d21f901097
Tested partially: handle allocation, connection, prepare, execute
...
Missing: Support for SQLBindCol & several catalog functions. The scsh 0.53 API abstraction need
s to be implemented on top of the ODBC 3 API.
2002-04-21 08:56:02 +00:00
570b5c11d2
Basic support for ODBC 3.0 API.
...
Tested partially: handle allocation, connection, prepare, execute
Missing: Support for SQLBindCol & several catalog functions. The scsh 0.53 API abstraction needs to be implemented on top of the ODBC 3 API.
2002-04-21 08:53:57 +00:00
0a8e312756
Compare symbols non-hygienic as well to support SREs within hygienic
...
macros.
2002-04-18 16:06:02 +00:00
3717053d60
Preserve thread fluids when forking.
2002-04-12 14:29:38 +00:00
349fff06c1
Basic FORK and and %FORK now start a new command level inside the
...
child process, thus preventing other threads from continuing to run.
Both take an additional optional argument that, when true, reverts the
old behavior of continuing all threads.
Moreover, use WITH-CONTINUATION NULL-CONTINUATION in CALL-TERMINALLY
to get us more space.
2002-04-10 15:03:02 +00:00
b222cfe4c3
Default argument for tty-info as promised in the manual.
2002-03-22 10:23:34 +00:00
c9c0535f41
Bug fix: 2 values were not converted to descriptors.
2002-03-22 10:22:59 +00:00
7fdde2527f
Plug space leak in BIND-LISTEN-ACCEPT-LOOK: Errno handlers for failed
...
connections would accumulate indefinitely.
2002-03-06 08:24:32 +00:00
cb6e8572d6
Added forgotten calls to with-cwd-aligned.
2002-03-05 16:57:43 +00:00
87d82f5011
Align cwd and umask.
2002-03-05 16:49:58 +00:00
a7517a3f05
Add record discloser for FDPORT-DATA.
2002-02-26 14:40:59 +00:00
5f7ed5befe
Fixed bug in test for sync-file.
2002-02-26 13:02:17 +00:00
13dabb8f4e
Always signal an error if time returns -1. Not uses errno-error as
...
only few systems set errno if mktime fails.
2002-02-26 10:53:32 +00:00
45305e7555
Set-cloexec to #t for unrevealed ports.
2002-02-26 08:42:56 +00:00
380fee6612
Make FLUSH-ALL-PORTS blocking so FORK will do something more sensible.
...
Previously,
(begin
(display "ha!") (newline)
(fork (lambda ()
'foo)))
would print "ha!" twice because FLUSH-ALL-PORTS would not finish
before the actual FORK.
2002-02-25 20:29:15 +00:00
09ab472695
+ Greet with scsh-version-string
...
+ call relinquish-timeslice twice before exit so the flushers may run
2002-02-25 09:00:39 +00:00
71cc8cf091
Fix for 0.6.1
2002-02-25 07:01:22 +00:00
de0de38bb6
Remove obsolete comment about START argument to CRE-SEARCH not being
...
supported.
2002-02-22 10:27:07 +00:00
68aa82c4e7
Remove superfluous END-VEC parameter to CRE-SEARCH.
2002-02-22 10:25:42 +00:00
dce89b2fa8
Translate submatches according to CRE:TVEC, fixing submatch bug
...
reported by Michel Schinz.
2002-02-22 10:22:38 +00:00
64cdef6fc7
Fix infinite recursion bug in HACK-BRACKET-SPEC:
...
The LOOSE and RANGES lists weren't getting sorted in the fixpoint
comparison.
2002-02-21 13:44:47 +00:00
be84d5ec47
Reinstate FILE-MATCH, accidentally left unported from 0.5.1.
...
Replace mention of SRFI 13 (not there yet) by STRING-LIB (gone soon).
2002-02-21 13:22:27 +00:00
b289d0dde7
Fix incomplete port to new regexp system.
2002-02-21 13:21:32 +00:00
4efe1a4cd8
+ Use S48's nonblocking facility to flush the ports
...
+ Deleted bogus select implementation
2002-02-21 06:55:37 +00:00
be129cadc7
Sometimes some OSs (OS X, Linux, Solaris) report EISDIR when unlinking a file.
2002-02-19 17:18:45 +00:00
d77257ae6b
More pervasive fix for ASCII NUL problem.
2002-02-16 18:04:49 +00:00
1f5b157c1d
Omit ASCII NUL from control character set.
2002-02-16 17:46:09 +00:00
a03bc65f4e
Remove BOW/EOW and WORD/WORD+ as they are not in POSIX, not supported
...
by a number of platforms, and their meaning is locale-dependent.
2002-02-16 16:57:25 +00:00
1b2096cfb4
Cancel some unnecessary GC_PROTECTs.
2002-02-14 14:00:27 +00:00
64221a6b95
Don't blindly remove a process object.
2002-02-14 13:32:19 +00:00
238ba3e8ba
Let weak-table-ref kill the entry if the value is #f.
2002-02-14 13:31:41 +00:00
36c5c82ec3
Added forgotten with-error-output-port stuff.
2002-02-14 13:30:53 +00:00
8d733eec78
env-path-list is a fluid.
2002-02-13 15:09:27 +00:00
54efae2318
+ Lock for the process table
...
+ Use wait-lock to glue waiting and the process object together
+ Delete the pid/weak-pointer pair in the process table if the process object
gets finalized
2002-02-13 14:56:11 +00:00
2b7266994d
Treat setting to #f as delete.
2002-02-13 14:36:10 +00:00
9e2d0ca60c
Minor corrections.
2002-02-13 14:34:52 +00:00
4d28e2bb33
/kick mainzelM
2002-02-13 08:51:01 +00:00
3662ea8d88
Fixed bug in bind-socket.
2002-02-13 08:28:59 +00:00
94d5ae9f71
GC_PROTECT'ed the necessary variables (specifically, where >1 arg to
...
a function 'may GC')
2002-02-12 15:49:39 +00:00
0025ee598b
Fixed typo.
2002-02-12 15:31:36 +00:00
90445340fe
GC_PROTECT'ed the necessary variables (specifically, where >1 arg to
...
a function 'may GC')
2002-02-12 13:43:37 +00:00
424109a452
GC_PROTECT'ed the necessary variables (specifically, where >1 arg to
...
a function 'may GC')
2002-02-12 13:38:20 +00:00
96b6558a78
GC_PROTECT'ed the necessary variables (specifically, where >1 arg to
...
a function 'may GC')
2002-02-12 13:37:36 +00:00
e316ce1109
+ Fixed bug in set-process-group
...
+ include <string.h>
+ save errno before starting timer interrupts
2002-02-11 17:32:20 +00:00
65407be26c
GC_PROTECT'ed the necessary variables (specifically, where >1 arg to
...
a function 'may GC')
2002-02-11 15:38:35 +00:00
6158e7a964
Fixed memory leak.
2002-02-07 10:11:20 +00:00
3ce5de8758
If $HOME is unset, consult (user-info (user-uid)) for the value of home-directory.
2002-02-06 15:17:19 +00:00
7acad294ac
+ Don't raise an error on ECHILD since ECHILD just indicates that
...
there are no childs ready.
+ Squirrel away errno before starting alarm interrupts
2002-02-06 09:47:27 +00:00
07f796459e
Align the resource before changing it in with-resource*.
2002-01-29 08:12:09 +00:00
369bc08f2e
GC protection fix for waitpid.
...
Note that there are probably more bugs similar to this one.
2002-01-27 20:20:09 +00:00
eknauel