[gwt-contrib] Change in gwt[master]: Changes StackTraceDeobfuscator to abstract and provide vario...

2013-03-18 Thread Brian Slesinsky

Brian Slesinsky has posted comments on this change.

Change subject: Changes StackTraceDeobfuscator to abstract and provide  
various factory methods.

..


Patch Set 2:

(3 comments)


File user/src/com/google/gwt/core/server/impl/StackTraceDeobfuscator.java
Line 65:   throw new MissingResourceException("Missing resource: "  
+ filePath, null, filePath);

Should be FileNotFoundException, to agree with the javadoc.


Line 86:   public static StackTraceDeobfuscator fromURL(final URL urlPath) {
Google style would be "fromUrl" instead of "fromURL". In GWT it seems to be  
inconsistent - there is UrlBuilder and UrlUtils but there is also  
GWT.getModuleBaseURL.



Line 89: return new URL(urlPath, fileName).openStream();
I don't know if this will throw FileNotFoundException if there is a 404 (or  
equivalent for other protocols).



--
To view, visit https://gwt-review.googlesource.com/2270
To unsubscribe, visit https://gwt-review.googlesource.com/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: I562e052caef8da7f3434319cf11b8984bc347fe5
Gerrit-PatchSet: 2
Gerrit-Project: gwt
Gerrit-Branch: master
Gerrit-Owner: Goktug Gokdogan 
Gerrit-Reviewer: Brian Slesinsky 
Gerrit-Reviewer: Goktug Gokdogan 
Gerrit-HasComments: Yes

--
--
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups "Google Web Toolkit Contributors" group.

To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.




[gwt-contrib] Change in gwt[master]: Changes StackTraceDeobfuscator to abstract and provide vario...

2013-03-18 Thread Goktug Gokdogan

Goktug Gokdogan has uploaded a new patch set (#2).

Change subject: Changes StackTraceDeobfuscator to abstract and provide  
various factory methods.

..

Changes StackTraceDeobfuscator to abstract and provide various factory  
methods.


This change will remove the need for extending StackTraceObfusctor most of  
the
time as in almost all scenarios it is extended to get the symbol map from  
jar

through the class loader (which makes sense).
Also a factory is provided to get the symbol map from the file system. For  
all
other cases, developers can construct the deobfuscator with a URL path or  
extend

the class to override openInputStream method.

Change-Id: I562e052caef8da7f3434319cf11b8984bc347fe5
Review-Link: https://gwt-review.googlesource.com/#/c/2270/
---
M user/src/com/google/gwt/core/server/impl/StackTraceDeobfuscator.java
M user/src/com/google/gwt/logging/server/StackTraceDeobfuscator.java
2 files changed, 68 insertions(+), 50 deletions(-)


--
To view, visit https://gwt-review.googlesource.com/2270
To unsubscribe, visit https://gwt-review.googlesource.com/settings

Gerrit-MessageType: newpatchset
Gerrit-Change-Id: I562e052caef8da7f3434319cf11b8984bc347fe5
Gerrit-PatchSet: 2
Gerrit-Project: gwt
Gerrit-Branch: master
Gerrit-Owner: Goktug Gokdogan 
Gerrit-Reviewer: Brian Slesinsky 
Gerrit-Reviewer: Goktug Gokdogan 

--
--
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups "Google Web Toolkit Contributors" group.

To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.




[gwt-contrib] Change in gwt[master]: Changes StackTraceDeobfuscator to abstract and provide vario...

2013-03-18 Thread Goktug Gokdogan

Goktug Gokdogan has posted comments on this change.

Change subject: Changes StackTraceDeobfuscator to abstract and provide  
various factory methods.

..


Patch Set 1:

(8 comments)


File user/src/com/google/gwt/core/server/impl/StackTraceDeobfuscator.java
Line 54:* {@link ClassLoader}.
Done


Line 59: return new StackTraceDeobfuscator() {
I don't think this will be a real issue, as if somebody really needs to  
know which anonymous class it is, it will be easy to follow by the line  
number.



Line 72:* Creates a deobfuscator that loads symbol map files from the  
given file path.

Done.
About parameter type: As in most google3 scenarios, directory name is  
usually available as string, so this is more convenient for developer.



Line 83:* Creates a deobfuscator that loads symbol map files from the  
given url path.

