a924387ffb
More statly bits.
Michael Bayne
2004-08-04 02:36:56 +00:00
6b0695af30
Track and report some connection manager stats over a short historical period so that we can see what sort of funny business is going on with the network thread when the process spikes up to 100% CPU.
Michael Bayne
2004-08-04 02:31:18 +00:00
ec022ed1b2
Ask the SceneProvider to handle a moveBody request if the body is already in the correct zone.
Ray Greenwell
2004-07-30 21:13:41 +00:00
2efe7cc777
Moved getPerformanceLevel to yohoho where it belongs.
Ray Greenwell
2004-07-29 21:51:38 +00:00
51ce5f3b4f
Qualify them error messages.
Ray Greenwell
2004-07-17 00:24:31 +00:00
db04c873b2
If our bundle is already ready to go, we may get the notification that it's ready prior to even waiting for it. Handle that case.
Ray Greenwell
2004-07-14 20:31:33 +00:00
0e853085d5
OK boys, we're going in. Revamped the resource manager now that it doesn't handle downloading of resources, it simply manages them in situ. Getdown is now responsible for all the downloading.
Michael Bayne
2004-07-14 14:06:46 +00:00
c7223a89e1
New commons world order.
Michael Bayne
2004-07-13 16:37:40 +00:00
1da753e14d
Brought things up to date with the latest Jakarta Commons world order.
Michael Bayne
2004-07-13 16:34:49 +00:00
a6f57205d7
Make sure we stop the drop sprite from dropping at high velocity when we suspend the puzzle. Also added some logging to debug the occasional obscure situation where the puzzle is suspended at just the right time to prevent the drop block from properly starting up again when we resume.
Michael Bayne
2004-07-10 05:18:59 +00:00
fe7e9dc413
Debugging code to try to track down the cause of the multiply released pending ratings. Somehow a game is getting ended twice even though the code should prevent it.
Michael Bayne
2004-07-10 01:32:54 +00:00
55df96a644
Throw an informative exception in this circumstance. The NPE that would follow doesn't tell us anything about the hosed tileset.
Michael Bayne
2004-07-09 04:00:20 +00:00
19abac5be1
Handle shorts. Maybe later we'll add support for pants.
Michael Bayne
2004-07-09 03:52:04 +00:00
f700be00db
Proving the value of the instant-review system, we repair this booch before it has a chance to bring everything to a screeching halt.
Michael Bayne
2004-07-07 04:08:41 +00:00
212ebb1b53
As Ray pointed out, clearing the oid upon destruction causes problems, so let's just not do it. We can accomplish our earlier goals, though with slightly less distinction making capability, in a different way.
Michael Bayne
2004-07-07 03:10:24 +00:00
761e05c540
Try to only log when something actually bad is going on.
Michael Bayne
2004-07-03 09:30:03 +00:00
f8e5c85d1f
Keep mum about connections being reset by their peer. It is a normal event in the wild, wild world of the Interweb.
Michael Bayne
2004-07-03 09:19:04 +00:00
124abb66f7
This also happens as a matter of course. Clients send invocation requests after things have already been processed and cleaned up on the server.
Michael Bayne
2004-07-03 07:24:39 +00:00
e57580626e
Clients come and go at the craziest of times. No need to worry about it.
Michael Bayne
2004-07-03 07:19:46 +00:00
90e5231bae
When a DObject is destroyed, set its oid to -1 so that it knows that it was once alive but is now gone. Catch attempts to start a transaction on a destroyed object and log them as such. Made isDestroyed() and isActive() final for wholly unfounded performance reasons.
Michael Bayne
2004-07-03 07:16:56 +00:00
0be671cd09
We aught to be usin' this here too.
Michael Bayne
2004-07-03 03:41:21 +00:00
e021264d52
If we encounter an error with a bundle while initializing the bundled component repository, instruct the bundle to delete itself so that we will redownload it next time the application is initialized.
Michael Bayne
2004-07-03 03:25:48 +00:00
9160490ffb
Made wipeBundle() accessible to external callers, moved a protected method into its happy place.
Michael Bayne
2004-07-03 03:21:05 +00:00
e1d5b096ed
Invoker refactoring.
