Re: [gwt-contrib] bug with the new snapshot

2014-10-03 Thread Julien Dramaix
Thanks Daniel,

I've tested your patch and it solves the issue

On Fri, Oct 3, 2014 at 7:41 AM, 'Daniel Kurka' via GWT Contributors <
google-web-toolkit-contributors@googlegroups.com> wrote:

> Hi Julien,
>
> I think this is related to us not waiting for the body to be loaded before
> injecting the recompile.nocache.js.
>
> Can you try: https://gwt-review.googlesource.com/#/c/9451/ and let me
> know if this fixes your issue?
>
> -Daniel
>
> On Thu, Oct 2, 2014 at 4:34 AM, Julien Dramaix 
> wrote:
>
>> This morning, I've updated my GWT snapshot to the last one and I'm not
>> able to run the super dev mode (with the flag -superDevMode) for my
>> applications.
>>
>> The problem comes from that the .nocache.js tries to add the compilation
>> dialog to the body before the body is initialized.
>>
>> Error:
>> Uncaught TypeError: Cannot read property 'appendChild' of null
>> in:
>> Dialog.prototype.show = function() {
>>   $doc.body.appendChild(this.__overlay);
>>   $doc.body.appendChild(this.__dialog);
>> };
>>
>> $doc.body is null when the method is called. This method should be called
>> when the html is fully loaded.
>>
>>
>> If I include the nocache.js script at the end of the body instead of in
>> the head of the html, it works fine.
>>
>> Julien
>>
>>
>>  --
>> You received this message because you are subscribed to the Google Groups
>> "GWT Contributors" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
>> To view this discussion on the web visit
>> https://groups.google.com/d/msgid/google-web-toolkit-contributors/CABb_3%3D5%3DMU8--RcXrHwjM8ZSPvB%2B2MG-7nAvF1Sq8BjDEeJ5Kw%40mail.gmail.com
>> 
>> .
>> For more options, visit https://groups.google.com/d/optout.
>>
>
>
>
> --
> Google Germany GmbH
> *Dienerstr. 12*
> *80331 München*
>
> Registergericht und -nummer: Hamburg, HRB 86891
> Sitz der Gesellschaft: Hamburg
> Geschäftsführer: Graham Law, Katherine Stephens
>
> --
> You received this message because you are subscribed to the Google Groups
> "GWT Contributors" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/google-web-toolkit-contributors/CALLujiqQnryHwbPUYRV1Uyborxpd31sFNHxSUkz063k7otE5Bw%40mail.gmail.com
> 
> .
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/CABb_3%3D5AriYNmoaFxdmKNzzECyh907ZT0AD8Kv7R4uHHiZ9muA%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.


[gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread Александр Новоселов
Will you update to Jetty 9.2 (with JVM 8 support)?
Any link to branch version?

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/ad0b7807-4da7-4a83-844d-dae11c3c786f%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread Manuel Carrasco Moñino
1.- History mechanism has changed in 2.7 to fulfil modern standards
https://gwt-review.googlesource.com/#/c/5356/.  BTW: you should't use
HistoryImpl and in general any Impl class, in general they are private API
2.- Use -XjsInteropMode  ONLY in the case that you have a need for
exporting gwt classes to JS or wrapping JS classes in gwt. Otherwise you
could have a increment of size and probably other issues. It is
experimental, even in 2.7.0, so don't use it for production or be prepared
to changes and breakages
3.- Code sizes in 2.6 and 2.7 should be very similar, but there are changes
in many places, so little differences are acceptable. Update to a new
version does not necessary bring a reduction on code size, but other
improvements.



On Fri, Oct 3, 2014 at 1:38 AM, confile 
wrote:

> Code size explodes when using GWT 2.7 with -XjsInteropMode JS
>
> I compiled the same code with GWT 2.6.1 and GWT 2.7 with argument strict
>
> code size GWT 2.6.1 1.351 MB
> code size  GWT 2.7  1.361 MB
>
> Then I compiled the same code with -XjsInteropMode JS flag on GWT 2.7
>
> code size  1.601 MB
>
>
> Why does GWT 2.7 increases the code size in every case and with Interop
> even by 300 kb?
>
>
> Am Freitag, 3. Oktober 2014 00:18:22 UTC+2 schrieb confile:
>
>> What is the difference between -XjsInteropMode JS and -XjsInteropMode
>> CLOSURE?
>>
>>
>>
>> Am Donnerstag, 2. Oktober 2014 23:49:17 UTC+2 schrieb John Stalcup:
>>>
>>> -XjsInteropMode JS
>>>
>>> On Thu, Oct 2, 2014 at 2:46 PM, confile 
>>> wrote:
>>>
 How can  JsInterop be activated in GWT 2.7? What flag does it?


