LGTM.
I think that using the timestamp of the ZipEntry is fine.
--
Bob Vawter
Google Web Toolkit Team
--~--~-~--~~~---~--~~
http://groups.google.com/group/Google-Web-Toolkit-Contributors
-~--~~~~--~~--~--~---
Why build support for Emma specifically? It seems like loading the
class bytes from the classpath if they're at least as new as the
source is a reasonable thing to do in the general case. Also, I think
that EmmaStrategy could be extracted into a general escape mechanism
for accessing external
Working on CssResource has shown a limitation in the Generator model
in that there are no provisions to accumulate state between
invocations of a Generator while rebinding a module. The general
problem with assuming stateful Generators is that the lifecycle and
invocation order of Generators
It was the issue of repeated restrict range of values tags that made
me decide that define-property had a far more obvious semantic when
it's repeated. And, as a convenience to developers, you can add
additional property values with the one tag.
--
Bob Vawter
Google Web Toolkit Team
I agree with Bruce's design. The ability to widen an
initially-restricted set is compelling.
New thread that introduces non-deferred-binding properties:
http://groups.google.com/group/Google-Web-Toolkit-Contributors/t/fa19b93ae079af12
--
Bob Vawter
Google Web Toolkit Team
On Thu, Sep 11, 2008 at 10:00 AM, Bruce Johnson [EMAIL PROTECTED] wrote:
Could we list some examples, so that we can refer to this thread in the
future to understand our motivations?
In summary, the motivation is to allow for properties that have no
well-defined set of possible values or the
On Thu, Sep 11, 2008 at 3:34 PM, Bruce Johnson [EMAIL PROTECTED] wrote:
I swear I'm not just trying to be a pain. For things like this that don't
quite merit a design doc ahead of time, these back-and-forth discussions
have to happen during the code review.
No offense taken or annoyance
Namely:
Hosted mode works fine but compiling tries to rebind twice (even
though I only have
one GWT.create()) and it can't create the printwriter because the
class it is trying to write
already exists. I guess I need a smarter generator. (?)
This is expected behavior. The generator gets
1) Avoiding slow-script warnings (SSWs) during RPC response deserialization
- directly-evalable responses is one approach, to the extent that the
response is written to be break-up-able (i.e. amenable to incremental
commands); in the design doc, it does not appear that this crucial problem
Changes from previous patch:
- SourceInfos are now created only via JProgram and JsProgram.
Descendant collection is now a field in the program types.
- Optimized getRoots() call in SourceInfo
- Addressed nits
SourceInfo is now an abstract class with a protected constructor.
Each AST
On Thu, Oct 2, 2008 at 4:56 PM, Lex Spoon [EMAIL PROTECTED] wrote:
1. Is there any reason to store all correlates as strings? The
I agree that the final API provided by the Correlation type should
be more structured than just a String, but I don't think that it
should hang on to the AST
Would it be correct to summarize the chief blocking issue as needing
to reconsider the equals / compareTo / hashCode functions on
SourceInfo and Correlation?
--
Bob Vawter
Google Web Toolkit Team
--~--~-~--~~~---~--~~
The attached Java file shows a rough sketch of what the
CompilationAnalysis API might look like. The basic idea is to map
J(s)Nodes, SourceInfos, and Correlations onto a simplified API
suitable for building reports from. The inner interfaces would likely
be top-level types in a separate
On Wed, Oct 15, 2008 at 1:39 PM, Lex Spoon [EMAIL PROTECTED] wrote:
Cool! LGTU.
Committed at r3760.
I'm going to follow up to this with a patch to pass in the caller of
makeChild() and makeSynthetic() since findCaller() was removed.
--
Bob Vawter
Google Web Toolkit Team
This patch simplifies the XML parse error detection to not look for a
specific style applied to the error document.
Diffstat:
XMLParserImplSafari.java |61 + 5 - 0 !
