3 *) Change the "which call to module_init() is this" check to a more
4 reliable approach. Doru Petrescu [pdoru@kappa.ro]
6 *) Close the old pipe file descriptor in apache main on USR1/HUP
7 (elimnates a small leak). James E. Jurach Jr. jjurach@fundsxpress.com
9 *) Fix a bug in fcgi_config_set_authoritative_slot(). Tetsuya Furukawa
10 [tetsuya@secom-sis.co.jp]
12 *) Eliminate the use of locks to assist in the clean shutdown of
13 applications. Instead, it is assumed that applications handle
14 termination signals properly (this is now embedded in the C
17 *) Fix Win32 process termination. Proper operation requires the use of
18 an updated application lib (termination is now signalled with an
19 Event and handled by specialized thread).
21 *) Docs cleanup. Andrew Benham [adsb@bigfoot.com]
23 *) Added code so if the last instance of a dynamic application died without
24 provocation, then don't restart it if singleThreshold > 0 (i.e. if the
25 configuration allows the last instance to be killed, then allow it to die).
26 Andrew Benham [adsb@bigfoot.com]
28 *) Fix the loadFactor calculation used to determine when dyanmic
29 applications could be killed off due to low demand [adsb@bigfoot.com].
31 *) Fix a deadlock condition that could occur with Win32 named
34 *) Fix a potential deadlock condition when FastCGI application
35 sent responses while still reading the client request (POST data).
39 *) Allow absolute pathnames in the -socket argument. Suggested by
40 Christian Jaeger [christian.jaeger@sl.ethz.ch].
42 *) Don't invoke suexec when the user/group for the fastcgi application
43 is the same as the apache main server. This is consistent with
44 apache's suexec handling. Suggested by Nikolaus Rath [Nikolaus@rath.org].
46 *) Reset the apache drop dead timer upon successful read or writes
47 to/from the client. This eliminates timeouts that were occuring
48 during the large file transfers to/from slow clients.
50 *) Support generic wrappers such as cgiwrap by eliminating dependencies
51 on Apache's SUEXEC, renaming the FastCgiSuexec directive
52 FastCgiWrapper and eliminating any checks regarding the target
53 application (this is the repsonibility of the wrapper).
55 *) Fix a nasty bug that occurred when a client aborted a POST request
56 before the connection to a dynamic FastCGI application was opened.
57 The application's lock file descriptor wasn't setup but was being
58 closed which resulted in FD0 being closed. Normally this is open to
59 /dev/null and should pose no problems except that because the FD
60 was available it was being returned by Apache's accept(). This
61 caused it to be registered for pool cleanup. mod_cgi though moves
62 the CGI stdin pipe to FD0 and thus it was getting waxed during pool
63 cleanup. Problem identified by checksum@163.net.
65 Changes with mod_fastcgi 2.2.8
67 *) Eliminate the concept of disabled applications. If a failure
68 occurs trying to setup an application (e.g. bind() error) its
69 tried repeatedly every init-start-delay seconds.
71 *) Tweak to Makefile.tmpl to support DSOs. Dave Hill [ddhill@zk3.dec.com]
73 Changes with mod_fastcgi 2.2.6
75 *) Shutdown the PM when Apache appears to have disappeared.
77 *) seteuid() tweak for HP-UX 11. Milton L. Hankins [mlh@swl.msd.ray.com]
79 *) (Win32) More dynamic fixes.
81 *) (Win32) Eliminate the per application named pipe mutex. Its now
82 per process to keep the lib happy (the use _FCGI_MUTEX_ should
83 removed from the lib). This allows pipe-based applications to
84 now handle multiple simoultanous requests (one per process).
86 *) Increase the number of open FDs we look to close when spawning apps.
88 *) Prevent an assert from popping unnecessarily.
90 *) (Win32) Add support for interpretter scripts.
92 *) (Win32) Fix named pipe handling (problems with large responses).
94 *) (Win32) Remove the "can exec" check.
96 Changes with mod_fastcgi 2.2.4
98 *) Beta WinNT support. David Allen [djallen@raleigh.ibm.com] and
99 Rob Saccoccio [robs@chelsea.net]
101 *) Remove request type restriction (GET and POST only) in order to support
102 Web DAV requests. Scott Robertson [sroberts@codeit.com]
104 *) Allow requests for URLs such as /server/some/other/qualifier to match a
105 FastCGI server defined as /server. This was done primarily for Java
106 Servlets, but is generically useful.
108 *) Change the comm between the PM and the request handlers from a regular
109 file to a pipe and an assortment of other dynamic fixes.
