Commit Graph

  • d0a3fa4c8e Occam's razor: presumably getDefinitionByName() returns the same class for the same name each time it's called. :) Ray Greenwell 2008-03-19 07:30:33 +00:00
  • bb5c4c224c The 'constructor' property is not to be trusted. Ray Greenwell 2008-03-19 07:28:05 +00:00
  • 51bb4d8e92 I am tentatively meddling with Ray's code because pointer comparison on the class fails for objects that are instantiated... elsewhere. Different application domain? Testing on the full class name ought to be safe. Par Winzell 2008-03-19 07:25:36 +00:00
  • 76f77b9748 Support Dictionary, too. Ray Greenwell 2008-03-18 22:13:53 +00:00
  • 783ea0d166 - use isPlainObject() - support Point and Rectangle. Ray Greenwell 2008-03-18 22:01:49 +00:00
  • 95e27a7943 Let's print a stack trace when there's an error in the completeCallback, it could be a ways inside it and we want to help people debug. Ray Greenwell 2008-03-14 23:08:21 +00:00
  • bb23a5caf1 Two fixes: - Added a mechanism for testing if something is already loaded after generation. DataPacks load immediately if passed a ByteArray. Loaders will always wait a frame, so I was going to change DataPack to follow that behavior, but it's nice that it just works right after construction. - If a singleton source was passed in and resulted in an error (or, now, it can already be loaded as well) then it would incorrectly return as the result the internal wrapper object for the single result. Ray Greenwell 2008-03-12 00:31:19 +00:00
  • 6e0bd68148 Roll back the interrupt() yet again - 1) it didn't fix the problem we were trying to fix and 2) it was causing problems with logoff. Mike Thomas 2008-03-10 17:20:03 +00:00
  • 1076052701 Remove the temporary debug logging. Mike Thomas 2008-03-08 01:34:42 +00:00
  • a5506cdbad Avoid putting an empty String at the end of the array in parseURLs(). Ray Greenwell 2008-03-08 01:20:48 +00:00
  • b097e46637 Need a default constructor visible to our descendants. Mike Thomas 2008-03-08 00:35:07 +00:00
  • 4fabfcf0f3 Resort to adding some temporary logging printouts to BlockingCommunicator to figure out what's keeping the reader thread from being shut down. Mike Thomas 2008-03-07 23:53:19 +00:00
  • 7480630c3b Touched up the examples. Added an example for the advanced case of loading classes at runtime. Ray Greenwell 2008-03-07 22:21:59 +00:00
  • 0af0ce8469 Re-instate the interrupt() but setup some additional exception handling so we don't spew to the logs unnecessarily due to this. Mike Thomas 2008-03-07 20:09:13 +00:00
  • 69646f624b Roll this back - calling interrupt() in this way caused the other end's writer to get fairly upset. Mike Thomas 2008-03-06 21:43:08 +00:00
  • d153a06aa2 Reinstate the kick in the pants for the reader thread. Previously we relied on the LogoffRequest being received by the server which resulted in the server closing its end of the socket which resulted in the reader being woken up to hear that its socket was closed and then it exited. We'll just have it exit immediately. It's possible that messages in transit to the reader that would have been read prior to full shutdown will now be dropped instead of delivered, but the client has already requested to logoff, so they can't really care about additional messages coming from the server. Michael Bayne 2008-03-06 19:14:19 +00:00
  • 8a745b30e0 Listen for SIGTERM as well and gracefully shutdown in that case. Michael Bayne 2008-03-05 23:28:28 +00:00
  • e8b5d41fb8 Split out the way we handle the preferred port so we can override that. Mike Thomas 2008-03-05 19:09:36 +00:00
  • 8a35738654 Fixed asdoc bug. Boy it's annoying when this happens, asdoc just gives "Error #1083" without telling you what happened or even which file. Ray Greenwell 2008-03-05 02:32:00 +00:00
  • e1ee1162f9 A bit of commenting touch-up. Ray Greenwell 2008-03-05 00:52:52 +00:00
  • aae7eb4e58 This isn't actually a valid asdoc tag. Ray Greenwell 2008-03-05 00:21:23 +00:00
  • a82ead698b Let's deprecate but not remove EmbeddedSwfLoader. Ray Greenwell 2008-03-05 00:20:05 +00:00