Done. (javadoc)

URL doesn't mean HTTP protocol. A file or resource from a jar can all be  
identified via URL. All other fromXXX methods above could have written by  
calling this method instead of anonymous class (I am just doing a minor  
optimization by not using this).


For example; JunitHost will be using this one to get the files from the  
servlet context..



Line 85:   public static StackTraceDeobfuscator fromFileSystem(final URL  
urlPath) {

Opps. Thanks for catching that.


Line 98:   private static class SymbolCache {
I also like them at the end if they are private. Especially, in this case  
keeping the public static methods on the top helps with discoverability.
Google java style guide is more helpful on this as it only asks for logical  
order. Anyway I am OK to change it if this is a strict rule or this class  
is going to be of the few exceptions..



Line 158:* Note that, this has no effect on permutations that the  
symbols have already been loaded for.

Done.
There are 3 logical states here; loaded / partially loaded / not loaded.  
Applied that to your suggestion.



Line 338:   protected abstract InputStream openInputStream(String fileName)  
throws IOException;

Added more description on the javadoc.

Currently this method is also called repeatedly which usually doesn't make  
sense if it is a "not found' error.
I like none of that, however I have other priorities right now. We can  
defer it as the problem was always there.



--
To view, visit https://gwt-review.googlesource.com/2270
To unsubscribe, visit https://gwt-review.googlesource.com/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: I562e052caef8da7f3434319cf11b8984bc347fe5
Gerrit-PatchSet: 1
Gerrit-Project: gwt
Gerrit-Branch: master
Gerrit-Owner: Goktug Gokdogan 
Gerrit-Reviewer: Brian Slesinsky 
Gerrit-Reviewer: Goktug Gokdogan 
Gerrit-HasComments: Yes

--
--
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups "Google Web Toolkit Contributors" group.

To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.




[gwt-contrib] Change in gwt[master]: Remove early optimization from Precompile.

2013-03-18 Thread Brian Slesinsky

Brian Slesinsky has posted comments on this change.

Change subject: Remove early optimization from Precompile.
..


Patch Set 2: Code-Review+1

It makes sense. It seems like the only way this could break something is if  
draftCompile was already broken? What do the numbers look like?


--
To view, visit https://gwt-review.googlesource.com/2280
To unsubscribe, visit https://gwt-review.googlesource.com/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: I26bad5f14735a9283021b53edadc5bde1a659c45
Gerrit-PatchSet: 2
Gerrit-Project: gwt
Gerrit-Branch: master
Gerrit-Owner: Roberto Lublinerman 
Gerrit-Reviewer: Brian Slesinsky 
Gerrit-Reviewer: Ray Cromwell 
Gerrit-HasComments: No

--
--
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups "Google Web Toolkit Contributors" group.

To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.




[gwt-contrib] Change in gwt[master]: Remove early optimization from Precompile.

2013-03-18 Thread Roberto Lublinerman

Roberto Lublinerman has uploaded a new patch set (#2).

Change subject: Remove early optimization from Precompile.
..

Remove early optimization from Precompile.

An early round of optimization results in increased compile time for large
projects. The reason is that the optimizations are run on a bigger
AST that does not benefit of the prunning made available by instantiating
a specific permutation.

Change-Id: I26bad5f14735a9283021b53edadc5bde1a659c45
Review-Link: https://gwt-review.googlesource.com/#/c/2280/
---
M dev/core/src/com/google/gwt/dev/jjs/JavaToJavaScriptCompiler.java
1 file changed, 0 insertions(+), 8 deletions(-)


--
To view, visit https://gwt-review.googlesource.com/2280
To unsubscribe, visit https://gwt-review.googlesource.com/settings

Gerrit-MessageType: newpatchset
Gerrit-Change-Id: I26bad5f14735a9283021b53edadc5bde1a659c45
Gerrit-PatchSet: 2
Gerrit-Project: gwt
Gerrit-Branch: master
Gerrit-Owner: Roberto Lublinerman 
Gerrit-Reviewer: Brian Slesinsky 
Gerrit-Reviewer: Ray Cromwell 

--
--
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups "Google Web Toolkit Contributors" group.

To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.




[gwt-contrib] Change in gwt[master]: Remove early optimization from Precompile.