 Am Donnerstag, 2. Oktober 2014 03:22:56 UTC+2 schrieb Brian Slesinsky:
>
> It's experimental and hidden behind a flag, but it's there.
>
>
> On Wed, Oct 1, 2014 at 6:20 PM, Cristian Rinaldi 
> wrote:
>
>> - JsInterop Preview is part of the release?
>>
>> El miércoles, 1 de octubre de 2014 16:15:26 UTC-3, Daniel Kurka
>> escribió:
>>
>>> Hi all,
>>>
>>> we just settled on a GWT 2.7 release plan:
>>>
>>>  - We *code freeze* on *October 7th* and branch for GWT 2.7.
>>>  - As soon as we have the *remaining patches submitted*, we put out
>>> a beta1 build, this should be no later than *October 7th.*
>>>  - Putting out a *beta1 externally* allows us to collect feedback
>>> on the new super dev mode integration externally as well.
>>>  - We are going to *flip incremental to default* tomorrow and *wait
>>> for 1-2 weeks* for google internal feedback, if there is no serious
>>> issues we are going to *put out RC1*
>>>  - GWT 2.7 will still be compatible with Java 6.
>>>
>>> Patches / Fixes that need to go in:
>>>  - Recompile on reload: https://gwt-review.goo
>>> glesource.com/#/c/9323/ (dankurka)
>>>  - Sending the wrong permutation to the client in SDM, if no files
>>> have changed (dankurka).
>>>  - Investigate why some people are seeing errors with incremental
>>>  not restricting to one permutation (dankurka).
>>>  - Public directories are not copied o the war directory when using
>>> SDM (skybrian).
>>>  - Restore Java 6 compatibility (skybrian).
>>>  - Document limitations of JsonUtils.safeEval and discourage usage
>>> (goktug) (promote Json.parse)
>>>
>>> Patches that are nice to have:
>>>  - Improve exception logging in SDM (goktug).
>>>
>>> *If you have any outstanding patches that you thing need to go into
>>> GWT 2.7, please bring them to our attention, by replying to this thread 
>>> or
>>> adding me as a reviewer on Gerrit and setting the topic to GWT2.7.*
>>>
>>> -Daniel
>>>
>>>  --
>> You received this message because you are subscribed to the Google
>> Groups "GWT Contributors" group.
>> To unsubscribe from this group and stop receiving emails from it,
>> send an email to google-web-toolkit-contributors+unsubscribe@
>> googlegroups.com.
>> To view this discussion on the web visit https://groups.google.com/d/
>> msgid/google-web-toolkit-contributors/1d5a4369-e03d-47ba-b8dd-
>> 5031e5460751%40googlegroups.com
>> 
>> .
>>
>> For more options, visit https://groups.google.com/d/optout.
>>
>
>  --
 You received this message because you are subscribed to the Google
 Groups "GWT Contributors" group.
 To unsubscribe from this group and stop receiving emails from it, send
 an email to google-web-toolkit-contributors+unsubscribe@
 googlegroups.com.
 To view this discussion on the web visit https://groups.google.com/d/
 msgid/google-web-toolkit-contributors/48e8f729-b645-
 4534-af9d-c5a9947dffd2%40googlegroups.com
 

Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread confile
Well, I need to override HistoryImpl to enable HTML5 pushstate api.
 

