeb3bdd21ec3a2stoddard * Win32: Reuse accept socket after transmitfile/close
f824925ac58ff729289c017235eeb3bdd21ec3a2stoddard This is not a bug, but would be nice to get this feature in
f824925ac58ff729289c017235eeb3bdd21ec3a2stoddard before ship.
2aae6faee508221efbeaba5547ca79b7a20ef047stoddard * Win32: Enable the winnt MPM to use the new scoreboard API
75960f20f88dad6bc67892c711c429946063d133stoddard * Win32: Non-blocking CGI pipes
75960f20f88dad6bc67892c711c429946063d133stoddard * Win32: Graceful restart of a service is broken (code does shutdown then start)
a5ed555df952c85bc1b179f5981e8a6c54ba16e6stoddard * Dynamically loadable MPMs
a5ed555df952c85bc1b179f5981e8a6c54ba16e6stoddard Enable users of the binary distributions of Apache to be able to
a5ed555df952c85bc1b179f5981e8a6c54ba16e6stoddard dynamically load the MPM.
40a5b7189dbbb28e107bf008ee625f2f0142c2ccdgaudet Dean says this probably isn't a good idea. For example consider
40a5b7189dbbb28e107bf008ee625f2f0142c2ccdgaudet -lpthread, you couldn't have the same core run with pthreads,
40a5b7189dbbb28e107bf008ee625f2f0142c2ccdgaudet prefork, and gnu pth without a lot of headaches is my guess.
a5ed555df952c85bc1b179f5981e8a6c54ba16e6stoddard * Platforms that do not support fork (primarily Win32 and AS/400)
a5ed555df952c85bc1b179f5981e8a6c54ba16e6stoddard Consider introducing HAVE_FORK feature macro. Architect start-up code
a5ed555df952c85bc1b179f5981e8a6c54ba16e6stoddard that avoids initializing all the modules in the parent process on
a5ed555df952c85bc1b179f5981e8a6c54ba16e6stoddard platforms that do not support fork.
56ca30c968906053ae61acb218420667bb58d996rbb * Clean the code. There are a lot of places we used APR but didn't
56ca30c968906053ae61acb218420667bb58d996rbb remove the hacks that were required for the cross-platform code in
56ca30c968906053ae61acb218420667bb58d996rbb 1.3. We need to make the code look like APR was supposed to be there.
70f6f32765cfaadd6da8de6f0fea97ddd72d8fadmanoj * Go throught the 1.3 Bug DB and research the bugs marked "suspended".
70f6f32765cfaadd6da8de6f0fea97ddd72d8fadmanoj People were told these would be considered for inclusion in Apache 2.0,
70f6f32765cfaadd6da8de6f0fea97ddd72d8fadmanoj it would be nice to actually do so.
0bff2f28ef945280c17099c142126178a78e1e54manoj * Win32: Migrate the MPM over to use APR thread/process calls. This
0bff2f28ef945280c17099c142126178a78e1e54manoj would eliminate some code in the Win32 branch that essentially
0bff2f28ef945280c17099c142126178a78e1e54manoj duplicates what is in APR.
0bff2f28ef945280c17099c142126178a78e1e54manoj Status: Bill <stoddard@raleigh.ibm.com> is working on this.
9c09943bad734ebd5c7cc10bd6d63b75c4c6e056stoddard * Move I/O layering into APR.
0bff2f28ef945280c17099c142126178a78e1e54manoj * There are still a number of places in the code where we are
0bff2f28ef945280c17099c142126178a78e1e54manoj loosing error status (i.e. throwing away the error returned by a
0bff2f28ef945280c17099c142126178a78e1e54manoj system call and replacing it with a generic error code)
75960f20f88dad6bc67892c711c429946063d133stoddard * Win32: Implement reliable piped logs on Windows
0bff2f28ef945280c17099c142126178a78e1e54manoj Status: Bill <stoddard@raleigh.ibm.com> has prototype code (not
0bff2f28ef945280c17099c142126178a78e1e54manoj reliable) in 1.3. Should be much cleaner with APR in 2.0.
aa1faea36e4ae357bc603a2337b6adc54f5daec1manoj * Switch to autoconf, et al. for configuration.