2013-03-18 Thread Roberto Lublinerman

Roberto Lublinerman has uploaded a new change for review.

  https://gwt-review.googlesource.com/2280


Change subject: Remove early optimization from Precompile.
..

Remove early optimization from Precompile.

An early round of optimization results in increased compile time for large
project. The reason is that the optimizations are run on a bigger
AST that does not benefit of the prunning made available by instantiating
a specific permutation.

Change-Id: I26bad5f14735a9283021b53edadc5bde1a659c45
---
M dev/core/src/com/google/gwt/dev/jjs/JavaToJavaScriptCompiler.java
1 file changed, 0 insertions(+), 8 deletions(-)



diff --git  
a/dev/core/src/com/google/gwt/dev/jjs/JavaToJavaScriptCompiler.java  
b/dev/core/src/com/google/gwt/dev/jjs/JavaToJavaScriptCompiler.java

index 5d1a2ab..17104e4 100644
--- a/dev/core/src/com/google/gwt/dev/jjs/JavaToJavaScriptCompiler.java
+++ b/dev/core/src/com/google/gwt/dev/jjs/JavaToJavaScriptCompiler.java
@@ -716,14 +716,6 @@
* reducing the total permutation count.
*/
   optimize(options, jprogram);
-} else {
-  /*
-   * Do only minimal early optimizations. This code path is used  
by the

-   * Precompile entry point. The external system might be able to
-   * perfectly parallelize the permutation compiles, so let's avoid
-   * doing potentially superlinear optimizations on the unified  
AST.

-   */
-  optimizeLoop("Early Optimization", jprogram, false);
 }
   }


--
To view, visit https://gwt-review.googlesource.com/2280
To unsubscribe, visit https://gwt-review.googlesource.com/settings

Gerrit-MessageType: newchange
Gerrit-Change-Id: I26bad5f14735a9283021b53edadc5bde1a659c45
Gerrit-PatchSet: 1
Gerrit-Project: gwt
Gerrit-Branch: master
Gerrit-Owner: Roberto Lublinerman 

--
--
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups "Google Web Toolkit Contributors" group.

To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.




[gwt-contrib] Change in gwt[master]: Attempt to continue after the JDT compiler aborts.

2013-03-18 Thread Brian Slesinsky

Brian Slesinsky has uploaded a new change for review.

  https://gwt-review.googlesource.com/2271


Change subject: Attempt to continue after the JDT compiler aborts.
..

Attempt to continue after the JDT compiler aborts.

This could occasionally work, but in my test case it just causes a
RuntimeException later due to violating an internal JDT invariant. Still, it
lets us print the names of a few more source files of death before dying.

Change-Id: Ia94d998c6b4e56d880570c726aa09f004a4f552c
---
M dev/core/src/com/google/gwt/dev/javac/JdtCompiler.java
1 file changed, 23 insertions(+), 6 deletions(-)



diff --git a/dev/core/src/com/google/gwt/dev/javac/JdtCompiler.java  
b/dev/core/src/com/google/gwt/dev/javac/JdtCompiler.java

index 05c038c..dfc37e0 100644
--- a/dev/core/src/com/google/gwt/dev/javac/JdtCompiler.java
+++ b/dev/core/src/com/google/gwt/dev/javac/JdtCompiler.java
@@ -226,11 +226,14 @@
   }

   private class CompilerImpl extends Compiler {
+private TreeLogger logger;
+private int abortCount = 0;

-public CompilerImpl() {
+public CompilerImpl(TreeLogger logger) {
   super(new INameEnvironmentImpl(),  
DefaultErrorHandlingPolicies.proceedWithAllProblems(),
   getCompilerOptions(), new ICompilerRequestorImpl(), new  
DefaultProblemFactory(