1 file changed, 1 insertion(+), 5 deletions(-)
--
Bob Vawter
Google Web Toolkit Team
Ok, I'll get started on the implementation. I'll keep the
CompilationAnalysis artifact type in the ext.linker package, but move
all of the inner interfaces into a sibling ext.soyc to keep the
namespace uncluttered.
--
Bob Vawter
Google Web Toolkit Team
Your space bar is on a hair trigger.
- Class comment of JsInliner has a semicolon starting a line
- Method comment of JsInliner.InvocationCountingVisitor.removeCountsFor
starts a line with a period.
Eclipse treats JavaDoc directives and html tags as acceptable places
for linebreaks. I've
Scott, Miguel, Kelly, and I sat down today to try to resolve some of
the problems with the existing WAR deployment design. Here is our
strawman proposal for how it should work in the future (we can go into
more detail on the rationale). Note that legacy mode would be
supported for 1.6 to
The cast to (T) is JsVisitor.accept() is necessary for javac 1.5.0_13
as shipped on OS X to compile the code.
Otherwise, LGTM.
What are your current thoughts on making runAsync work with XS and
SSO linkers, since they use function closures?
I'll follow up with an updated SOYC reports CL
On Tue, Nov 11, 2008 at 2:02 PM, Lex Spoon [EMAIL PROTECTED] wrote:
One big question: why a separate packer step? My naive understanding
is that the point of a linker is to decide where the various bits of
compiler output will go on disk. Is there a reason to discontinue
that? It sounds
I'm working on a srawman proposal for this; expect a copy later today.
--
Bob Vawter
Google Web Toolkit Team
--~--~-~--~~~---~--~~
http://groups.google.com/group/Google-Web-Toolkit-Contributors
-~--~~~~--~~--~--~---
On Tue, Nov 11, 2008 at 2:39 PM, Ray Cromwell [EMAIL PROTECTED] wrote:
Something tells me the packer proposal might be very useful. To
package GWT apps for offline Android execution, Android demands that
various artifacts be placed in precise directories on disk before
being packaged into an
What exactly is the reason we can't do the same here?
subclipse won't handle linked resources.
Even if it is an
issue, why would we not leave the files in an eclipse subdirectory and have
interested users simply copy them to their trunk directory so we don't trash
any other similarly named
On Wed, Nov 19, 2008 at 9:32 AM, dflorey [EMAIL PROTECTED] wrote:
Do I have to pass Constants, ImageBundle and ImmutableResourceBundle
into a widget to achieve what I'm looking for or is there a smarter
way?
My observation has been that most localized or resource-enable widgets
use a
Adding new resource types is done by defining a new interface that
derives from ResourcePrototype and annotating that new resource with
an @ResourceGeneratorType annotation. Take a look at TextResource as
an example of a simple resource type.
In the simplest case, you could define a new
LGTM
--
Bob Vawter
Google Web Toolkit Team
--~--~-~--~~~---~--~~
http://groups.google.com/group/Google-Web-Toolkit-Contributors
-~--~~~~--~~--~--~---
On Fri, Nov 21, 2008 at 1:37 PM, dflorey [EMAIL PROTECTED] wrote:
1. The CssResources are very powerfull but why do I have to write
a .css file?
An early draft of CssResource with only class-based selectors used
Java annotations as the basis for providing the content of the rule.
It turned
The attached patch against the 1.6 branch adds
EmittedArtifact.getLastModified() which is used to set the mtime on
files written to disk by the compiler. This resolves issue 2280.
The default implementation in the EmittedArtifact base class is
concrete to avoid breaking changes and always
On Thu, Dec 4, 2008 at 1:46 PM, Scott Blum [EMAIL PROTECTED] wrote:
Make that r4250 + r4251; I forgot to commit the calling code in
StandardGeneratorContext.
LGTM
--
Bob Vawter
Google Web Toolkit Team
--~--~-~--~~~---~--~~
Notes:
- Having two servlet containers that need to be shipped with the
distribution seems like it will be problematic from the maintenance
perspective. We should aggressively deprecate the current entry
points if it is decided that we do want to ship both containers in
1.6.