Michael Bayne
2004-06-29 03:22:22 +00:00
434a7cafb0
We need to set the difficulty before we set the board seed because that triggers the generation of the boards and the difficulty needs to have been broadcast prior to that.
Michael Bayne
2004-06-22 15:35:10 +00:00
a4cdf5382e
Removed vestigal constant.
Michael Bayne
2004-06-22 14:26:10 +00:00
2f0b9939d7
Some cleaning up of the difficulty handling so that we can use it as the basis for our star system.
Michael Bayne
2004-06-22 14:08:58 +00:00
031331a5c3
Regenerated from service definitions.
Michael Bayne
2004-06-22 13:57:44 +00:00
425e8a6dcb
If we fail to resolve our jar file, it is most likely corrupt, so delete it and allow the resource manager to redownload.
Michael Bayne
2004-06-19 08:09:33 +00:00
8f921179b1
Let there be one ant script.
Michael Bayne
2004-06-19 06:52:40 +00:00
5b68088b7c
Download the properties file ourselves before handing it to Java to parse into a properties object so that we can log the text if Java chokes.
Michael Bayne
2004-06-18 17:41:39 +00:00
03ca013e8d
Allow system broadcasts to be INFO or ATTENTION level.
Ray Greenwell
2004-06-17 18:28:58 +00:00
5282160ab4
Need to set the content length in here as well.
Michael Bayne
2004-06-17 03:02:55 +00:00
fc8213493e
Alas, such a simple fix isn't going to do the job. We need to be smarter about handling resources that didn't provide a content length.
Michael Bayne
2004-06-16 09:44:23 +00:00
53eca560f6
Cope with the current situation where PA's servers somehow strip the content-length header even though the servlet container is providing one.
Michael Bayne
2004-06-16 09:25:52 +00:00
407d4ac3b1
Added a static broadcast().
Ray Greenwell
2004-06-15 22:03:51 +00:00
14c9e6bc2c
Fail more gracefully in this circumstance, not that there's much we can do.
Michael Bayne
2004-06-15 18:25:13 +00:00
136da4e780
Gracefully deal with DSet modifications, rather than brutishly recreating the entire table.
Ray Greenwell
2004-06-15 17:58:38 +00:00
1e5ff6dd92
Broke out textual parsing into another method to ease subclassing.
Ray Greenwell
2004-06-15 03:25:19 +00:00
87bbff083f
Guts exported to ObjectEditorTable.
Ray Greenwell
2004-06-14 01:09:23 +00:00
0af79e2719
Invoker unit namestravaganza.
Michael Bayne
2004-06-13 08:51:57 +00:00
86e2a25580
Allow invoker units to be easily named.
Michael Bayne
2004-06-13 08:03:45 +00:00
e5b1d3eda2
Go up to days.
Ray Greenwell
2004-06-04 22:14:11 +00:00
d25c7adba6
added getSelectedEntry.
Ray Greenwell
2004-06-03 19:11:18 +00:00
c8f519fc85
Rolled SettedObject right into DObject, because it's just too useful.
Ray Greenwell
2004-06-03 18:15:03 +00:00
2d90dd40a1
A special field editor that may be used to display or edit DSets.
Ray Greenwell
2004-06-03 17:15:38 +00:00
ae40f1ae0b
SettedObject interface. We may just want to roll this into DObject.
Ray Greenwell
2004-06-03 16:34:07 +00:00
dec746b983
Use ClassUtil.getFields(clazz)
Ray Greenwell
2004-06-03 15:56:25 +00:00
fadc9b82d9
Other entities may also want to queue up a shutdown.
Michael Bayne
2004-06-03 09:03:06 +00:00
0dfbb137e5
Added a mechanism for being informed when the connection manager fails to bind to its socket.
Michael Bayne
2004-06-03 09:02:47 +00:00
62b3ca4ba7
Apparently the proper way to remove a range from a List is to call subList(range).clear();
Ray Greenwell
2004-05-25 21:58:14 +00:00
685f989dc9
A one shot component for tiling an area with a background tiler.
Michael Bayne
2004-05-21 07:36:17 +00:00
9a998f9e69
Forgot something import-ant.