Am Freitag, 3. Oktober 2014 16:01:31 UTC+2 schrieb Manuel Carrasco Moñino:
>
> 1.- History mechanism has changed in 2.7 to fulfil modern standards 
> https://gwt-review.googlesource.com/#/c/5356/.  BTW: you should't use 
> HistoryImpl and in general any Impl class, in general they are private API
> 2.- Use -XjsInteropMode  ONLY in the case that you have a need for 
> exporting gwt classes to JS or wrapping JS classes in gwt. Otherwise you 
> could have a increment of size and probably other issues. It is 
> experimental, even in 2.7.0, so don't use it for production or be prepared 
> to changes and breakages
> 3.- Code sizes in 2.6 and 2.7 should be very similar, but there are 
> changes in many places, so little differences are acceptable. Update to a 
> new version does not necessary bring a reduction on code size, but other 
> improvements.
>
>
>
> On Fri, Oct 3, 2014 at 1:38 AM, confile  > wrote:
>
>> Code size explodes when using GWT 2.7 with -XjsInteropMode JS 
>>
>> I compiled the same code with GWT 2.6.1 and GWT 2.7 with argument strict
>>
>> code size GWT 2.6.1 1.351 MB
>> code size  GWT 2.7  1.361 MB
>>
>> Then I compiled the same code with -XjsInteropMode JS flag on GWT 2.7
>>
>> code size  1.601 MB
>>
>>
>> Why does GWT 2.7 increases the code size in every case and with Interop 
>> even by 300 kb?
>>
>>
>> Am Freitag, 3. Oktober 2014 00:18:22 UTC+2 schrieb confile:
>>
>>> What is the difference between -XjsInteropMode JS and -XjsInteropMode 
>>> CLOSURE?
>>>
>>>
>>>
>>> Am Donnerstag, 2. Oktober 2014 23:49:17 UTC+2 schrieb John Stalcup:

 -XjsInteropMode JS

 On Thu, Oct 2, 2014 at 2:46 PM, confile  
 wrote:

> How can  JsInterop be activated in GWT 2.7? What flag does it?
>
>
> Am Donnerstag, 2. Oktober 2014 03:22:56 UTC+2 schrieb Brian Slesinsky:
>>
>> It's experimental and hidden behind a flag, but it's there.
>>
>>
>> On Wed, Oct 1, 2014 at 6:20 PM, Cristian Rinaldi  
>> wrote:
>>
>>> - JsInterop Preview is part of the release?
>>>
>>> El miércoles, 1 de octubre de 2014 16:15:26 UTC-3, Daniel Kurka 
>>> escribió:
>>>
 Hi all,

 we just settled on a GWT 2.7 release plan:

  - We *code freeze* on *October 7th* and branch for GWT 2.7.
  - As soon as we have the *remaining patches submitted*, we put 
 out a beta1 build, this should be no later than *October 7th.*
  - Putting out a *beta1 externally* allows us to collect feedback 
 on the new super dev mode integration externally as well.
  - We are going to *flip incremental to default* tomorrow and *wait 
 for 1-2 weeks* for google internal feedback, if there is no 
 serious issues we are going to *put out RC1*
  - GWT 2.7 will still be compatible with Java 6.

 Patches / Fixes that need to go in:
  - Recompile on reload: https://gwt-review.goo
 glesource.com/#/c/9323/ (dankurka)
  - Sending the wrong permutation to the client in SDM, if no files 
 have changed (dankurka).
  - Investigate why some people are seeing errors with incremental 
  not restricting to one permutation (dankurka).
  - Public directories are not copied o the war directory when using 
 SDM (skybrian).
  - Restore Java 6 compatibility (skybrian).
  - Document limitations of JsonUtils.safeEval and discourage usage 
 (goktug) (promote Json.parse)

 Patches that are nice to have:
  - Improve exception logging in SDM (goktug).

 *If you have any outstanding patches that you thing need to go into 
 GWT 2.7, please bring them to our attention, by replying to this 
 thread or 
 adding me as a reviewer on Gerrit and setting the topic to GWT2.7.*

 -Daniel

  -- 
>>> You received this message because you are subscribed to the Google 
>>> Groups "GWT Contributors" group.
>>> To unsubscribe from this group and stop receiving emails from it, 
>>> send an email to google-web-toolkit-contributors+unsubscribe@
>>> googlegroups.com.
>>> To view this discussion on the web visit 
>>> https://groups.google.com/d/msgid/google-web-toolkit-contrib
>>> utors/1d5a4369-e03d-47ba-b8dd-5031e5460751%40googlegroups.com 
>>> 
>>> .
>>>
>>> For more options, visit https://groups.google.com/d/optout.
>>>
>>
>>  -- 
> You received this message because you are subscribed to the Google 
> Groups "GWT Contributors" group.
> To unsubscribe from this group and stop receiving emails from it, send 
> an email to google-web-toolk

Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread Manuel Carrasco Moñino
There is no plans to upgrade Jetty. It implies some changes in
JettyLauncher and WebServer, I have already a patch for that but not
committed yet because I have not tested enough. If someone else demands it
we could consider to upgrade it.

Anyway I think it should be nice not to bundle jetty in the .jar deployed
in maven repos, but set it as a dependency so as it's easier to experiment
other versions.