- Several
On Fri, Dec 5, 2008 at 12:02 PM, Lex Spoon [EMAIL PROTECTED] wrote:
Here is an updated version that puts the deferred JS files for each
permutation into a separate subdirectory of deferredjs. That is,
where there used to be a '-', there is now a '/'.
LGTM.
--
Bob Vawter
Google Web Toolkit
dev/core/src/com/google/gwt/core/ext/linker/AbstractLinker.java
| 7170 +1 - 0 !
What's with the @SuppressWarnings tags?
dev/core/src/com/google/gwt/core/ext/linker/impl/SelectionScriptLinker.java
| 20 13 +7 - 0 !
emitSelectionScript(): add a comment to explain
LGTM
--
Bob Vawter
Google Web Toolkit Team
--~--~-~--~~~---~--~~
http://groups.google.com/group/Google-Web-Toolkit-Contributors
-~--~~~~--~~--~--~---
I was trying to demonstrate the practical difference that runAsync
would make when compared to a monolithic deployment to someone today,
and there was no simple way to turn off runAsync that wasn't the
-disableAggressiveOptimizations flag.
This patch adds an undocumented flag -disableRunAsync
Bob, do you have any idea how hard it would be to trigger off of
module settings rather than a command line flag? I believe this will
require threading through the module meta-data into parts of the
compiler that currently don't have access to it.
I looked at making it deferred-binding
This isn't a compiler bug, it's caused by hosted mode being perhaps slightly
overeager? The deal is, the custom field serializer for Arrays.asList()
(com.google.gwt.user.client.rpc.core.java.util.Arrays.ArrayList_CustomFieldSerializer),
has a method getArray0() that uses a JSNI reference
This patch adds an interface and a generator to build a thing called
an AsyncProxy. What this allows you to do is to use a
command-oriented API with runAsync and not have to have your calling
code necessarily be aware of how or when the underlying implementation
gets loaded with runAsync. It's
[Redacted summary]
AsyncProxy:
- Less boilerplate code at the call sites and on the thing being async-loaded.
- Centralized error handling while the concrete type is outstanding.
- Better isolation of concrete types from calling environment, to
give CodeSplitter more room to work.
-
Initial implementation committed at r4347 for more general review with
a desk review of the implementation by Ray. I'll address any
follow-up reviews with a second commit.
Changes from previous patch:
- Added a @DefaultValue annotation to allow @AllowNonVoid getters to
return whatever value
On Thu, Dec 18, 2008 at 6:04 PM, Ray Cromwell cromwell...@gmail.com wrote:
I floated an idea a long time ago of property aliasing/subtyping
(see here:
http://groups.google.com/group/Google-Web-Toolkit-Contributors/browse_thread/thread/8bd9fe0a2676c4a2)
I like the aliasing idea, although
when-property-is name=user.agent value=ie6 /
none
when-property-is name=user.agent value=ie7 /
/none
(you would alias ie8ie7 and ie7ie6, or you'd need to add value=ie8
in the none, which isn't quite forwards compatible)
As a pattern, this seems non-obvious and ultimately harmful in the
I've working on a branch that allows JavaScriptObjects to implement
interfaces with methods. Due to the lack of polymorphic dispatch of
JSOs, for a given method-bearing interface, there can be exactly one
JSO type implementing the interface. The question at hand is whether
or not to allow
https://google-web-toolkit.googlecode.com/svn/changes/bobv/jso_single_impl_r4432
.
--
Bob Vawter
Google Web Toolkit Team
--~--~-~--~~~---~--~~
http://groups.google.com/group/Google-Web-Toolkit-Contributors
-~--~~~~--~~--~--~---
On Tue, Jan 13, 2009 at 10:41 AM, Scott Blum sco...@google.com wrote:
Sure, is there a design doc for this so that I know what I'm looking at?
