4e3129d521
Also reworked the AppClient class (to use AppTransport)
2014-03-28 00:32:53 +01:00
a817f25943
Added missing include
2014-03-27 23:51:31 +01:00
72fc0b581b
Fixed detecting failed tests
2014-03-27 00:28:49 +01:00
56449b4ce0
Not every test has a protocol associated anymore
2014-03-27 00:22:58 +01:00
0840dbebd8
Nicer output
2014-03-24 06:31:23 +09:00
6d7fe7045a
The "appserver" test is currently expected to fail
2014-03-24 01:22:41 +09:00
becab80217
Reviewed the AppServer code
2014-03-23 15:32:02 +09:00
2ee7645798
Renamed some functions and callbacks for consistency
2014-03-23 14:36:54 +09:00
515825fa53
Also run tests against actual names
2014-03-23 14:14:21 +09:00
ba6fe20346
Unifying the tests
2014-03-23 14:11:49 +09:00
ed27d70cce
There may be more than one message in incoming TCP data
2014-03-23 12:58:39 +09:00
74216ee392
Code cleanup
2014-03-23 12:53:55 +09:00
1ee2e1e8b9
Added code to send messages back to clients (untested)
2014-03-23 01:58:07 +09:00
82ead01089
Attempt to spare transport plug-ins from handling acknowledgements
2014-03-23 01:47:32 +09:00
1cd74f6859
Beginning to send acknowledgements
2014-03-23 00:39:09 +09:00
c2dcd53cce
Creating message acknowledgements is only relevant internally
2014-03-23 00:36:41 +09:00
32593c6e25
Preparing support for acknowledgements
2014-03-22 23:40:28 +09:00
362d266383
Link the transport plug-ins with libApp explicitly (as required for MacOS X)
2014-03-22 23:32:17 +09:00
77e759f3ae
Fixed link-time flags for libSystem
2014-03-22 23:13:27 +09:00
b8346239d6
Alias callbacks as calls
2014-03-22 22:19:32 +09:00
8ed2b6baf5
Also check the message's type
2014-03-22 22:19:16 +09:00
7e0d8adfd5
Code cleanup
2014-03-22 22:08:16 +09:00
d067afbfdb
Fixed a crash when the address is invalid
2014-03-22 22:07:25 +09:00
97269fc9d4
Added missing #include <string.h>
2014-03-22 21:57:41 +09:00
0dadf2321f
The UDP tests are now expected to complete
2014-03-22 19:07:30 +09:00
2194e8b36c
Fixed memory corruption
2014-03-22 19:03:31 +09:00
727a33b6eb
Getting closer to functional UDP calls
2014-03-22 19:00:53 +09:00
3017f213fe
Checking if the method name is preserved after serialization (for calls)
2014-03-22 18:48:18 +09:00
5c4a33b3ed
Added some debugging information
2014-03-22 18:48:03 +09:00
2c51e00280
More complete test code
2014-03-22 18:39:44 +09:00
30ffe2ea3e
Added a basic test for AppMessages
2014-03-22 18:35:56 +09:00
eb198d1adf
Also free the AppMessage
2014-03-22 18:21:41 +09:00
cad944b9ca
Also delete any AppTransportClient upon teardown
2014-03-22 17:20:25 +09:00
e29f6370cc
Also implement the helper deleting clients
2014-03-22 17:19:27 +09:00
7721bb1b30
More complete cleanup (for servers in particular)
2014-03-22 16:39:30 +09:00
9b265d7c29
Check that messages are sent from clients
2014-03-22 15:58:31 +09:00
30d9d25cd2
Let UDP servers accept incoming messages
2014-03-22 15:51:00 +09:00
4351745c06
More portable shebang, from 7heo <7heo@mail.com>
...
Original commit message:
« The path /bin/sh for the shebang is more portable
than /usr/bin/env, according to Sven Mascheck:
http://in-ulm.de/~mascheck/various/shebang/#env
and:
http://in-ulm.de/~mascheck/various/shells/
Additionally, a space after the shebang first two
characters (#!) is said to be more portable:
http://in-ulm.de/~mascheck/various/shebang/#blankrequired
I fixed the test script shebang accordingly. »
It doesn't say the presence of a space makes it more portable; it says it was a rumor that could never be proven true, for some completely obsolete releases of UNIX. I did not apply that part.
Thanks!
2014-03-13 13:42:43 +09:00
d77c87cc58
Newer version of the script upstream
2014-03-13 13:42:06 +09:00
0fd8da5e6b
Newer versions of the helper scripts upstream
2014-01-23 12:42:25 -06:00
88a40ac5bb
Code cleanup
2013-10-25 22:58:23 +02:00
f392915040
Slightly improved the tests
2013-10-25 22:24:38 +02:00
745654a94e
Fixed the URL for the online documentation
2013-07-24 03:38:13 +02:00
d0246c83ee
Re-generated the Makefiles
2013-07-17 03:34:52 +02:00
d9846e9f6b
Using AF_* constants instead of PF_* (for POSIX compliance)
2013-07-17 03:34:12 +02:00
0246b35efb
Using AF_* constants instead of PF_* (for POSIX compliance)
2013-07-17 03:33:07 +02:00
2fd22109e4
Workaround for a build issue
2013-07-17 03:02:37 +02:00
8067a45422
Re-generated the Makefiles
2013-03-24 17:14:02 +01:00
4cdbf793b9
Re-generated the Makefiles
2013-03-17 00:21:48 +01:00
e244567ef9
Check that the service name is defined
2013-01-11 02:21:29 +01:00
