2 Last modified at [$Date: 1999/11/04 18:25:23 $]
6 2.0: In pre-alpha development
10 * Vague goal of an alpha or beta release in 1999. Commit-then-review
15 * Pipes to CGI scripts are not being timed out
19 Status: Paul Reder is working on this.
22 Status: Paul Reder is working on this.
25 Status: Manoj has posted an patch to fix this.
26 <19991103003605.A20612@samosa.mindspring.com>
28 * Windows NT port isn't done
29 Status: Bill is working on MPM and APR. Remaining work:
30 1. Add back ability to run Apache as a service
31 2. Fix Win9* specific code in the winnt MPM
32 3. Get the MPM working in multi process mode (one parent & one child)
33 4. Test access logging with multiple threads. Will the native file I/O
34 calls serialize automagically like the CRT calls or do we need to
35 add region locking each time we write to the access/error logs?
37 * Current 2.0 code is not tested on many Unix platforms. Make 2.0
38 work on most, if not all the systems 1.3 did
40 Status: Autoconf and APR will get us most of the way there.
42 * The module API is a weird combination of the old table and the new
43 hook system. Switch completely to the new hook system
44 Status: Ben Laurie is working on this.
46 RELEASE NON-SHOWSTOPPERS BUT WOULD BE REAL NICE TO WRAP THESE UP:
48 * Win32: Redirect stderr to the error log. This is tougher than it appears
49 because Apache for Windows now uses native file I/O libraries rather than
50 the CRT library. No problem redirecting native I/O STDERR to the error log.
51 However, we need to replace all the fprintf(stderr...) calls with something
52 that will write to the native stderr handle instead of the CRT stderr handle.
55 * Win32: Migrate the MPM over to use APR thread/process calls.
57 * Move I/O layering into APR.
59 * There are still a number of places in the code where we are loosing error
60 status (i.e. throwing away the error returned by a system call and replacing
61 it with a generic error code)
63 * APRize the stat function. CRT stat() is about 30% slower on Windows than
64 the equivalent native Windows call. The APR call should return the Unix
65 style stat structure, just to keep it familier.
67 * Implement reliable piped logs on Windows
68 Status: Bill has prototype code (not reliable) in 1.3. Should be much
69 cleaner with APR in 2.0.
71 * Switch to autoconf, et al. for configuration.
73 * Use APR to get rid of more platform dependancies.
74 Status: Ryan Bloom <rbb@raleigh.ibm.com> is working on this.
76 * The connection status table is not very efficient. Also, very few stats
77 are exported to the connection status table (easy to fix), and mod_status
80 Other bugs that need fixing:
82 * MaxRequestsPerChild measures connections, not requests.
83 Until someone has a better way, we'll probably just rename it
84 "MaxConnectionsPerChild".
86 * Regex containers don't work in an intutive way
87 Status: No one has come up with an efficient way to fix this
88 behavior. Dean has suggested getting rid of regex containers
91 * Areas where APRFile is being used need to be cleaned-up
93 * SIGSEGV on Linux seems to only kill a thread, not a whole process;
94 we need to work around this, probably by bouncing the signal to
95 the sigwait thread. But this will hurt debugability.
97 Other features that need writing:
99 * Finish infrastructure in core for async MPMs
102 * TODO in source -- just do an egrep on "TODO" and see what's there
104 * Odd comments in source (egrep for "ZZZ") need to be cleaned-up
106 * Work on other MPMs. Possible MPMs discussed or in progress include:
108 - Dean Gaudet's async MPM
111 - Zach Brown's <zab@zabbo.net> Linux siginfo MPM
114 Documentation that needs writing:
116 * The concept of MPMs, especially if we ship more than one MPM for a
119 * New directives in the various MPMs
122 Status: Ben Laurie has written some hooks documentation
123 (apache-2.0/htdocs/hooks.html)
125 * Changes since 1.3.9 can be more easily seen in the commitlog file
126 dev.apache.org:/home/cvs/CVSROOT/commitlogs/apache-2.0
127 which includes some of Roy's comments when the changes were
128 committed in rough change-sets by purpose. Note that the commitlog
129 does not show the contents of new files until later.
133 * Mike Abbott's <mja@trudge.engr.sgi.com> patches to improve
135 Status: These were written for 1.3, and are awaiting a port to
140 * What do we do about mod_proxy?
142 * Which MPMs will be included with Apache 2.0?
144 * Is conf/highperformance.conf-dist obsolete? It looks obsolete.