Not at present, it's pretty much as we discussed. Methods on
SingleJsoImpl are renamed so that JSO$ can unambiguously implement
them as trampoline
On Tue, Jan 13, 2009 at 10:49 AM, BobV b...@google.com wrote:
On Tue, Jan 13, 2009 at 10:41 AM, Scott Blum sco...@google.com wrote:
Sure, is there a design doc for this so that I know what I'm looking at?
I've added a section at the bottom of
http://code.google.com/p/google-web-toolkit/wiki
That's the flag I asked about. In general it makes perfect sense to
update the AST to hold any information that is needed. In this case,
though, if I understand correctly, the information is soley determined
by the target type of the cast.
I put the flag in there so that JsoNormalizer
http://gwt-code-reviews.appspot.com/2801/diff/1/7
File src/com/google/gwt/gen2/widgetbase/public/FastTree.css (right):
http://gwt-code-reviews.appspot.com/2801/diff/1/7#newcode66
Line 66: background: #e3e8f3 url('%selectionBar%') repeat-x;
This could be an @sprite rule that references an
to obfuscated identifiers.
You can either check out the branch or use the following merge command:
svn merge -r4602:HEAD
http://google-web-toolkit.googlecode.com/svn/changes/bobv/elide_rpc_type_names_r4602
This initial version makes the type removal opt-in because it's
fundamentally incompatible
On Thu, Feb 5, 2009 at 12:28 AM, BobV b...@google.com wrote:
http://google-web-toolkit.googlecode.com/svn/changes/bobv/elide_rpc_type_names_r4602
Future changes:
- A compiler flag or optional module to neuter Class.getName().
I've added an -XdisableClassMetadata flag to this branch
On Fri, Feb 6, 2009 at 5:10 PM, BobV b...@google.com wrote:
On Fri, Feb 6, 2009 at 4:29 PM, Scott Blum sco...@google.com wrote:
FYI: this will break some JRE code that depends on inspecting Class names of
array types to do optimized things.
This has been fixed in the branch. I've added
/svn/changes/bobv/elide_rpc_type_names_r4602
Major changes:
- Added configuration property to change generated RPC code to
replace type names on the wire with short idents.
- Allow JSNI to refer to class literals via (@com.foo.Bar::class) syntax.
- Adds an -XdisableClassMetadata flag
On Tue, Feb 10, 2009 at 7:38 PM, Ray Cromwell cromwell...@gmail.com wrote:
Bob, you are awesome. I'm excited this went from design to
implementation so quickly, since it solves a lot of headache in my
current codebase. I only wish it made it into 1.6. :)
s/Bob/GWT team/
Thanks Ray. I'm sure
caller property is unreliable in GWT web mode code and the
caller chain peters out after a few iterations for some unknown
reason. It has its own stack trace property on exceptions, somewhat
like the one Firefox offers.
On 13-Feb-09, at 11:04 AM, BobV wrote: The attached patch implements
web
Is @noflip{} insufficient?
If you go down this route, it would be necessary to add -begin and
-end suffixes as well.
border-begin: 4px;
padding-begin: 5px;
--
Bob Vawter
Google Web Toolkit Team
--~--~-~--~~~---~--~~
Also, what about four-valued properties?
--
Bob Vawter
Google Web Toolkit Team
--~--~-~--~~~---~--~~
http://groups.google.com/group/Google-Web-Toolkit-Contributors
-~--~~~~--~~--~--~---
the
JSORestrictionsChecker. TypeOracle now has its own variation of
JTypeOracle's computeSingleImplData() method.
Please execute this merge command against trunk:
svn merge -r4770:HEAD
http://google-web-toolkit.googlecode.com/svn/changes/bobv/singlejsoimpl_sans_annotation_r4770
Diffstat:
dev/core/src/com
Thanks for the review.
SerializerBase contains the comment Relies on monotonic behavior of
hashcodes in web mode. I see no problem with this (and I've wanted to do so
for JRE stuff at times), but I don't think we've ever explicitly stated this
as a requirement of System.identityHashCode().
Committed with changes from Matt and jgw at r4807.
