- d3fca05 Build fix: follow on to r3562. by scottb@google.com · 17 years ago
- 1b420bf 1) Adds the concept of module "configuration" properties which do not affect deferred-binding decisions and may be set to any string value. These configuration properties are exposed to Generators and Linkers as a means of providing global configuration information. by scottb@google.com · 17 years ago
- 1fbb51a Two tweaks to improve Array instantiation performance. by scottb@google.com · 17 years ago
- 78b9fec Support Java 6 as a source language. For 1.5 VMs, rewrite v6 class files to v5. by scottb@google.com · 17 years ago
- 76edf77 Merge disparate code for creating JDT compiler options; simple refactor. by scottb@google.com · 17 years ago
- 3b1edcc Change instance initializer in About.java to static. Most of the work, though, was on the Ant files such that once the filtered copies have already been generated, they will re-generate IFF the gwt OR svn versions have changed (i.e. after an svn up). by fabbott@google.com · 17 years ago
- 5da9113 Merge 1.5 branch changes. by bobv@google.com · 17 years ago
- 9f89297 Adds Emma integration for client classes running in hosted mode. Here are the key parts: by scottb@google.com · 17 years ago
- 406fcd8 Fixes a bug in ASM where invalid input LVT entries can translate into negative length output LVT entries, which causes a ClassFormatError when the JVM tries to load it. by scottb@google.com · 17 years ago
- 3bfc7e7 Fixes issue #2607; TreeItemLogger now eagerly grabs exception info from a thrown exception to break lingering dependency on user code. Partial back-port from OOPHM. by scottb@google.com · 17 years ago
- 10363be Trap unchecked exceptions thrown by generators. by scottb@google.com · 17 years ago
- 6e2abd9 Adds getLastModified() to CompilationUnit and its dependencies. by scottb@google.com · 17 years ago
- 566806a Merging releases/1.5 into trunk by scottb@google.com · 17 years ago
- bd9fe8c Merging releases/1.5 into trunk by scottb@google.com · 17 years ago
- 788dbe1 Merging releases/1.5 into trunk by scottb@google.com · 17 years ago
- 40b80b6 Fixes issue #2379; unstable generators should not crash the compiler or hosted mode. by scottb@google.com · 17 years ago
- 0e2dc628 by fabbott@google.com · 17 years ago
- 5f9b1c6 Adds WindowScrollListener support to the Window. This patch also moves the initHandlers method out of the *Template.js files and into Window, reducing the size of the nocache files by 1.5k. The Window class now injects the initWindowHandlers method onto the outer window as needed, then removes it immediately. by jlabanca@google.com · 17 years ago
- 4a87ea1 Adding SVN branding of About.* with svn branch@version information by fabbott@google.com · 17 years ago
- 9da5012 When tightening method calls, use a more reliable technique to by spoon@google.com · 17 years ago
- 96c7556 Merging releases/1.5 into trunk by scottb@google.com · 17 years ago
- 60d329f Semantics of JClassType.isDefaultInstantiable() now actually match the Javadoc. GWTRunnerGenerator now deals better with nested classes. by scottb@google.com · 17 years ago
- 7c699eb Catch our ThreadDeath to avoid tripping up debuggers. by scottb@google.com · 17 years ago
- 9dbc150 Fix to propagate OOMs correctly. by scottb@google.com · 17 years ago
- 4bc5bb3 Merging releases/1.5 into trunk by scottb@google.com · 17 years ago
- 10d0258 Merging releases/1.5 into trunk by scottb@google.com · 17 years ago
- f1e5adb Fixes a race condition between the final worker thread and the main thread seeing that no threads are left. by scottb@google.com · 17 years ago
- 7207a89 Better cloning support for JGwtCreate expressions; does not try to create new instantiation expressions late in the game. by scottb@google.com · 17 years ago
- 1d181d9 Remove mutable static data to instance data for thread safety. by scottb@google.com · 17 years ago
- 2269498 Merging releases/1.5 into trunk by scottb@google.com · 17 years ago
- d45f90c Refactor of multithreading compiler. This splits the logic out into new classes to get it out of the main GWTCompiler class. It also uses java.util.concurrency instead of primitive locks, so it should be a bit easier to determine what's going on. by scottb@google.com · 17 years ago
- 68cd70e Make compiler go faster. Particularly when you have: by scottb@google.com · 17 years ago