   Locale.getDefault()));
+  this.logger = logger;
 }

 /**
@@ -248,7 +251,19 @@

 @Override
 public void process(CompilationUnitDeclaration cud, int i) {
-  super.process(cud, i);
+  try {
+super.process(cud, i);
+  } catch (AbortCompilation e) {
+abortCount++;
+String filename = new String(cud.getFileName());
+logger.log(TreeLogger.Type.ERROR,
+"JDT aborted: " + filename + ": " + e.problem.getMessage());
+return; // continue without it; it might be a server-side class.
+  } catch (RuntimeException e) {
+logger.log(TreeLogger.Type.ERROR,
+"JDT died after " + abortCount + " previous errors", e);
+throw new AbortCompilation(cud.compilationResult, e);
+  }
   ClassFile[] classFiles = cud.compilationResult().getClassFiles();
   Map results = new LinkedHashMapCompiledClass>();

   for (ClassFile classFile : classFiles) {
@@ -739,17 +754,19 @@
   icus.add(new Adapter(builder));
 }

-compilerImpl = new CompilerImpl();
+compilerImpl = new CompilerImpl(logger);
 try {
   compilerImpl.compile(icus.toArray(new  
ICompilationUnit[icus.size()]));

 } catch (AbortCompilation e) {
-  if (e.problem.getOriginatingFileName() != null) {
+  if (e.problem == null) {
+logger.log(TreeLogger.Type.ERROR, "JDT compiler aborted");
+  } else if (e.problem.getOriginatingFileName() == null) {
+logger.log(TreeLogger.Type.ERROR, "JDT compiler aborted: " +  
e.problem.getMessage());

+  } else {
 String filename = new String(e.problem.getOriginatingFileName());
 TreeLogger branch = logger.branch(TreeLogger.Type.ERROR,
 "At " + filename + ": " + e.problem.getSourceLineNumber());
 branch.log(TreeLogger.Type.ERROR, "JDT compiler aborted: " +  
e.problem.getMessage());

-  } else {
-logger.log(TreeLogger.Type.ERROR, "JDT compiler aborted: " +  
e.problem.getMessage());

   }
   throw new UnableToCompleteException();
 } finally {

--
To view, visit https://gwt-review.googlesource.com/2271
To unsubscribe, visit https://gwt-review.googlesource.com/settings

Gerrit-MessageType: newchange
Gerrit-Change-Id: Ia94d998c6b4e56d880570c726aa09f004a4f552c
Gerrit-PatchSet: 1
Gerrit-Project: gwt
Gerrit-Branch: master
Gerrit-Owner: Brian Slesinsky 

--
--
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups "Google Web Toolkit Contributors" group.

To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.




[gwt-contrib] Re: Change in gwt[master]: Updates ICU4J from 4.4.2 to 50.1.1

2013-03-18 Thread n . giamouris
Thanks for your reply. Pull as in git-svn. My bad - synced trunk but not 
tools :/

On Tuesday, March 19, 2013 12:20:06 AM UTC+2, John A. Tamplin wrote:
>
> On Mon, Mar 18, 2013 at 5:30 PM, > wrote:
>
>>
>> Any updates on this? Just did a pull and the jar is still missing.
>
>
> When you say a pull, did you mean svn up?  gwt/tools is still managed 
> under svn, and I picked it up just a couple of days ago when I synced.
>
> -- 
> John A. Tamplin 

-- 
-- 
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups 
"Google Web Toolkit Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.




[gwt-contrib] Re: Change in gwt[master]: Updates ICU4J from 4.4.2 to 50.1.1

2013-03-18 Thread John A. Tamplin
On Mon, Mar 18, 2013 at 5:30 PM,  wrote:

>
> Any updates on this? Just did a pull and the jar is still missing.


When you say a pull, did you mean svn up?  gwt/tools is still managed under
svn, and I picked it up just a couple of days ago when I synced.

-- 
John A. Tamplin

-- 
-- 
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups 
"Google Web Toolkit Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.




[gwt-contrib] Change in gwt[master]: Changes StackTraceDeobfuscator to abstract and provide vario...

2013-03-18 Thread Brian Slesinsky

Brian Slesinsky has posted comments on this change.

Change subject: Changes StackTraceDeobfuscator to abstract and provide  
various factory methods.

..


Patch Set 1:

(8 comments)


File user/src/com/google/gwt/core/server/impl/StackTraceDeobfuscator.java
Line 54:* {@link ClassLoader}.
"that loads symbol map and source map files as resources under the given  
resource path. Uses StackTraceObfuscator's ClassLoader."



Line 59: return new StackTraceDeobfuscator() {
Stack traces will be easier to read if you convert these anonymous classes  
into nested classes, so it's clearer which one is in use.