--
Bob Vawter
Google Web Toolkit Team
--~--~-~--~~~---~--~~
http://groups.google.com/group/Google-Web-Toolkit-Contributors
-~--~~~~--~~--~--~---
The attached patch fixes web-mode compilation for SingleJsoImpl JSO
types that are never explicitly referenced by user code.
Consider the case where you're only ever using interfaces in your code
and never explicitly refer to the implementing JSO type:
client/Foo.java:
interface Foo { methods
On Thu, Feb 26, 2009 at 4:25 AM, Scott Blum sco...@google.com wrote:
1) Shove every singleJsoImpl in up front. The easiest way to do this would
be to add additional seedTypeNames in getCompilationUnitDeclarations() and
make no changes to AbstractCompiler.
Done.
The name of this type seems
On Sat, Feb 21, 2009 at 2:47 AM, Scott Blum sco...@google.com wrote:
On Fri, Feb 20, 2009 at 3:09 AM, nicolas de loof nicolas.del...@gmail.com
wrote:
I may be wrong, but doesn't the ProcessBuilder use the OS interpreter to
parse the forked process command line arguments ? (This may be JRE
Otherwise, LGTM.
http://gwt-code-reviews.appspot.com/9802
--~--~-~--~~~---~--~~
http://groups.google.com/group/Google-Web-Toolkit-Contributors
-~--~~~~--~~--~--~---
On Tue, Mar 3, 2009 at 2:16 AM, Scott Blum sco...@google.com wrote:
SourceGen: all this really does is make it so that when you click on a
JsniMethodBody in the debugger, you can see the source.
Is SourceGenerationVisitor specified to be able to recreate the
original Java input?
Also, why
@Lex or Scott,
Can you review this mostly-deletion patch?
The attached patch cleans up how tag interfaces are implemented by
SingleJsoImpl types.
It is legal for any number of JSO subclasses to implement a tag
interface (one that implies no methods), so for simplicity's sake, it
is assumed
The attached patch adds the header X-GWT-Permutation to GWT RPC
requests. The value can be retrieved on the server-side by calling
RemoteServiceServlet.getPermutationStrongName(). Note that the
permutation's strong name (seen in XYZZY.cache.html) is distinct from
the strong name of the RPC
Scott reviewed this. Committed at r4942.
--
Bob Vawter
Google Web Toolkit Team
--~--~-~--~~~---~--~~
http://groups.google.com/group/Google-Web-Toolkit-Contributors
-~--~~~~--~~--~--~---
Reviewers: scottb, spoon,
Description:
This patch adds deprecation warnings when JSNI methods refer to
deprecated types, methods, or fields. It also adds support for the
@SuppressWarnings annotation on native methods.
Please review this at http://gwt-code-reviews.appspot.com/10802
Affected
Reviewers: scottb, spoon, google-web-toolkit-contributors,
Description:
This patch increases the amount of data in the exported symbol maps to
include
- js ident
- JSNI ident (for methods and fields)
- (enclosing) class name
- member name (for methods and fields)
- origin URI
On Wed, Mar 11, 2009 at 9:13 PM, Bruce Johnson br...@google.com wrote:
Bob has started the process of landing ClientBundle (the fancy new name for
IRB) into the GWT trunk, so this kind of functionality will have a proper
ongoing home sooner rather than later.
ImageResource's sprite builder is
You can use a TextResource from the incubator and dynamically add a
script tag to the document.
I suspect that once ClientBundle is integrated into trunk, it would be
worthwhile to talk about a proper JavaScriptResource type.
--
Bob Vawter
Google Web Toolkit Team
On Fri, Mar 13, 2009 at 10:29 AM, j...@google.com wrote:
LGTM
Thanks or the review. Committed at r5010
If the class names are the same, should it then use other data for a
more complete ordering, and therefore deterministc output?
For the purposes of that comparator, the two class entries
The attached patch extracts a configuration API for use by
RemoteServiceProxy. This semantic API allows developers to override
RPC proxy behavior based on intent, rather than interpreting the
status of a RequestBuilder after the fact.