111 *) Log dynamic process termination scheduling and the resulting
112 process exit notification.
114 *) Change the default singleThreshhold from 10 to 0 to prevent the last dynamic
115 application process from being killed off due to low demand.
117 *) Clean up FastCGI application pathnames (e.g. remove duplicate slashes).
119 *) Fix bugs that prevented dynamic processes from being shutdown when
120 the load subsided. Lars Heete [heete@do.isst.fhg.de]
122 *) Prevent dynamic processes from being scheduled to be started before
123 the init-start-delay or restart-delay period expires. This prevents
124 the queuing of process start requests while an application which has
125 a long initialization period starts up.
127 *) When suexec is enabled, allow the process manager to become root again in
128 order to signal applications it spawned.
130 *) Change the "server started" log message level from INFO to WARN.
132 *) Fix dynamic server "Connection Refused" handling.
134 *) Fix demand-based dynamic process spawing (uses a more portable approach).
136 *) Add -idle-timeout arg to FastCgiServer, FastCgiExternalServer, and
137 FastCgiConfig. mod_fastcgi will now abort a connection if inactive for
138 longer than this period. It applies to the initiation of connections as
139 well (and thus is similar to appConnTimeout). Default is 30 seconds.
141 *) appConnTimeout is 0 by default now resulting in blocking connect()s. This
142 is more platform portable/predicable.
144 *) Leave STDOUT and STDERR open to the main server error_log. This should
145 eliminate a pile of protocol errors and having to track down 3rd party
146 libs writing to stderr inadvertantly. This doesn't mean these shouldn't
147 be fixed in the application to use the FastCGI I/O, it just allows these
148 applications to run when previously they'd crash and burn.
150 *) Add the pass-header arg to server directives.
152 *) Allow supplementary groups to be passed to spawned FastCGI applications (as
153 is done for CGI) - at least until PR2580 is resolved.
155 *) Initialize the default (empty) environment in a more socially acceptable
158 *) Miscellaneous doc improvements based on notes I've collected up over the
161 *) Fix a bug in the stderr handling introduced in 2.2.2. There were conditions
162 that resulted in not or improperly terminated strings.
164 *) Fix the call to setsockopt() to disable Nagle. [based on a bug report by
165 Johannes Plassmann <johannes.plassmann@pdb.siemens.de>]
167 Changes with mod_fastcgi 2.2.2
169 *) Added support for blocking connect()s by setting appConnTimeout to 0.
170 Non-blocking connect()s (the default) can be troublesome on some platforms.
171 Its expected that blocking connect()s will become the default (and
172 non-blocking connect()s will become optional) in the next release.
174 *) Dump a compile time error if the version of Apache is too old.
176 *) Wrap the SIGPIPE handler manipulation code such that it is only applicable
177 to Apache releases prior to 1.3.6.
179 *) Minor tweaks for RUSSIAN_APACHE. [Sergey Gershtein <sg@mplik.ru>]
183 Always restart a failed dynamic application if it is the last instance.
184 This means there once started there will always be at least one process
185 instance of a dynamic application.
187 Send PLEASE_START to the PM when a connect() results in an ECONNREFUSED.
188 ECONNREFUSED means the listen queue is full (or there isn't one). Asking
189 the PM to start (another) application instance may help empty it faster.
191 Change two sleep() calls to select() based snoozes because alarm() is in
192 effect and sleep() and alarm() don't always play nice together.
194 Fixed a couple of error messages.
196 *) Fix -listen-queue-depth arg on FastCgiConfig (dynamic). Previously it
197 was ignored and the default was always used.
199 *) Allow the -initial-env argument to be used to pass variables from the
200 Apache process environment to the FastCGI server (by specifying a
201 variable name without the "=" or a value). [suggested by
202 Martin Lichtin <lichtin@oanda.com>]
204 *) Cleanup some debug macros.
206 *) Improved script stderr handling. [based on suggestions by
207 David Birnbaum <davidb@chelsea.net>]
209 *) Added IRIX and FreeBSD to the list of supported platforms and other minor
210 updates to the INSTALL doc.
212 *) Changed the default listen-queue-depth (FCGI_DEFAULT_LISTEN_Q) from
213 5 to 100. Its still configurable with the -listen-queue-depth option.