Line 72:* Creates a deobfuscator that loads symbol map files from the  
given file path.
"loads symbol and source map files from the given directory". Maybe the  
argument should be a File?



Line 83:* Creates a deobfuscator that loads symbol map files from the  
given url path.

"loads symbol and source map files from a URL beneath the given parent."

This seems rather dicy. We will end up doing two HTTP requests for each  
file (one for the symbol map and another for the source map). If the file  
doesn't exist, there's only one request, but still someone could use this  
API as a way to do a denial-of-service attack on whichever server we're  
loading the files from. Maybe leave out this constructor for now?



Line 85:   public static StackTraceDeobfuscator fromFileSystem(final URL  
urlPath) {

"fromUrl(final URL parentDir)"


Line 98:   private static class SymbolCache {
According to GWT's usual stye, nested classes go above static methods.  
However, I think it would be better to put the inner classes at the end.



Line 158:* Note that, this has no effect on permutations that the  
symbols have already been loaded for.

"This will only have an effect on symbol maps that haven't been loaded yet."


Line 338:   protected abstract InputStream openInputStream(String fileName)  
throws IOException;
"This method will be called with a filename (not including a slash) for a  
symbol map or a source map file and the implementation should open the file  
in the appropriate directory, or throw an IOException if the file wasn't  
found. The returned InputStream doesn't need to be buffered."


However, maybe we want to do a better job of logging unexpected errors? It  
seems like reporting a simple log message (not a full stack trace) for the  
first few missing files would make it easier to fix configuration errors,  
instead of dropping all IOExceptions silently like we do now. And if it's  
not just a missing file then I think it deserves a full stack trace.


So perhaps we should have a specific exception for a normal file not found  
error, so we can catch it separately?



--
To view, visit https://gwt-review.googlesource.com/2270
To unsubscribe, visit https://gwt-review.googlesource.com/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: I562e052caef8da7f3434319cf11b8984bc347fe5
Gerrit-PatchSet: 1
Gerrit-Project: gwt
Gerrit-Branch: master
Gerrit-Owner: Goktug Gokdogan 
Gerrit-Reviewer: Brian Slesinsky 
Gerrit-HasComments: Yes

--
--
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups "Google Web Toolkit Contributors" group.

To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.




Re: [gwt-contrib] dev tools

2013-03-18 Thread Stephen Haberman

> Chrome already has a protocol for most of this stuff, a wire-debugging
> protocol and source-maps.

Yes, I know.

My point was that FF/Chrome should use the same wire protocol.

That way if GPE ever learns to control the browser/JS debugger (e.g.
for SuperDevMode debugging while staying in Eclipse/GPE), it doesn't
have to implement a different protocol for each browser.

Perhaps it is on the Firefox team's shoulders to play nice and be
compatible with the Chrome protocol; same thing with IE.

My point is that a pro-active "let's share" initiative might be in
order here, to save the IDE developers of the world a lot of hurt.

Vs. the Selenium/WebDriver re-active approach of only standardizing on
a protocol after 5-10 years of wasting the Selenium team's lives trying
to trick each browser into doing what they want.

- Stephen

-- 
-- 
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups 
"Google Web Toolkit Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.




[gwt-contrib] Re: Change in gwt[master]: Updates ICU4J from 4.4.2 to 50.1.1

2013-03-18 Thread n . giamouris

Any updates on this? Just did a pull and the jar is still missing.

On Wednesday, February 6, 2013 8:24:52 PM UTC+2, Brian Slesinsky wrote:
>
> Brian Slesinsky has abandoned this change. 
>
> Change subject: Updates ICU4J from 4.4.2 to 50.1.1 
> .. 
>
>
> Abandoned 
>
> Submitted. 
>
> -- 
> To view, visit https://gwt-review.googlesource.com/1700 
> To unsubscribe, visit https://gwt-review.googlesource.com/settings 
>
> Gerrit-MessageType: abandon 
> Gerrit-Change-Id: I75c9e0e07c4d29a75573701a676a19d6e83b91d9 
> Gerrit-PatchSet: 3 
> Gerrit-Project: gwt 
> Gerrit-Branch: master 
> Gerrit-Owner: John A. Tamplin > 
> Gerrit-Reviewer: Andrew Bachmann > 
> Gerrit-Reviewer: Brian Slesinsky > 
> Gerrit-Reviewer: John A. Tamplin > 
>

-- 
-- 
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups 
"Google Web Toolkit Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.