  • e53103cf92 Introducing.. MultiLoader! This was driven by API simplifcation, but it turns out to be a really cool class. - Replaces some of the guts of DataPack. - Replaces DataPackLoader. - Replaces EmbeddedSwfLoader, although I'm not going to remove that just yet. - Can be used to coordinate loading of multiple instances of just about anything. Ray Greenwell 2008-03-04 23:36:13 +00:00
  • b279ef6eb8 Added Util.isPlainObject(). Ray Greenwell 2008-03-04 23:33:59 +00:00
  • 4f0d94df11 Whoopsie, we want to build against flex, not link it in. Ray Greenwell 2008-03-04 04:55:15 +00:00
  • 8493434099 ArrayUtil.copyOf() Tom Conkling 2008-03-04 03:18:31 +00:00
  • 0b068d5189 You know, generifying that correctly is actually a big scary tunnel of dealing with our generated services & whatnot, so that's a much deeper rabbit hole than I was thinking when I just did that one big, so I'm going to leave that well enough alone for now instead of half-assing it. Dave Hoover 2008-03-03 22:11:45 +00:00
  • 8c5e04aac5 Generify. Dave Hoover 2008-03-03 22:06:49 +00:00
  • 938be601e2 Screw it, let's just do the right thing. We'll forget later. - Use the chaining method to get a resource from ResourceManager. - Fixed getAll so that it will not overwrite a found property with one in the parent. Made it also understand locales and do the chaining. Ray Greenwell 2008-02-24 04:13:11 +00:00
  • 8823d592fd Added getAll(). Ray Greenwell 2008-02-24 02:58:26 +00:00
  • da02f02777 Updated to work with the revamped resource handling of flex 3. We're supposed to use ResourceManager.getString(bundleName, rsrcName) and that will do cool locale chaining, but then we're looking up the bundle every time. We may want to revisit this in the future. (Cool chaining: we can compile the app with the local order as [ "jp_JP", "en_US" ], and then if we request the "b.buy" key from the "game" bundle, it will look in japanese for b.buy, but fall back to english if not found.) Ray Greenwell 2008-02-24 02:48:37 +00:00
  • 32354987f2 Updated for flex 3. Ray Greenwell 2008-02-22 22:40:51 +00:00
  • f52f686f19 Use the enhanced for-loop. Ray Greenwell 2008-02-21 02:30:26 +00:00
  • 5cbc654b13 Hm, this wasn't validating the values inside objects. Fixed. Ray Greenwell 2008-02-21 02:28:59 +00:00
  • 6dee96bb15 use Function.call(), not Function.apply() Tom Conkling 2008-02-20 00:55:28 +00:00
  • 8929c2e89e added RingBuffer.every(), RingBuffer.forEach(), RingBuffer.indexOf() Tom Conkling 2008-02-20 00:40:43 +00:00
  • 8a7f4c4121 Clarify that SetListeners get all Entry events, not just the ones for a particular set Charlie Groves 2008-02-19 23:05:30 +00:00
  • 7564f88796 Expose the type on Entry* events to SetListener and SetAdapter Charlie Groves 2008-02-19 20:40:09 +00:00
  • 875d4988a3 Added Comparators.COMPARABLE. Ray Greenwell 2008-02-19 00:40:36 +00:00
  • cde838dc45 Call the Flash libraries {library}lib.swc to match the pattern of flexlib.swc and also to make the Maven artifact not look like a meta-artifact that brings in all Narya/Nenya/Vilya dependencies. Michael Bayne 2008-02-18 19:12:29 +00:00
  • 4bb7a2f6a3 Build our Flash libraries like we build our Java libraries. Michael Bayne 2008-02-18 18:33:25 +00:00
  • 8ccdf19a42 Fix docs, clarify. Ray Greenwell 2008-02-05 23:57:57 +00:00
  • 81c4fb64db - Use safeXML utility methods. - Some doc improvements on StringUtil. Ray Greenwell 2008-02-05 23:48:07 +00:00
  • 3d2166cd8d Added utility methods for safe XML operations. Ray Greenwell 2008-02-05 23:47:20 +00:00
  • 2ab08d0a6f Removed dependance on mx.utils.StringUtil. Ray Greenwell 2008-02-04 21:03:01 +00:00
  • 6c70ca11be Commenting improvements. Ray Greenwell 2008-02-04 21:02:42 +00:00
  • 206fd85f99 Updated StringUtil.toString() to print Objects (associative hashes). Added some comments about inadequacies. Ray Greenwell 2008-02-01 19:34:54 +00:00
  • 4f5d065405 Access this static method in a static way. Dave Hoover 2008-01-31 21:26:37 +00:00
  • 80cbcdb396 Format our timestamp like the timestamp in our Java logs. Ray Greenwell 2008-01-27 22:51:19 +00:00