214 This should help eliminate the FAQ - Why do I see "Connection Refused"
217 *) Fix a bug in FastCgiExternalServer that broke support for external servers
218 on other hosts. [Dave Neuer <dneuer@futuristics.net>]
220 Changes with mod_fastcgi 2.2.1
222 *) Updates to the INSTALL doc to describe building as a DSO.
224 *) If the FastCgiIpcDir directive was in use and "httpd -t" was issued while
225 Apache was running ("apachectl restart" does this), the contents of the
226 dynamic directory (dynamic sockets and the mbox) would be blown away
229 *) Add an extern declaration for ap_sys_siglist (its exposed but not in any
230 Apache header file) to prevent a compile error on systems without
233 *) During auth requests, the subprocess_env table was left holding variables
234 sent to the auth FastCGI server (including REMOTE_PASSWD!) which means they
235 were passed to other processing phases (such as CGI/FastCGI). The
236 subprocess_env is now restored to its pre-auth condition.
238 *) Added a FastCGI Authorizer Role compatibility mode which implements the
239 specification to the tee. Use -compat arg with any of the Auth server
240 directives. This is intended for new and existing FastCGI authorizer
241 applications that require compatibility with other server implementations.
243 *) Fix logging from the process manager. Previously, all calls to the
244 logging routines used NULL for the server_rec. This works fine in a dev
245 environment, but results in most of these messages being tossed by
246 log_error_core() because DEFAULT_LOGLEVEL is too low. Odd logic.
248 *) Always setup fr->header so that in the event the FastCGI server
249 doesn't get a valid FastCGI protocol header over the wire, we
250 can still print our error message to the log from do_work().
252 *) Fix a compile error in open_connection_to_fs() on systems with TCP_NODELAY
253 defined (Don Locrasto [locrasto@iceminer.com]).
255 Changes with mod_fastcgi 2.2.0
257 *) Serious rewrite of mod_fastcgi.html. An example conf will have
258 to wait until another time - I'm burned out on docs updates.
260 *) Add to the mod_fastcgi to the server version string.
262 *) SCO doesn't like the const in the arg to inet_addr() and
265 *) Use Apache's NET_SIZE_T to get the correct arg type for getsockopt().
267 *) Deleted redundant header files from fcgi.h.
269 *) Full path names are no longer required for directives which take
270 file names as arguments. Like other Apache directives you can
271 specify paths from server_root (don't start with a "/").
273 *) The fcgiKillMgr is gone. This intermediate process sat between
274 the Apache parent and the FastCGI Process Manager under Apache
275 1.2. The process manager tries to name itself fcgi_pm (it used
278 *) Authentication, Authorization, and Access phases are now
279 supported. I've bent the FastCgi spec a bit: as many of the standard
280 environment variables (that were easy) are sent (the spec says don't
281 send some by name), all headers (except Status) sent by the auth
282 FastCgi server are passed to subprocesses (CGI/FastCGI invocations)
283 as environment variables rather than just those prefixed by
284 "Variable-". Custom responses for auth failures aren't supported
285 (yet.. we'll see what the demand for this is like). In addition to
286 the FastCGI protocol defined environment variable "FCGI_ROLE" being
287 set to "AUTHORIZER" an environment variable "FCGI_APACHE_ROLE" is
288 set indicating which of the three phases is being processed. See
289 the docs for info on the new per-directory directives:
291 FastCgiAuthenticator, FastCgiAuthenticatorAuthoritative,
292 FastCgiAuthorizer, FastCgiAuthorizerAuthoritative,
293 FastCgiAccessChecker, FastCgiAccessCheckerAuthoritative
295 *) The code has been broken into logical chunks making figuring it
296 out much easier. All of the #defines a user may want to change are
297 now in mod_fastcgi.h.
299 *) All the logging has been converted to the "supported" Apache
300 logging routines, ap_log_error() and ap_log_rerror(). Log entries
301 for request specific errors should now be directed to the correct
302 server errorlog (if your running more than one). Some attempt has
303 been made at setting appropriate log levels (comments of course are
304 welcome) and printing errno info when its of value. All of log
305 messages now have "FastCGI" in them (nice for grep).
307 *) All of the calls to Apache routines have been updated to the ap_
308 convention, eliminating the need for the compat header.
310 *) By default we no longer flush every piece of data we get from the
311 FastCGI server to the client. This allows the FastCGI server to
312 release without having to wait for the client flush to complete.