Re: [gwt-contrib] dev tools

2013-03-18 Thread Ray Cromwell
Chrome already has a protocol for most of this stuff, a wire-debugging
protocol and source-maps. IntelliJ/WebStorm have direct support for this.
WebStorm 6 for example + Chrome will do much of what the Firefox
announcement talks about. It even has "Live Editing"




On Mon, Mar 18, 2013 at 10:36 AM, Stephen Haberman  wrote:

> Hey,
>
> I know I've been quiet on the GWT front lately, but I've really enjoyed
> seeing all of the reviews/commits going by. Great stuff.
>
> You've probably seen it, but there's a Mozilla post about next-gen dev
> tools:
>
> http://paulrouget.com/e/devtoolsnext
>
> I posted a comment on it, but given we have GWT/Googlers here on the
> list, wanted to reiterate: it seems like if we ever want SuperDevMode
> editing/debugging to be anywhere near as sexy/seamless as JvmDevMode
> debugging, we'd want a simple *cross-browser* protocol for doing all of
> the IDE-ish things like step in/out/reload CSS/etc.
>
> I don't think GWT has the man power, either in Google or out, to build
> per-browser hacks into GPE to do all of this sexy stuff from scratch
> (i.e. we're trying to get out of the per-browser/extension game).
>
> Googlers, or anyone else, any comments on whether Chrome devs are
> thinking the same thing? I know Chrome has wire protocols, but
> specifically on a standard debugging protocol...
>
> Seems like a lot of the new Chrome dev tool stuff I've seen lately is
> about rebuilding an IDE inside of Chrome, which just seems so very
> backwards from what GWT users would like to see.
>
> - Stephen
>
> --
> --
> http://groups.google.com/group/Google-Web-Toolkit-Contributors
> ---
> You received this message because you are subscribed to the Google Groups
> "Google Web Toolkit Contributors" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>
>

-- 
-- 
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups 
"Google Web Toolkit Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.




[gwt-contrib] Change in gwt[master]: Changes StackTraceDeobfuscator to abstract and provide vario...

2013-03-18 Thread Goktug Gokdogan

Goktug Gokdogan has uploaded a new change for review.

  https://gwt-review.googlesource.com/2270


Change subject: Changes StackTraceDeobfuscator to abstract and provide  
various factory methods.

..

Changes StackTraceDeobfuscator to abstract and provide various factory  
methods.


This change will remove the need for extending StackTraceObfusctor most of  
the
time as in almost all scenarios it is extended to get the symbol map from  
jar

through the class loader (which makes sense).
Also a factory is provided to get the symbol map from the file system. For  
all
other cases, developers can construct the deobfuscator with a URL path or  
extend

the class to override openInputStream method.

Change-Id: I562e052caef8da7f3434319cf11b8984bc347fe5
---
M user/src/com/google/gwt/core/server/impl/StackTraceDeobfuscator.java
M user/src/com/google/gwt/logging/server/StackTraceDeobfuscator.java
2 files changed, 63 insertions(+), 50 deletions(-)



diff --git  
a/user/src/com/google/gwt/core/server/impl/StackTraceDeobfuscator.java  
b/user/src/com/google/gwt/core/server/impl/StackTraceDeobfuscator.java

index 44d209b..7c4f191 100644
--- a/user/src/com/google/gwt/core/server/impl/StackTraceDeobfuscator.java
+++ b/user/src/com/google/gwt/core/server/impl/StackTraceDeobfuscator.java
@@ -25,9 +25,11 @@
 import java.io.IOException;
 import java.io.InputStream;
 import java.io.InputStreamReader;
+import java.net.URL;
 import java.util.HashMap;
 import java.util.HashSet;
 import java.util.Map;
+import java.util.MissingResourceException;
 import java.util.Scanner;
 import java.util.Set;
 import java.util.concurrent.ConcurrentHashMap;
@@ -45,7 +47,48 @@
  * directory is written. By default, the final symbolMaps  
directory is

  * war/WEB-INF/deploy/yourmodulename/symbolMaps/.
  */
-public class StackTraceDeobfuscator {
+public abstract class StackTraceDeobfuscator {
+
+  /**
+   * Creates a deobfuscator that loads symbol map files from the given  
resource path using