  • bab168d443 Removed MethodQueue.setStage(), use the same trick we use in AnimationManager and that I saw in use in Tweener: just listen for ENTER_FRAME on some random sprite. Ray Greenwell 2008-01-24 21:47:11 +00:00
  • cd421f7fa7 Attempt to quit with an error code when an assertion fails. This only works in the standalone debug player, if the swf is a trusted local file. Ray Greenwell 2008-01-24 18:09:49 +00:00
  • 24f557c0d7 Speeling Dave Hoover 2008-01-22 23:57:54 +00:00
  • 2ce35db583 Remove unneeded casts. Dave Hoover 2008-01-17 01:25:52 +00:00
  • 346e4ee781 Documentation improvements. Ray Greenwell 2008-01-11 19:07:14 +00:00
  • 34875278b3 Only run ninstall on "unix" operating systems (which includes Mac OS X). Michael Bayne 2008-01-08 22:22:38 +00:00
  • 643982e798 If we just bail after a connection closes during resolution, the ClientManager still thinks we're resolved and can just resume the session if the client reconnects. Resuming fails because we don't have _clobj to resume with, so the client fails to connect over and over again until the half-connected client is flushed from ClientManager's cache several minutes later. Just clear the client out of the cache immediately instead. Charlie Groves 2008-01-08 04:12:35 +00:00
  • 460d826d91 Avoid calling the type constant for an event with only one type TYPE. Follow actionscript convention and create a reasonable name for it. Ray Greenwell 2008-01-08 01:35:01 +00:00
  • a5b132757b We don't need these methods, really. They are freaking out asdoc. Ray Greenwell 2007-12-22 01:51:46 +00:00
  • 148766903e created RingBuffer collection class Tom Conkling 2007-12-21 01:57:26 +00:00
  • 9cb8f380b0 Now you can have a simple Object hash that understands Hashable keys. (Because it just uses a Map of your choosing under the hood.) There is really no new functionality here, but it is neat, except for the one stupid thing: keys get converted to Strings when iterating with a 'for' loop. Ray Greenwell 2007-12-21 00:29:04 +00:00
  • 768e6a76be Create our shutdowner list at class init time so that it is safe to register a shutdowner at any time. Michael Bayne 2007-12-20 22:05:49 +00:00
  • d70b867ddb Make null keys work with SortedHashMap. Ray Greenwell 2007-12-20 21:21:44 +00:00
  • c90e8ef83d First pass at SortedHashMap. Ray Greenwell 2007-12-20 20:20:04 +00:00
  • 62e8a39e52 Enable idle reporting. Ray Greenwell 2007-12-19 20:00:36 +00:00
  • dd8fcfc0a7 Allow subclasses to translate the status if a nonstandard one is in use. Ray Greenwell 2007-12-19 19:14:47 +00:00
  • 5e2a70bf5c Nix marker interface that's not used by anything. If we do ever want something like this, we can think about where the interface would be useful. Michael Bayne 2007-12-18 00:19:56 +00:00
  • 6e55c2df8b Use samskivert signal wrapper. Michael Bayne 2007-12-17 22:23:51 +00:00
  • 2bf1c107d1 Added a simple utility method to initialize an object with properties specified in an initObject. This is a fairly common practice in flash and let's go ahead and use it sometimes, even though it's prone to runtime errors. Our version of this has a defaults Object as well as another Object that can be specified to mask properties from the initObject. Ray Greenwell 2007-12-13 23:25:29 +00:00
  • 73ba77af8e Boolean reduction. Ray Greenwell 2007-12-13 23:21:34 +00:00
  • 4a86621196 Added parseUnsignedInteger. Ray Greenwell 2007-12-13 23:21:04 +00:00
  • 901cb095ae Weakening the contract on Set.toArray(), to remove the non-modification guarantee from the resulting array. Users requiring strong non-modification guarantees can manually do a deep or a shallow copy of the array, as appropriate for their task. Robert Zubeck 2007-12-10 20:35:45 +00:00
  • dd85d0047e Allow the currently registered ChatFilters to be retrieved. Ray Greenwell 2007-12-10 03:33:15 +00:00
  • 274811e8f6 Accept an optional Random for shuffle(). Ray Greenwell 2007-12-06 03:34:59 +00:00
  • fcdace5589 Some method documentation. Ray Greenwell 2007-12-05 02:25:44 +00:00