313 The old behaviour can revived with the -flush argument to AppClass.
315 *) I renamed some of the directives for consistency (the old names
318 AppClass -> FastCgiServer,
319 FCGIConfig -> FastCgiConfig,
320 ExternalAppClass -> FastCgiExternalServer
322 *) Directive arguments are no longer case sensitive.
324 *) The module now uses hard_timeout() rather than soft_timeout. This
325 means that the module will longjump out of the request if the
326 drop-dead timeout expires (set with Apache's Timeout directive, the
327 default is 5min) or if the client closes the connection (SIGPIPE).
328 This is more typical for a Apache modules and I'm not convinced we
329 properly handle all of the error cases well enough to use
330 soft_timeout (I can think of one place we don't anyway). This means
331 your FastCGI application can see SIGPIPE.
333 *) SIGPIPE is now ignored by default in FastCGI servers spawned by
334 Apache. Without this the default behaviour is exit().
336 *) It should (I think) run now under DSO now.
338 *) The bug which prevented sending of binary data in 2.1b1 is fixed.
340 *) The broken -initial-env bug ins 2.1b1 is fixed.
342 *) Maybe some other stuff.
346 *** Well there's a big gap here. Maybe I'll go back and extract the CVS
347 commit notes and stick 'em in here.. when I get some free time. ;)
352 *** Originally from docs/README..
354 Apache[X1.03.01]/mod_fastcgi[02.00.05] 19970909 unsupported
355 From: Stanley Gambarin <stanleyg@cs.bu.edu>
357 *) Yet more changes to the source distribution. Separated out
358 the Tcl dynamic string and buffer libraries into separate files.
359 Did the same for the OS library, however, it is not really an
360 abstraction... more like a bunch of wrappers. Added to DEVNOTES
361 a note about the problem of Apache not allowing including header
364 Apache[X1.03.01]/mod_fastcgi[02.00.04] 19970908 unsupported
365 From: Stanley Gambarin <stanleyg@cs.bu.edu>
367 *) More changes to the installation scripts. Theoretically we
368 should be able to to use the same Makefile/installation script
369 for both Apache 1.2.x and 1.3.x. Need some sort of the abstraction
370 layer to use the same source code for both 1.2.x and 1.3.x sources.
372 Apache[X1.03.01]/mod_fastcgi[02.00.03] 19970905 unsupported
373 From: Stanley Gambarin <stanleyg@cs.bu.edu>
375 *) Created Makefile to compile the module into the archive library,
376 just like the proxy module. Also initial draft of the installation
377 script, which should work for both 1.2 and 1.3 sources of Apache.
379 Apache[01.02.04]/mod_fastcgi[02.00.02] 19970903 unsupported
380 From: Stanley Gambarin <stanleyg@cs.bu.edu>
382 *) More source reorg. Separated out some header files. Need much
383 more work on this. Updated the TODO file. Added fcgivers.h and
384 CHANGES file to track the history.
386 Apache[01.02.04]/mod_fastcgi[02.00.01] 19970902 unsupported
387 From: Stanley Gambarin <stanleyg@cs.bu.edu>
389 *) Source reorganization. This is done to provide a basis for later
390 functionality implementation. See TODO file for more information
391 of what possible future enhancements are possible/desired.
393 Apache[01.02.04]/mod_fastcgi[02.00.00] 19970902 unsupported
394 From: David MacKenzie <djm@va.pubnix.com>
396 *) Create the "dynamic" dir and "mbox" as the user specified in the
397 User and Group directives instead of as root. They are created with
398 restrictive permissions, and then the module checks to see whether the
399 user specified in the User and Group directives has read, write, and
400 execute permissions on them. Those permissions had been explicitly
401 denied by creating them as "root".
403 *) Some documentation was garbled or incomplete.
405 *) When a FastCGI application can't be execl'd, the code used a
406 value of errno that may have been stomped on by intervening system
407 calls. That happened on BSD/OS 2.1, where the error_log was
408 reporting errno=25 (NOTTY) instead of the correct errno=13 (EACCES).
409 An intervening stdio call was setting errno as a side-effect, which
410 it has the right to do.
412 *) Erroneous fprintf arguments, a missing return value, an unused
413 function, missing declarations, and other problems detected by gcc
416 *) Duplicated code to create the lock file name merged into a single
417 function, closing several memory leaks.