+   * {@link ClassLoader}.
+   */
+  public static StackTraceDeobfuscator fromResource(String symbolMapsPath)  
{
+final String basePath = symbolMapsPath.endsWith("/") ?  
symbolMapsPath : symbolMapsPath + "/";
+final ClassLoader classLoader =  
StackTraceDeobfuscator.class.getClassLoader();

+return new StackTraceDeobfuscator() {
+  protected InputStream openInputStream(String fileName) throws  
IOException {

+String filePath = basePath + fileName;
+InputStream inputStream =  
classLoader.getResourceAsStream(filePath);

+if (inputStream == null) {
+  throw new MissingResourceException("Missing resource: " +  
filePath, null, filePath);

+}
+return inputStream;
+  }
+};
+  }
+
+  /**
+   * Creates a deobfuscator that loads symbol map files from the given  
file path.

+   */
+  public static StackTraceDeobfuscator fromFileSystem(final String  
symbolMapsDirectory) {

+return new StackTraceDeobfuscator() {
+  protected InputStream openInputStream(String fileName) throws  
IOException {
+return new FileInputStream(new File(symbolMapsDirectory,  
fileName));

+  }
+};
+  }
+
+  /**
+   * Creates a deobfuscator that loads symbol map files from the given url  
path.

+   */
+  public static StackTraceDeobfuscator fromFileSystem(final URL urlPath) {
+return new StackTraceDeobfuscator() {
+  protected InputStream openInputStream(String fileName) throws  
IOException {

+return new URL(urlPath, fileName).openStream();
+  }
+};
+  }

   /**
* A cache that maps obfuscated symbols to arbitrary non-null string  
values. The cache can assume

@@ -105,36 +148,16 @@
   private static final int LINE_NUMBER_UNKNOWN = -1;
   private static final String SYMBOL_DATA_UNKNOWN = "";

-  private final boolean lazyLoad;
-  private final File symbolMapsDirectory;
   private final Map sourceMaps = new  
HashMap();

   private final SymbolCache symbolCache = new SymbolCache();
+  private boolean lazyLoad = false;

   /**
-   * Creates a deobfuscator that loads symbol map files from the given  
directory. Symbol maps are
-   * generated into the location specified by the GWT compiler  
-deploy command line

-   * argument.
-   *
-   * @param symbolMapsDirectory the symbolMaps directory,  
with or without trailing

-   *directory separator character
+   * If set to {@code true}, only symbols requested to be deobfuscated are  
cached and the rest is
+   * discarded. This provides a large memory savings at the expense of  
occasional extra disk reads.
+   * Note that, this has no effect on permutations that the symbols have  
already been loaded for.

*/
-  public StackTraceDeobfuscator(String symbolMapsDirectory) {
-this(symbolMapsDirectory, false);
-  }
-
-  /**
-   * Creates a deobfuscator that loads symbol map files from the given  
directory. Symbol maps are
-   * generated i

[gwt-contrib] Change in gwt[master]: fix mismatch in Double/Float.isInfinite between dev / produc...

2013-03-18 Thread Brian Slesinsky

Brian Slesinsky has posted comments on this change.

Change subject: fix mismatch in Double/Float.isInfinite between dev /  
production

..


Patch Set 1: Code-Review+1

--
To view, visit https://gwt-review.googlesource.com/2250
To unsubscribe, visit https://gwt-review.googlesource.com/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: I6e7a891ff1339a64f3d9aaa57978fac6b23fdb64
Gerrit-PatchSet: 1
Gerrit-Project: gwt
Gerrit-Branch: master
Gerrit-Owner: Daniel Kurka 
Gerrit-Reviewer: Brian Slesinsky 
Gerrit-Reviewer: Matthew Dempsky 
Gerrit-Reviewer: Thomas Broyer 
Gerrit-HasComments: No

--
--
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups "Google Web Toolkit Contributors" group.

To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.




[gwt-contrib] Change in gwt[master]: Avoid running JsInliner on the full JavaScript tree.

2013-03-18 Thread Roberto Lublinerman

Roberto Lublinerman has abandoned this change.

Change subject: Avoid running JsInliner on the full JavaScript tree.
..