Diffstat:
client/rpc/RpcRequestBuilder.java | 212
On Wed, Mar 18, 2009 at 7:44 PM, Ray Cromwell cromwell...@gmail.com wrote:
My only criticism of this design is that it only captures a part of the use
cases where the transport mechanism needs override. That is, it helps in
cases where one is making HTTP requests, and needs to modify HTTP
The original idea was to have some kind of ReachabilityOracle that
could tell a Linker if some Java method were still contained in the
compiled JS output. The Generator would create a task-list of
resources to add to the compiled output, and it would be the Linker's
responsibility to add that
On Wed, Mar 25, 2009 at 11:44 AM, codesite-nore...@google.com wrote:
Author: sco...@google.com
Date: Wed Mar 25 11:12:37 2009
New Revision: 5080
Modified:
trunk/dev/core/src/com/google/gwt/dev/PermutationWorkerFactory.java
Log:
Removes the PermutationWorkerFactory.lazyFactories
After 16+ months of on-and-off development, ClientBundle (nee
ImmutableResourceBundle) is moving to GWT trunk.
$ find user/src/com/google/gwt/resources/
user/test/com/google/gwt/resources/ -name *.java | xargs wc | grep
total
9871 34391 317990 total
The
On Wed, Mar 25, 2009 at 7:57 PM, Ray Cromwell cromwell...@gmail.com wrote:
Awesome. With regards to StyleInjector, one feature I'd ask for is the
ability to control where in the cascade it injects. For example, I use IRB
in Chronoscope to inject GWT widget Themes, but what happens is, this
The attached patch re-adds image de-duplication to ClientBundle's
ImageBundleBuilder. (It was there at some point, but disappeared along
the way). The patch also ensures that the ImageResource test is run
in a no-inlining permutation, which would have allowed me to see this
error on my mac.
Reviewers: rjrjr,
Description:
This version of StyleInjector improves upon the incubator version by
extending the IE6 version to include a workaround to limit the number of
times createStyleSheet is called. It also adds injectStyleSheetAfter()
and injectStyleSheetBefore() to allow for greater
Reviewers: jat, scottb,
Description:
This patch fixes a compiler exception relating to SingleJsoImpl types
that have array return types. The bug stems from the fact that
JArrayType extends JClassType, but JArrayType.isClass() returns null.
The fix is to add an explicit check for array types to
On Tue, Mar 31, 2009 at 8:12 PM, Scott Blum sco...@google.com wrote:
This seems kinda hacky since you could recompute the seed func name at
compile time during GenerateJavaScriptAST?
I thought about that, but discounted it because
GenerateJavaScriptAST would have to take a look at the pruned
Committed at r5142.
--
Bob Vawter
Google Web Toolkit Team
--~--~-~--~~~---~--~~
http://groups.google.com/group/Google-Web-Toolkit-Contributors
-~--~~~~--~~--~--~---
[+Ray Cromwell]
@RayC,
What was the exact problem that you were having without an
insertBefore/After method? These methods are somewhat problematic to
implement in a least-surprise way on IE owing to the pigeonhole
principle unless the API is changed to return some kind of token instead
of a
StyleInjector committed to trunk at r5185.
@RayC,
I added an assertion that there is a head tag.
@RayR,
The line lengths seem to be ok according to the eclipse formatting
rules checked into trunk.
I added notes to the Javadoc indicating that not all StyleElements
returned are necessarily
I'm in the process of migrating ClientBundle into trunk and it
occurred to me that even if you remove the use of all deprecated types
from your code, your module might still inherit an otherwise-unused
module.