- 6e306e6 Disables constant folding in two cases where a division by zero could result. by spoon@google.com · 17 years ago
- 5543272 GWTShellServlet was keeping files locked too long due to extra inefficient connection acquisition. by scottb@google.com · 17 years ago
- 4c26ea3 Build break; forgot to check in dependency of r3247. by scottb@google.com · 17 years ago
- 22113cd Peephole optimization to speed up basic visitation a tiny bit. by scottb@google.com · 17 years ago
- 5cc0248 Turned off spammy generator perf logging. by scottb@google.com · 17 years ago
- 1eadbdc Avoid recomputing strongname twice. by scottb@google.com · 17 years ago
- fc9aa12 Merging revs 3170 to 3188 from GWT 1.5 branch back into trunk. by zundel@google.com · 17 years ago
- 0eb88b4 Merging releases/1.5@r3125:r3170 into trunk. by scottb@google.com · 17 years ago
- 6851019 We were failing to staticify certain call sites simply because no new static impl methods are created. by scottb@google.com · 17 years ago
- 0391cd0 Merging releases/1.5@r3078:r3125 into trunk. by zundel@google.com · 17 years ago
- 6ea01d7 This makes LongLib.toString faster by only dividing once each time through the loop by spoon@google.com · 17 years ago
- 0986505 Merging releases/1.5@r3048:r3078 into trunk. by jlabanca@google.com · 17 years ago
- 6435ea2 Merging releases/1.5@r2994:r3048 into trunk. by scottb@google.com · 17 years ago
- 41eaeda Tweaks the toString() of new array nodes to look more like Java source code. by spoon@google.com · 17 years ago
- 1ab58a6 Merging releases/1.5@2992:2994 trunk to pick up ResourceOracleImpl fix with nested active packages. by scottb@google.com · 17 years ago
- 9de92df Merging releases/1.5@2941:2990 trunk. by scottb@google.com · 17 years ago
- 88ad0ca Adds additional GWTBridge methods to allow useful faking-out for non-GWTTestCase JUnit support, and potentially server-side use. by scottb@google.com · 17 years ago
- 337e447 When replacing a field reference with a nullField reference, override the type of the reference by spoon@google.com · 17 years ago
- bd0a191 Rescue volatile fields that have a live assignment to them. This is helpful by spoon@google.com · 17 years ago
- 36f5b20 Move multi-expressions out of the test of an if. That is: by spoon@google.com · 17 years ago
- 5de7128 This replaces r2963. It adds a very narrow static eval rule for by spoon@google.com · 17 years ago
- d71242d The JS inliner counts the number of invocations of each function, treating by spoon@google.com · 17 years ago
- 9cba78e When making a static implementation method, set the original parameter types by spoon@google.com · 17 years ago
- 6ac141e If an empty multi-expression somehow sneaks through to GenerateJavaScriptAST, by spoon@google.com · 17 years ago
- c624dcb Rolling back r2922, which was not approved. by scottb@google.com · 17 years ago
- b3cb274 Fixes a crazy IE7 bookmark bug. When a user bookmarks a page with a history by knorton@google.com · 17 years ago
- 7db8583 adding logs to JUnit stuff. Helpful for disambiguating which of several testcases with same method names runs, but primarily aimed at diagnosing a remoteweb test timeout... surely not all of the clients really go AWOL all at once?? by fabbott@google.com · 17 years ago
- b22fa6d Added a banner with gwt.jjs.traceMethods to show where permuations start. by scottb@google.com · 17 years ago
- c11ec95 Sort & format to remove checkstyle warning. by scottb@google.com · 17 years ago
- e729920 This file is a relic. by scottb@google.com · 17 years ago
- 7dde34c Fixes a refresh bug with FF2 where it tries to reload an old (now non-existent) strongname script instead of the correct strongname script. by scottb@google.com · 17 years ago
- 6f7f570 I screwed up my assertions here. by scottb@google.com · 17 years ago
- e13eda7 Fixes issue 1695 by zundel@google.com · 17 years ago
- d342e12 The same resource can actually be mapped into the logical space more than once if it appears on multiple super paths. by scottb@google.com · 17 years ago
- cba4727 Fixes several potential crashes when dealing with units that have serious errors. by scottb@google.com · 17 years ago
- eb274f0 Use DOM manipulation to inject stylesheets referenced from gwt.xml files. by bobv@google.com · 17 years ago