Abandoned

Submitted, thanks!

--
To view, visit https://gwt-review.googlesource.com/1930
To unsubscribe, visit https://gwt-review.googlesource.com/settings

Gerrit-MessageType: abandon
Gerrit-Change-Id: Iba3cedf4f001f216509f2c3d003ed7fe25725708
Gerrit-PatchSet: 9
Gerrit-Project: gwt
Gerrit-Branch: master
Gerrit-Owner: Roberto Lublinerman 
Gerrit-Reviewer: Brian Slesinsky 
Gerrit-Reviewer: John A. Tamplin 
Gerrit-Reviewer: Ray Cromwell 
Gerrit-Reviewer: Roberto Lublinerman 
Gerrit-Reviewer: Thomas Broyer 

--
--
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups "Google Web Toolkit Contributors" group.

To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.




[gwt-contrib] dev tools

2013-03-18 Thread Stephen Haberman
Hey,

I know I've been quiet on the GWT front lately, but I've really enjoyed
seeing all of the reviews/commits going by. Great stuff.

You've probably seen it, but there's a Mozilla post about next-gen dev
tools:

http://paulrouget.com/e/devtoolsnext

I posted a comment on it, but given we have GWT/Googlers here on the
list, wanted to reiterate: it seems like if we ever want SuperDevMode
editing/debugging to be anywhere near as sexy/seamless as JvmDevMode
debugging, we'd want a simple *cross-browser* protocol for doing all of
the IDE-ish things like step in/out/reload CSS/etc.

I don't think GWT has the man power, either in Google or out, to build
per-browser hacks into GPE to do all of this sexy stuff from scratch
(i.e. we're trying to get out of the per-browser/extension game).

Googlers, or anyone else, any comments on whether Chrome devs are
thinking the same thing? I know Chrome has wire protocols, but
specifically on a standard debugging protocol...

Seems like a lot of the new Chrome dev tool stuff I've seen lately is
about rebuilding an IDE inside of Chrome, which just seems so very
backwards from what GWT users would like to see.

- Stephen

-- 
-- 
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups 
"Google Web Toolkit Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.




[gwt-contrib] Change in gwt[master]: fix mismatch in Double/Float.isInfinite between dev / produc...

2013-03-18 Thread Matthew Dempsky

Matthew Dempsky has posted comments on this change.

Change subject: fix mismatch in Double/Float.isInfinite between dev /  
production

..


Patch Set 1: Code-Review+1

--
To view, visit https://gwt-review.googlesource.com/2250
To unsubscribe, visit https://gwt-review.googlesource.com/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: I6e7a891ff1339a64f3d9aaa57978fac6b23fdb64
Gerrit-PatchSet: 1
Gerrit-Project: gwt
Gerrit-Branch: master
Gerrit-Owner: Daniel Kurka 
Gerrit-Reviewer: Matthew Dempsky 
Gerrit-Reviewer: Thomas Broyer 
Gerrit-HasComments: No

--
--
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups "Google Web Toolkit Contributors" group.

To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.




[gwt-contrib] Re: ie9 bug fixes. (issue1383809)

2013-03-18 Thread t . broyer

Marius: I agree with your comments, but this change has actually been
merged already (despite not being closed here).

Please file an issue and possibly provide a patch (preferably to Gerrit
rather than Rietveld)

As for the reasons of the method overrides, it probably has to do with
this:
http://stackoverflow.com/questions/2717252/document-body-scrolltop-is-always-0-in-ie-even-when-scrolling
(though I haven't tested in IE9 how it actually behaves; keeping in mind
that the DOMImplIE9 is only used for "IE9 in IE9 Standards mode", i.e.
documentMode >= 9)

http://gwt-code-reviews.appspot.com/1383809/

--
--
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups "Google Web Toolkit Contributors" group.

To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.




[gwt-contrib] Re: gwt rebranding

2013-03-18 Thread Jens


> Ah ah, love that new acronym: GWT = Gorgeous Web Toolkit/box ;-)
> (I prefer toolkit to toolbox, it rhymes with GWiT)
>

Ah right, as everyone likes rhymes "toolkit" should be preferred :) Just 
gives a better flow during a talk.

-- J.

-- 
-- 
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups 
"Google Web Toolkit Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.