Ray Greenwell
2004-05-19 22:09:04 +00:00
0091a445f5
Don't actually register our scene manager until all of the our subclasses have finished running didStartup. This is a little hacky, but it's far easier than requiring that all our subclasses call super() at the end, and booching it if someday they don't.
Ray Greenwell
2004-05-19 22:04:45 +00:00
680fc5bb38
Still having trouble tracking down a bug related to games that seem to end twice. This should help.
Ray Greenwell
2004-05-19 01:24:10 +00:00
72ea6248ae
Zone names are Name objects.
Michael Bayne
2004-05-18 11:04:08 +00:00
60f9159f25
When flushing inactive clients, don't let a choke in one prevent us from flushing the rest.
Ray Greenwell
2004-05-10 17:45:27 +00:00
11fba91964
Fixed pause-related bug. Lord forgive me for mucking about in here so much, but now I've cleaned up the code that dicks with mouse listeners while paused and moved it into an inner class so that it's easier to understand.
Ray Greenwell
2004-05-05 02:38:22 +00:00
1a0ae950f4
Changed module name to just "puzzle" (we're not longer in yohoho).
Ray Greenwell
2004-05-04 23:36:01 +00:00
d3176c2d9c
Sigh, fixed so that the first click is not eaten if the user unpauses with the keyboard.
Ray Greenwell
2004-04-30 00:02:21 +00:00
cb7007226b
Use the good'ole MouseHijacker to prevent the click that unpauses to pass through to the board view.
Ray Greenwell
2004-04-29 00:28:05 +00:00
d5889452eb
Fixed click-to-unpause for non-blockade duty puzzling...
Ray Greenwell
2004-04-28 18:57:55 +00:00
aa7537d526
When paused, don't install the mouse-clicky unpauser on the boardview, but rather on the whole panel. This allows us to unpause by clicking even if the boardview is removed during pausing, as it is now in blockade duty puzzles.
Ray Greenwell
2004-04-14 00:36:49 +00:00
7308feea9b
More Name conversion.
Ray Greenwell
2004-04-09 17:36:47 +00:00
1d474b4863
By default, do not remove whitespace as part of Name normalization. This is to prevent clashes between an NPP name like "Bad Bob" and a player name like "Badbob". Something more sophisticated, like maybe a specific Name subclass for yohoho usernames, is left as an exercise to the Mike, if desired. Note that crew/flag name normalization is still as strict as ever.
Ray Greenwell
2004-04-06 17:42:49 +00:00
1dc8070910
Checking these in for Ray so that things will compile.
Michael Bayne
2004-04-06 03:53:26 +00:00
c50dbe8b72
Allow the field editor to edit arrays of floats. The sword runtime config currently uses them but it wasn't actually possible to edit them. (Not that I'm messing with the values or the sword code, I just noticed it was seemingly borked.)
Eric Lundberg
2004-04-05 23:44:28 +00:00
beb2539663
Get rid of the not needed code that I had previously checked in for a short stint of actually being used.
Eric Lundberg
2004-03-31 02:16:12 +00:00
a0718bc6cd
Report the locale we're using. It's handy.
Michael Bayne
2004-03-31 02:09:37 +00:00
74eb2bef88
Added code to allow object visitors to visit a specific section.
Eric Lundberg
2004-03-30 01:30:38 +00:00
15b1d7137c
Dead code pruning.
Ray Greenwell
2004-03-28 00:36:39 +00:00
1af8c46fcf
A bit 'o logging in places where it's handy and not too verbose.
Michael Bayne
2004-03-10 16:20:04 +00:00
8a54e85914
Switched to a safe subscriber.
Michael Bayne
2004-03-06 14:55:50 +00:00
e39ff5ffd3
Whoops, we forgot to check in the main attraction.
Michael Bayne
2004-03-06 14:55:03 +00:00
1a4bdbc5c2
Distributed objects go in the data package, not client. While I was in here, I went ahead and cleaned up the custom field editor handling per Ray's earlier suggestion.