What do you think about adding another tag to gwt.xml as follows?
module
deprecated
On Mon, Apr 6, 2009 at 1:28 PM, Bruce Johnson br...@google.com wrote:
Sounds pretty useful. We should lock its behavior down more, though. Maybe
Locking it down is just going to get in the way because we can't cover
all of the types of messages that you'd necessarily want to be able to
convey
I don't think that handles the most common use case, though. Wouldn't the
most common thing be, Don't use this; use that? Most people will go to
the absolute minimum trouble necessary, so it's unlikely people would be
included to create an actual URL to point to. (IMHO)
I'm thinking of the
How does that manifest as XML?
deprecated superceded-by=othermodule href=http://google.com/something; /
where both attributes are optional. The URL will be presented by
using the AbstractTreeLogger.log() method that takes a HelpInfo
object.
--
Bob Vawter
Google Web Toolkit Team
On Tue, Apr 7, 2009 at 1:35 PM, Eric Ayers zun...@google.com wrote:
[ERROR] The specified property 'name' is not of the correct type; found
'ConfigurationProperty' expecting 'BindingProperty'
ModuleDefSchema line 794 is producing a bad error message, that should
have included the name of the
failure is what we think it
is.
Review by: bobv (TBR)
LGTM
--
Bob Vawter
Google Web Toolkit Team
--~--~-~--~~~---~--~~
http://groups.google.com/group/Google-Web-Toolkit-Contributors
-~--~~~~--~~--~--~---
Reviewers: spoon, scottb,
Description:
Some user-agents return status code 0 when operating off of file: URLs.
This patch gets Showcase working from local files.
Please review this at http://gwt-code-reviews.appspot.com/20801
Affected files:
On Mon, Apr 13, 2009 at 11:38 AM, Scott Blum sco...@google.com wrote:
LGTM.
Thanks. Committed at r5215.
--
Bob Vawter
Google Web Toolkit Team
--~--~-~--~~~---~--~~
http://groups.google.com/group/Google-Web-Toolkit-Contributors
Reviewers: bruce, scottb, knorton, spoon,
Description:
This patch will alter the encoding of the initial fragment so that the
browser will evaluate it in chunks. Chunking the evaluation has
beneficial usability effects because the browser will periodically
unblock the event loop and this helps
On Mon, Apr 13, 2009 at 2:40 PM, sco...@google.com wrote:
The code LG, one high-level comment about the design. Assuming we're ok
with the design implications, this seems good.
The never-fail way to do this would be to have JsSourceGeneration
visitor indicate offset values of top-level
On Mon, Apr 13, 2009 at 2:55 PM, Scott Blum sco...@google.com wrote:
Actually, I am going to have to torpedo this (sorry!). It just occurred to
me that while comments are only a hypothetical problem, string literals are
a real, actual problem that could cause problems today. A string literal
I've deprecated the public ImmutableResourceBundle interfaces in the
incubator project to encourage users to migrate to the ClientBundle
series of interfaces that are now part of GWT trunk. The IRB code in
the incubator project will not be removed for the foreseeable future,
but all future
On Tue, Apr 14, 2009 at 9:02 AM, dflorey daniel.flo...@gmail.com wrote:
Good news! Are there plans to consolidate ClientBundle and i18n
messages constants?
It would be very nice to have a single bundle for all resources of a
widget to streamline the widget cusomization.
As an interim step,
Here's my original email to Freeland:
What does this mean for generators that are sensitive to the
user.agent property (or any other property that may be extended)?
Imagine that a common component has a CssResource with an @if
user.agent safari rule. The end-developer comes along and extends
On Wed, May 6, 2009 at 4:15 PM, Alex Epshteyn
alexander.epsht...@gmail.com wrote:
Yes, we discussed it some time ago. How is that coming along?
The symbol maps are emitted into the -aux directory.
@Alex, what technique are you using?
--
Bob Vawter
Google Web Toolkit Team
On Wed, May 6, 2009 at 4:49 PM, Alex Epshteyn
alexander.epsht...@gmail.com wrote:
I'm using static code rewriting at the Java source level to maintain a
virtual stack. It's actually working out very well so far. Getting
surprisingly good results for code size, efficiency, and correctness.
1 - 100 of 723 matches
Mail list logo