- b6003c7 Added a comment since the java package restrictions impact the serializability of classes. by mmendez@google.com · 17 years ago
- eccf548 Fixed a pair of NPEs that can occur when running a test case that has compile errors. by scottb@google.com · 17 years ago
- fa698bc Fixes Style/Script references to preserve module declaration order in Linker. by scottb@google.com · 17 years ago
- 7fef3d5 A swapped division in revision 2290 effectively disabled JsInliner's check for whether a method is too large to inline. Re-enabling the check causes output size for Showcase and QueWeb to go back down about 3%. by spoon@google.com · 17 years ago
- dc3e029 Fixes issue 2426: Passing a char to a JSNI method. by zundel@google.com · 17 years ago
- f9e1ef3 Support jars and zips on the classpath. by scottb@google.com · 17 years ago
- 2b7149a Fixed a build break introduced in r2845. Added java.io.Serializable as an implemented interface of java.lang.String, but failed to add a serialVersionUID field. This resulted in a serialVersionUID warning which caused the JdtCompilerTest to fail. by mmendez@google.com · 17 years ago
- 5174749 Merged changed/mmendez/STOB r2670:2814 into the trunk. Updated STOB to properly deal with parameterized and arrays types that can lead to infinite recursions if you are not careful. by mmendez@google.com · 17 years ago
- 66ba3f3 Fixed the package used for Serializable since it lives in java.io; enable a TypeOracle to be built without using the standard classes. by mmendez@google.com · 17 years ago
- af66423 Fixes Buganizer #1174117. JavaScript expressions like (a?b:c)?d:e were emitted without the parentheses: a?b:(c?d:e). by spoon@google.com · 17 years ago
- 4a75b36 Dropping the "unknown classpath entry" to TRACE. by scottb@google.com · 17 years ago
- fe3fb84 Provides an escape hatch to avoid metadata-related deprecation warnings. by scottb@google.com · 17 years ago
- 0a64dd0 Torque down log level on invalid classpath entries. by scottb@google.com · 17 years ago
- 7ac9704 Removing some build warnings. by scottb@google.com · 17 years ago
- 26faa98 Rolling back ill-advised FSOD changes, plus added clarifying doc to the misleading-named Utility.getFileFromClassPath(). by scottb@google.com · 17 years ago
- 8559e54 NPE fix. by scottb@google.com · 17 years ago
- 4b09f81 Make JsniRef's constructor protected instead of private, so that it can be extended. by spoon@google.com · 17 years ago
- a9dbeae Make JsniRef be non-final so that the Eclipse plugin can extend it. by spoon@google.com · 17 years ago
- 55e1527 Prevent a module from refreshing twice on initial load. by scottb@google.com · 17 years ago
- 7696398 Fixes a bug where the module's source oracle was failing to refresh. by scottb@google.com · 17 years ago
- 453f525 Aggressively clear transient state between runs. by scottb@google.com · 17 years ago
- b5841e9 Fixes a bug in ResourceOracleImpl that caused rerooted resources to always appear stale whenever any resource was stale. by scottb@google.com · 17 years ago
- 63f3260 FLAMING SWORD OF DEATH by scottb@google.com · 17 years ago
- 0babd08 Use the correct Java class file version constant instead of a handcoded number. by scottb@google.com · 17 years ago
- 640c5df Added warning comments regarding ASM and COMPUTE_MAXS. by scottb@google.com · 17 years ago
- 40e024e - Native methods are now changed to non-native thunks via bytecode rewriting, as part of other rewrites. by scottb@google.com · 17 years ago
- 4827a4c Fixes a bug in r2611: when doing a sloppy scan for JSNI by spoon@google.com · 17 years ago
- ecd4fc8 Speeds up hosted mode's initial creation of JavaScriptObject$.class by avoiding having to compile all JSO subtypes; instead they get synthesized based on TypeOracle supertype information. by scottb@google.com · 17 years ago
- 4f451d8 Fixes a ClassCircularityError for real. Note: this work may be temporary pending redesign of JSNI injection. by scottb@google.com · 17 years ago
- 9df9e16 Warning removal and format. by scottb@google.com · 17 years ago
- d05b318 Optimizes LongFromJSNIChecker's execution time significantly. An initial sloppy pass of FindJSNIRefs quickly checks for problems; only if problems are found is a strict parse done. by scottb@google.com · 17 years ago
- 5dac97a Forgot to commit this earlier; no need to force JSO to be loaded. by scottb@google.com · 17 years ago