1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
|
Fri Mar 16 16:17:24 2001 Carlos O'Ryan <coryan@uci.edu>
* java/EAC/User Manual.doc:
* java/EAC/User_Manual.doc:
Renamed documentation file, many a script gets confused with
names that have spaces in them.
Sun Jun 4 14:53:05 2000 Darrell Brunsch <brunsch@uci.edu>
* ImageProcessing/framework/MedJava.html:
* ImageProcessing/framework/test.html:
* apps/NexusII/start.html:
Added in missing CVS Id strings.
Tue Aug 24 13:09:15 1999 Everett Anderson <eea1@cs.wustl.edu>
* src/*, tests/*, netsvcs/*: Implemented major changes to the
package system. Now all files descend from the JACE directory
under the main archive level. All files are further grouped by
their subpackage names. This is the way Java usually wants it,
and makes JACE javadoc friendly.
Rewrote and added javadoc comments throughout JACE.
Changed all concurrency mechanisms to inherit from AbstractLock,
obeying all its requirements such as not releasing the lock
unless the accessing thread is the owner.
Added the token service.
Added and rewrote tests for Concurrency mechs and ServiceConfig.
Rewrote network services to inherit from JACE.netsvcs.Server and
JACE.netsvcs.Handler, abstracting much of the work out to these
base classes.
Debugging is now OFF by default.
Wed Aug 18 19:43:23 1999 Everett Anderson <eea1@cs.wustl.edu>
* src/GetOpt.java: Added another constructor to GetOpt with an
option to return args even if not in the optstring. This is
useful for the default case in a switch (for printUsage
and exit).
Wed Aug 18 13:35:46 1999 Everett Anderson <eea1@cs.wustl.edu>
* src/TimerQueue.java: Now when createInternalThread is specified
in the constructor, it is set to be a daemon thread. Thus, if a
TimerQueue thread is the only one executing, the VM exits.
Wed Jul 14 11:27:47 1999 Everett Anderson <eea1@cs.wustl.edu>
* src/StrategyAcceptor.java, AcceptStrategy.java
StrategyAcceptor now inherits from Acceptor. AcceptStrategy
can now delegate to an Acceptor, rather than having its own
SOCKAcceptor. Its default is to delegate to a generic
Acceptor instance (that uses a SOCKAcceptor).
Fri Jul 09 13:58:42 1999 Everett Anderson <eea1@cs.wustl.edu>
* src/Acceptor.java, TimeValue.java
Solved a few javadoc warnings.
Fri Jul 09 11:08:32 1999 Everett Anderson <eea1@cs.wustl.edu>
* src/TimedWait.java, Stream*.java, Token.java
Fixed a bug in TimedWait from the recent changes. Made
Streams (and Modules) use absolute times for timeouts.
Thu Jul 08 14:54:36 1999 Everett Anderson <eea1@cs.wustl.edu>
* src/Condition.java, TimedWait.java, EventHandler.java,
MessageQueue.java, ServiceObject.java, SvcHandler.java,
Task.java, Mutex.java, Semaphore.java, Token.java,
TimerQueue.java
Changed the semantics of Condition, TimedWait, MessageQueue,
Task, Mutex, Semaphore, and Token to use absolute times for
their timeouts.
Changed the semantics of EventHandler, ServiceObject, and
SvcHandler such that handleTimeout receives a TimeValue
representing when the event occured.
Changed TimerQueue internally to reflect the above changes.
Also made better checks in Token and Mutex so that non-owners
can call release without adverse effects. I plan to try to
do this with Semaphore and RWMutex as well.
Thu Jul 08 11:20:17 1999 Everett Anderson <eea1@cs.wustl.edu>
* src/TimeValue.java: Added relativeTimeOfDay(offsets)
methods which return TimeValues. These should help people
since many classes will be changed to use
absolute TimeValues shortly to correspond to ACE's versions.
Wed Jul 07 22:42:27 1999 Everett Anderson <eea1@cs.wustl.edu>
* src/Token.java: Fixed several bugs in tryacquire and renew.
Specifying a null TimeValue in acquire or renew now means to wait
until the lock is obtained. Specifying TimeValue.zero in acquire
means just to poll and don't call sleepHook(). In renew, it is
the same as specifying a requePosition of 0 -- doesn't give up
the token.
Wed Jul 07 21:54:35 1999 Everett Anderson <eea1@cs.wustl.edu>
* src/TimedWait.java: Fixed the semantics of
timedWait(TimeValue tv). Now when tv represents zero, it has
polling semantics, checking the condition once and returning
on success or throwing a TimeoutException on failure. Before,
it blocked until notified. (Also affects synch mechanisms)
Wed May 20 14:37:48 CDT 1998 Everett Anderson <eea1@cs.wustl.edu>
Reset the ChangeLog file for future users. If anyone wants more complete
file by file changes, they can be provided with the CVS log command.
Updated the JACE page at http://www.cs.wustl.edu/~eea1/JACE.html to reflect
the Version 1.4.4 changes, which included:
* Removed CUP and JLex from JACE. The java_cup directory is still in CVS,
but won't be included in future versions. This means that the option to
load C++ ACE ServiceConfigurator files is gone.
* Changed the SOCKStream system so it uses DataInputStream and DataOutputStream
to maintain compatibility with other languages and remove the deprecation
warnings from using PrintStream. (Can't use Reader/Writers since they
write 16-bit chars not 8-bit bytes)
* Made lots of modifications to comments and code to remove javadoc warnings
and deprecation warnings. They were all straight-forward changes such as
making sure a @throws line was in the comments for all exceptions a method
had, etc.
|