On Fri, Oct 3, 2014 at 10:44 AM, Александр Новоселов 
wrote:

> Will you update to Jetty 9.2 (with JVM 8 support)?
> Any link to branch version?
>
> --
> You received this message because you are subscribed to the Google Groups
> "GWT Contributors" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/google-web-toolkit-contributors/ad0b7807-4da7-4a83-844d-dae11c3c786f%40googlegroups.com
> 
> .
>
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/CAM28XAvgNuq2rew1u57KTPwHKXe2jvkjR611nxuEUFOeCh-OpA%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread Manuel Carrasco Moñino
On Fri, Oct 3, 2014 at 4:05 PM, confile 
wrote:

> Well, I need to override HistoryImpl to enable HTML5 pushstate api.
>

HistoryImpl is now a private class inside History, maybe we could discuss
whether to change its visibility. Maybe @Daniel could help here.


>
>
> Am Freitag, 3. Oktober 2014 16:01:31 UTC+2 schrieb Manuel Carrasco Moñino:
>>
>> 1.- History mechanism has changed in 2.7 to fulfil modern standards
>> https://gwt-review.googlesource.com/#/c/5356/.  BTW: you should't use
>> HistoryImpl and in general any Impl class, in general they are private API
>> 2.- Use -XjsInteropMode  ONLY in the case that you have a need for
>> exporting gwt classes to JS or wrapping JS classes in gwt. Otherwise you
>> could have a increment of size and probably other issues. It is
>> experimental, even in 2.7.0, so don't use it for production or be prepared
>> to changes and breakages
>> 3.- Code sizes in 2.6 and 2.7 should be very similar, but there are
>> changes in many places, so little differences are acceptable. Update to a
>> new version does not necessary bring a reduction on code size, but other
>> improvements.
>>
>>
>>
>> On Fri, Oct 3, 2014 at 1:38 AM, confile 
>> wrote:
>>
>>> Code size explodes when using GWT 2.7 with -XjsInteropMode JS
>>>
>>> I compiled the same code with GWT 2.6.1 and GWT 2.7 with argument strict
>>>
>>> code size GWT 2.6.1 1.351 MB
>>> code size  GWT 2.7  1.361 MB
>>>
>>> Then I compiled the same code with -XjsInteropMode JS flag on GWT 2.7
>>>
>>> code size  1.601 MB
>>>
>>>
>>> Why does GWT 2.7 increases the code size in every case and with Interop
>>> even by 300 kb?
>>>
>>>
>>> Am Freitag, 3. Oktober 2014 00:18:22 UTC+2 schrieb confile:
>>>
 What is the difference between -XjsInteropMode JS and -XjsInteropMode
 CLOSURE?



 Am Donnerstag, 2. Oktober 2014 23:49:17 UTC+2 schrieb John Stalcup:
>
> -XjsInteropMode JS
>
> On Thu, Oct 2, 2014 at 2:46 PM, confile 
> wrote:
>
>> How can  JsInterop be activated in GWT 2.7? What flag does it?
>>
>>
>> Am Donnerstag, 2. Oktober 2014 03:22:56 UTC+2 schrieb Brian Slesinsky:
>>>
>>> It's experimental and hidden behind a flag, but it's there.
>>>
>>>
>>> On Wed, Oct 1, 2014 at 6:20 PM, Cristian Rinaldi 
>>> wrote:
>>>
 - JsInterop Preview is part of the release?

 El miércoles, 1 de octubre de 2014 16:15:26 UTC-3, Daniel Kurka
 escribió:

> Hi all,
>
> we just settled on a GWT 2.7 release plan:
>
>  - We *code freeze* on *October 7th* and branch for GWT 2.7.
>  - As soon as we have the *remaining patches submitted*, we put
> out a beta1 build, this should be no later than *October 7th.*
>  - Putting out a *beta1 externally* allows us to collect feedback
> on the new super dev mode integration externally as well.
>  - We are going to *flip incremental to default* tomorrow and *wait
> for 1-2 weeks* for google internal feedback, if there is no
> serious issues we are going to *put out RC1*
>  - GWT 2.7 will still be compatible with Java 6.
>
> Patches / Fixes that need to go in:
>  - Recompile on reload: https://gwt-review.goo
> glesource.com/#/c/9323/ (dankurka)
>  - Sending the wrong permutation to the client in SDM, if no files
> have changed (dankurka).
>  - Investigate why some people are seeing errors with incremental
>  not restricting to one permutation (dankurka).
>  - Public directories are not copied o the war directory when
> using SDM (skybrian).
>  - Restore Java 6 compatibility (skybrian).
>  - Document limitations of JsonUtils.safeEval and discourage usage
> (goktug) (promote Json.parse)
>
> Patches that are nice to have:
>  - Improve exception logging in SDM (goktug).
>
> *If you have any outstanding patches that you thing need to go
> into GWT 2.7, please bring them to our attention, by replying to this
> thread or adding me as a reviewer on Gerrit and setting the topic to
> GWT2.7.*
>
> -Daniel
>
>  --
 You received this message because you are subscribed to the Google
 Groups "GWT Contributors" group.
 To unsubscribe from this group and stop receiving emails from it,
 send an email to google-web-toolkit-contributors+unsubscribe@
 googlegroups.com.
 To view this discussion on the web visit
 https://groups.google.com/d/msgid/google-web-toolkit-contrib
 utors/1d5a4369-e03d-47ba-b8dd-5031e5460751%40googlegroups.com
 