Michael Bayne
2004-03-06 12:00:39 +00:00
32dee3cbaf
405 modified source files and 17,367 lines of diffs later we now enforce more discipline when handling names in our code base. Any user entered name should find its way into a Name object as soon as it comes out of a text field or whatnot, and stay that way until it makes its way into a text field or into a database record (for which String objects are vastly simpler because of JORA magic).
Michael Bayne
2004-03-06 11:29:19 +00:00
92b716d790
seems to work like a champ.
Eric Lundberg
2004-03-05 02:40:18 +00:00
20541d11dc
Added code to support serializing and deserializing of Streamable objects in the prefs. Thus we can now have DSets of arbitrary data types in our run time config objects, with swanky custom UI no less.
Eric Lundberg
2004-03-04 02:50:30 +00:00
4e7ea75efa
Changed to use the custom ui if it is available from the configObject.
Eric Lundberg
2004-03-04 02:43:48 +00:00
97207fca74
Initial checkin of the new base config object for run time configs. Allows derived config objects to provide custom editing UI.
Eric Lundberg
2004-03-04 02:42:31 +00:00
df0cf0cf98
Go back to only logging verbose sound events when the runtime adjust in on.
Ray Greenwell
2004-02-26 18:42:51 +00:00
40e04de557
Pruned imports.
Michael Bayne
2004-02-25 14:52:16 +00:00
9ccff79f52
Regenerated.
Michael Bayne
2004-02-25 14:51:51 +00:00
d4595269d5
Brought things back up to date; pruned imports; made Eclipse suggested cleanups.
Michael Bayne
2004-02-25 14:51:26 +00:00
d618d3c771
Eclipse suggested cleanup.
Michael Bayne
2004-02-25 14:50:28 +00:00
6cd4e73720
Import cleanup.
Michael Bayne
2004-02-25 14:49:56 +00:00
ef748636e2
Various Eclipse suggested bits.
Michael Bayne
2004-02-25 14:41:47 +00:00
61d46f2768
Import cleanup.
Michael Bayne
2004-02-25 14:39:34 +00:00
5e00fadbfc
Force the service files to be compiled before we run genservice.
Michael Bayne
2004-02-25 10:02:05 +00:00
7ac816f12c
- Trashed any reference to line.drain(), we're not going back. - Removed the RuntimeAdjust for tweaking the Line's buffer size. - Feed data to the Line in chunks the same size as its buffer. - Rather than sleeping 3 seconds, calculate how long it would take to play the line's entire buffer and sleep that long. It turns out that with the buffer size I'm using and the sample rates our sounds are encoded at, this is less than 400 ms. - Went ahead and increased the maximum number of simultaneous line spoolers to 12.
Ray Greenwell
2004-02-24 08:08:25 +00:00
796c953c89
Window.show() is deprecated in jdk 1.5, replaced with setVisible(true).
Ray Greenwell
2004-02-22 18:55:26 +00:00
bfe9cbba04
Changed the variable named "enum", since that's a keyword in jdk 1.5
Ray Greenwell
2004-02-22 18:52:33 +00:00
ebdf693143
Log the culprit when we report and continue from a fatal error.
Michael Bayne
2004-02-21 02:18:56 +00:00
a7908454a1
Removed insano tracking that allowed me to find the IntroPirate bug.
Ray Greenwell
2004-02-21 01:00:23 +00:00
350c49835c
Let's use proper grammar.
Michael Bayne
2004-02-21 00:54:51 +00:00
f8f4c1a611
If a client sends more than 100 messages in 10 seconds, give them the heave ho.
Michael Bayne
2004-02-21 00:52:16 +00:00
461a72bd94
Mark each upstream message with a the time at which it was decoded from incoming network data.
Michael Bayne
2004-02-21 00:51:02 +00:00
a26e460fd6
Catch fatal errors thrown by invoker units as well. Also only log the stack trace if we're not going to rethrow the error.
Michael Bayne
2004-02-21 00:22:30 +00:00
813e7b0d47
Attempt to catch and recover from fatal errors. If some rogue event listener tries to 'new int[Integer.MAX_VALUE]' or infinitely loop, they will get the smack down and we will go on about our business.
Michael Bayne
2004-02-21 00:04:40 +00:00
Michael Bayne