  • c42510b325 Another place where we call IDataInput's readBytes. Let's just make sure we never pass it 0 for the length. Ray Greenwell 2007-12-05 02:11:45 +00:00
  • a920f02030 Found and fixed the "Mother Fucker of a Bug(tm)" that was causing Zyraxxis to hork. Deep within our streaming code, ArrayMask objects were being created with zero length when streaming non-final arrays with a zero length. This tickled a "feature" of Adobe's IDataInput code, which reads *all available bytes* if you pass readBytes() a length of 0. What's really annoying is that this worked fine if you were notified of the zero-length array via a PropertySetEvent, because that event would come in on its own frame and so if the array's length was zero then there were no more bytes to read. But if you entered a game and had to receive all the usercode properties at once, it booched. There are clumps of my pulled-out hair all around my desk from that. And: this is ultimately my own fault, as my old code didn't handle the case of the passed-in ByteArray having 0 length. Ray Greenwell 2007-12-05 02:09:59 +00:00
  • 2b5933015e Some documentation. Ray Greenwell 2007-12-04 03:24:52 +00:00
  • 1b054677c2 Allow ValueEvents to be cancelable. Ray Greenwell 2007-12-04 03:20:26 +00:00
  • 4ad889c106 properly set keyword properties on Set.as Tom Conkling 2007-12-03 21:43:10 +00:00
  • 507f214016 whoops - added Set.as to svn Tom Conkling 2007-12-03 21:26:43 +00:00
  • c5a557214b ISet -> Set Tom Conkling 2007-12-03 21:22:53 +00:00
  • 635bb0a13f Track that we're switching servers so that the client can differentiate during did logoff between a real logoff and a server switch. Michael Bayne 2007-12-03 21:21:50 +00:00
  • 289475b253 Keyword substitution. Dave Hoover 2007-12-03 21:07:34 +00:00
  • 9be9742398 added HashSet and ISet to Narya Tom Conkling 2007-12-03 20:42:19 +00:00
  • 4c4d253974 Let's add a couple more very handy helper methods. Michael Bayne 2007-12-03 19:54:58 +00:00
  • cb46d3a1af Allow the unit to throw InvocationException and do the right thing with it. Michael Bayne 2007-12-03 19:45:53 +00:00
  • c935c46612 Handle the client connecting and disconnecting before client object resolution completed, or connecting, disconnecting and reconnceting before client object resolution completes a little more gracefully. Michael Bayne 2007-12-03 18:28:07 +00:00
  • 362e506fd6 don't need insert - you can add values with ArrayUtil.splice() Nathan Curtis 2007-11-30 02:08:13 +00:00
  • 9ddea518fc added an insert method Nathan Curtis 2007-11-30 01:54:51 +00:00
  • be0eaf6a22 Make it explicit in the function signature that we're just doing an "Object == null" comparison. Otherwise the user might pass in an undefined, and expect a strict equality test. Robert Zubeck 2007-11-28 02:34:50 +00:00
  • 983adf75e5 Nix the wacky arguments thing. Let the caller simply beware. Michael Bayne 2007-11-28 01:03:49 +00:00
  • 69a1d2ec23 Slight change to usage and documentation. Michael Bayne 2007-11-28 00:47:16 +00:00
  • 95adfeb646 Regenerated services. Michael Bayne 2007-11-28 00:25:35 +00:00
  • 9e835ff6e9 Added a way to package up an action and data and send it to all appropriate servers for execution. The class in question must exist on the remote server but all of our servers are all always running the same code, so this follows. Michael Bayne 2007-11-28 00:25:23 +00:00
  • 195f78376b Haha, there was already one in StringUtil. I was sure I'd checked... Par Winzell 2007-11-24 21:50:54 +00:00
  • 9a8df29c8e I'm afraid we really do need a hashcode method... Par Winzell 2007-11-24 21:41:53 +00:00
  • 63db77667a Some type safety, redundant cast removal, and other code hygiene. Michael Bayne 2007-11-24 19:40:19 +00:00
  • 6eb2c79b19 Flag overrides in this template. Dave Hoover 2007-11-21 19:16:05 +00:00
  • aa418e19ee If the reboot time requested is at exactly the right time for the warning, we can use that warning and not require to go to the next level. Mike Thomas 2007-11-20 18:26:31 +00:00
  • 31dbdc78be A simple Flash assert class. Where's a preprocessor when you need one... Robert Zubeck 2007-11-16 22:56:08 +00:00