f03d292915be9977eaf74e9be7b0404aec226f84manoj Status: Manoj has placed an initial version into the 2.0
f03d292915be9977eaf74e9be7b0404aec226f84manoj repository. Todos include:
3304cbd819df02e7548e9338dc0afa8d3ba29358manoj - "make install" should work
f03d292915be9977eaf74e9be7b0404aec226f84manoj - a tool to simplfy third-party module building should be
f03d292915be9977eaf74e9be7b0404aec226f84manoj written. Something like apxs
f03d292915be9977eaf74e9be7b0404aec226f84manoj - Add a lot more checks to satisfy the various platforms, e.g.
f03d292915be9977eaf74e9be7b0404aec226f84manoj for threading
aa1faea36e4ae357bc603a2337b6adc54f5daec1manoj * Use APR to get rid of more platform dependancies.
aa1faea36e4ae357bc603a2337b6adc54f5daec1manoj Status: Ryan Bloom <rbb@raleigh.ibm.com> is working on this.
302dc1f7b3feee23a91ad8f3cf3cb2edd95a557bmanoj * The connection status table is not very efficient. Also, very few stats
302dc1f7b3feee23a91ad8f3cf3cb2edd95a557bmanoj are exported to the connection status table (easy to fix), and mod_status
2a6c49cfaef5979a5a06098f3ce987cd76769409manojOther bugs that need fixing:
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj * MaxRequestsPerChild measures connections, not requests.
d5defd5a0c5cdbaf74b85939484dc2b6c8317d19manoj Until someone has a better way, we'll probably just rename it
d5defd5a0c5cdbaf74b85939484dc2b6c8317d19manoj "MaxConnectionsPerChild".
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj * Regex containers don't work in an intutive way
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj Status: No one has come up with an efficient way to fix this
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj behavior. Dean has suggested getting rid of regex containers
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj completely.
0bff2f28ef945280c17099c142126178a78e1e54manoj * SIGSEGV on Linux (glibc 2.1.2) isn't caught properly by a
0bff2f28ef945280c17099c142126178a78e1e54manoj sigwaiting thread. We need to work around this, perhaps unless
0bff2f28ef945280c17099c142126178a78e1e54manoj there is hope soon for a fixed glibc.
2a6c49cfaef5979a5a06098f3ce987cd76769409manojOther features that need writing:
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj * Finish infrastructure in core for async MPMs
40a5b7189dbbb28e107bf008ee625f2f0142c2ccdgaudet Status: post 2.0
dbf0c7bef06259486cd2748a2d0e82f27e099d6efielding * TODO in source -- just do an egrep on "TODO" and see what's there
2a6c49cfaef5979a5a06098f3ce987cd76769409manojDocumentation that needs writing:
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj * The concept of MPMs, especially if we ship more than one MPM for a
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj given platform
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj * New directives in the various MPMs
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj * API documentation
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj Status: Ben Laurie has written some hooks documentation
d6b3cb141f0667101c1bca883ad15b383402c93bfielding * Changes since 1.3.9 can be more easily seen in the commitlog file
d6b3cb141f0667101c1bca883ad15b383402c93bfielding dev.apache.org:/home/cvs/CVSROOT/commitlogs/apache-2.0
d6b3cb141f0667101c1bca883ad15b383402c93bfielding which includes some of Roy's comments when the changes were
d6b3cb141f0667101c1bca883ad15b383402c93bfielding committed in rough change-sets by purpose. Note that the commitlog
d6b3cb141f0667101c1bca883ad15b383402c93bfielding does not show the contents of new files until later.
2a6c49cfaef5979a5a06098f3ce987cd76769409manojAvailable Patches:
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj * Mike Abbott's <mja@trudge.engr.sgi.com> patches to improve
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj performance
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj Status: These were written for 1.3, and are awaiting a port to
2a6c49cfaef5979a5a06098f3ce987cd76769409manojOpen issues:
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj * What do we do about mod_proxy?
2a6c49cfaef5979a5a06098f3ce987cd76769409manoj * Which MPMs will be included with Apache 2.0?
d6b3cb141f0667101c1bca883ad15b383402c93bfielding * Is conf/highperformance.conf-dist obsolete? It looks obsolete.