419 *) FCGIConfig -minProcesses didn't allow a value of 0, so the last
420 FCGI app in a given dynamic class would live forever, even if it
421 hadn't been requested for weeks.
423 *) The test for whether to keep looking for dynamic app victims to
424 kill was backwards, so no dynamic apps were ever selected as
427 *) The killInterval and updateInterval were being ignored; the
428 sigsuspend() forced recalculations only when a child died,
429 instead of at the intervals specified.
431 *) When a dynamic app couldn't be started, the program SEGV'd when
432 trying to free the ipcAddrPtr twice.
434 *) Remove the dead lock file and socket when cleaning up after a
435 server whose last child has been killed, so the FastCGIHandler
436 isn't fooled into thinking there is still a process serving
439 *) Make file locking robust in the presence of signals.
441 *) Rename some badly named variables.
443 *) Fix typos in many comments.
445 *) Fix some memory and file descriptor leaks.
447 *) Make the blocking kill of a server closer to working.
449 *) More fixes to calculations in dynamic application management.
452 *** Originally from docs/README.OMI..
454 What's New: Version 2.0b1, 16 Apr 1997
456 *) Implemented a mechanism by which FastCGI applications are started by
457 the web server on the first request and continue running. Also, a
458 heuristics have been implemented to allow for dynamic killing of the
459 running FastCGI apps. The configuration options are supplied via
460 the new FCGIConfig directive, that is described in the
463 *) When performing internal redirect, since the original request's body
464 has already been read, do not allow the redirected request to think
465 that is has one [body].
467 *) More conditional compilation for OS/2.
469 *) Fixed occassional "Assertion failed: len > 0, file mod_fastcgi.c"
471 *) Fixed "failed assertion `count >= 0 && count <= bufPtr->length'"
473 *) Added bflush() call after bwrite() in DrainReqOutBuf function to
474 force some output to be written in the period of inactivity. This
475 circumvents the Apache's buffering during the server-push.
477 *) Added a header-parser function placeholder for compatibility with an
480 *) Additional checks have been placed with regard to permissions of the
481 FastCGI processes. These include the existence of the file,
482 ExecCGI and IncludesNOEXEC checks, disallowing of nph- scripts, etc.
484 *) Implemented a restart cleanup, so that no parentlesses FastCGI
485 applications are left after Apache is restarted and/or terminated.
486 This was accomplished via two independent mechanisms, where the
487 first one prevents a start of the process manager on the first
488 reading of the configuration files. The second mechanism deals away
489 from Apache's implementation of fork() (via spawn_child()) and
490 implements its own forking, thus removing any dependencies on the
491 Apache to cleanup processes during termination phase.
493 *) Removed the definition of the Sigfunc, as Apache 2.0b10 defines it
494 itself in the file conf.h
496 What's New: Version 1.4.3, 15 Jan 1997
498 *) Fixed compilation warnings for various platforms, as well as
499 conditional compilation for OS/2.
501 What's New: Version 1.4.2, 12 Dec 1996
503 *) mod_fastcgi.c is ported to Apache 1.2b. Any further development
504 will proceed under this version of Apache web server.
506 *) As the result of porting, the "include virtual" construct of SSI
507 will now work correctly using either <Location> or <Directory>
510 What's New: Version 1.4.1, 4 Dec 1996
512 *) Checks have been removed from the ScanCGIHeaders that provided
513 for the presence of both Status and Location headers as being
514 an error. Contradictory to CGI/1.1 Internet Draft, both of
515 these headers are used by the current CGI applications.
517 What's New: Version 1.4, 22 Nov 1996
519 *) Added the -port option to AppClass, allowing TCP/IP communication.
520 Added the -socket option to AppClass, allowing Unix domain
521 communication via a configurable pathname.
523 *) Added the ExternalAppClass directive, allowing TCP/IP
524 communication with remote FastCGI applications.
526 *) The handler had its own code for generating HTTP response headers;
527 now it uses Apache's. This reduces the size of the module.
528 More importantly, it fixes the bug in which "include virtual"
529 sees the HTTP response headers.
531 *) The response header parser performed very little checking.
532 Now the parser enforces the guidelines in the CGI/1.1
533 Internet-Draft: Status and Location are mutually exclusive,
534 Location can only be a response to GET or HEAD,
535 CGI response headers can't be repeated, etc. (The CGI response
536 headers are the ones the handler interprets: Status, Location,
539 *) The response header parser used to miss CGI headers
540 with no whitespace after the colon, e.g. "Status:200 OK".
542 *) The response header parser sometimes interpreted the first line
543 of content as an RFC822 continuation line.
545 *) The handler implemented a nonstandard version of Location
546 which never used internal redirects. The handler
547 now attempts to implement Location as specified in the
548 CGI/1.1 Internet-Draft. The module documentation explains
551 *) Error log entries from the response header parser were
552 pretty uninformative; they are better now. If a header is
553 malformed, the log entry includes it. If the headers are
554 unterminated rather than malformed, the log entry says that,
555 and says how many bytes were received from the app.
557 *) When the application manager forked a new process, and that
558 process ran into trouble before executing the first instruction
559 of application code, the process used to exit with status =
560 errno. This made certain configuration problems (e.g. incorrect
561 file permissions when server parent is root) quite difficult
562 to pin down. Now the failing child process opens up the error log
563 and writes an informative entry before exiting.
565 *) The module now correctly handles a slash at the end of the
566 DocumentRoot directive. This was a one-line fix.
568 What's New: Version 1.3.3, 17 Oct 1996
570 *) The module now registers its request handler under the name
571 fastcgi-script in addition to the name application/x-httpd-fcgi.
572 This was a one-line addition to the module, but had quite
573 a large impact on the documentation and sample configuration.
575 What's New: Version 1.3.2, 27 Sept 1996
577 *) On some systems (SunOS, Linux?), fopen for append has a bug
578 that strikes when two processes append to the same file.
579 This bug causes the process manager to corrupt the error log.
580 Work around the bug by calling open, then fdopen. A patch
581 to Apache 1.1.1 is also required, as described in mod_fastcgi.html.
582 (Reported by Bob Ramstad.)
584 What's New: Version 1.3.1, 17 Sept 1996
586 *) On Linux and SunOS, the Apache default user_id (-1) is not a
587 legal value of uid_t, so cast it. (Reported by Bob Ramstad,
588 Scott Langley, others.)
590 *) On some systems (Linux, some Solaris?, Irix?), connect requires
591 write access to a Unix Domain socket, so provide it. (Reported
592 by Scott Langley, freeform@wired.com, others.)
594 *) If you hit a .fcg file, but there's no AppClass defined, the
595 error should be NOT_FOUND instead of SERVER_ERROR. (Reported
598 What's New: Version 1.3, 4 Sept 1996
600 *) Module sometimes busy-waited in FastCgiDoWork. Fixed.
602 *) Module violated the Apache buff abstraction.
603 Fixed. Might now work with SSL (not tested).
604 As part of the fix, eliminated support for
605 Apache 1.0x versions.
607 *) Module failed to chmod the Unix domain listening sockets
608 it created, so protections were set according to the current
611 *) Module forked too many process manager processes: One
612 per AppClass. When the Apache parent ran as root, these
613 process manager processes ran as root. New process
614 manager is a single process (only started if AppClass is
615 used) and runs with same privileges as other children.
616 The new process manager doesn't do any polling, so
617 there's less system overhead than before.
619 *) AppClass insisted on getting at least two arguments. Fixed.
621 *) Module used setjmp/longjmp, causing compiler warnings
622 on some platforms (e.g. Linux.) The module no longer
625 *) Module created listening sockets in /tmp, where they were
626 sometimes wiped out by cleanup scripts. Added FastCgiIpcDir
627 directive to give control over the location of listening
630 *) Module wrote error log entries without a timestamp. Fixed.
632 *) AppClass directive wrote error messages to stderr in addition
633 to returning a char * message to the Apache core. Fixed.
635 What's New: Version 1.2, 3 June 1996
637 *) Ported from Apache-1.0-based code to Apache-1.1b2 internals by
640 Add version string: APACHEVERSION macro to mod_fastcgi.c
642 *) chown FastCGI socket to user_id and setuid to user_id for app
645 *) Modify GetFromStream() by having it call an OS dependent
646 function GetStreamSize(FILE *) which uses FILE internal data member.
647 Linux users might need to modify GetStreamSize().
649 What's New: Version 1.1, 10 May 1996
651 *) If you specify a non-existent executable in the AppClass directive,
652 or if the file exists and it does not have execute permission you
653 get a constant stream of error messages telling you that "program
654 terminated due to a signal".
656 *) The mod_fastcgi module should use the standard Apache error logging
657 facility instead of writing to stderr.
659 Version 1.0, 30 April 1996