 .

 For more options, visit https:

Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread Thomas Broyer


On Friday, October 3, 2014 4:09:33 PM UTC+2, Manuel Carrasco Moñino wrote:
>
>
>
> On Fri, Oct 3, 2014 at 4:05 PM, confile  
> wrote:
>
>> Well, I need to override HistoryImpl to enable HTML5 pushstate api.
>>
>
> HistoryImpl is now a private class inside History, maybe we could discuss 
> whether to change its visibility. Maybe @Daniel could help here.
>

When Daniel proposed the change, we decided it was OK to break 
compatibility with, say, Johannes Barop's gwt-pushstate. The rationale was 
that History was explicitly about using the "hash" part of the URL and if 
you wanted something else, or a pluggable implementation, you should use 
another API (Places are pluggable, and you could easily build another API 
similar to History but that would operate differently, whether pluggable or 
not).
AFAICT the whole discussion is in the review you linked to (or possibly 
another one, as there had been several tries IIRC)

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/63d381ce-ca34-4394-81c7-ea59b8f1e713%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread confile
Well if you do not use Places for example when you use GWTP you need the 
HistoryImpl to make GWT-pushstate working.



Am Freitag, 3. Oktober 2014 17:11:35 UTC+2 schrieb Thomas Broyer:
>
>
>
> On Friday, October 3, 2014 4:09:33 PM UTC+2, Manuel Carrasco Moñino wrote:
>>
>>
>>
>> On Fri, Oct 3, 2014 at 4:05 PM, confile > > wrote:
>>
>>> Well, I need to override HistoryImpl to enable HTML5 pushstate api.
>>>
>>
>> HistoryImpl is now a private class inside History, maybe we could discuss 
>> whether to change its visibility. Maybe @Daniel could help here.
>>
>
> When Daniel proposed the change, we decided it was OK to break 
> compatibility with, say, Johannes Barop's gwt-pushstate. The rationale was 
> that History was explicitly about using the "hash" part of the URL and if 
> you wanted something else, or a pluggable implementation, you should use 
> another API (Places are pluggable, and you could easily build another API 
> similar to History but that would operate differently, whether pluggable or 
> not).
> AFAICT the whole discussion is in the review you linked to (or possibly 
> another one, as there had been several tries IIRC)
>

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/d3babcef-5e2a-4a2e-a8a2-95a3dd88531f%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread Richard Wallis
Hi Thomas,

Please can you link to the discussion where you decided it was ok to break
compatibility with deferred binding implementations of the HistoryImpl.

On Fri, Oct 3, 2014 at 5:15 PM, confile 
wrote:

> Well if you do not use Places for example when you use GWTP you need the
> HistoryImpl to make GWT-pushstate working.
>
>
>
> Am Freitag, 3. Oktober 2014 17:11:35 UTC+2 schrieb Thomas Broyer:
>>
>>
>>
>> On Friday, October 3, 2014 4:09:33 PM UTC+2, Manuel Carrasco Moñino wrote:
>>>
>>>
>>>
>>> On Fri, Oct 3, 2014 at 4:05 PM, confile 
>>> wrote:
>>>
 Well, I need to override HistoryImpl to enable HTML5 pushstate api.

>>>
>>> HistoryImpl is now a private class inside History, maybe we could
>>> discuss whether to change its visibility. Maybe @Daniel could help here.
>>>
>>
>> When Daniel proposed the change, we decided it was OK to break
>> compatibility with, say, Johannes Barop's gwt-pushstate. The rationale was
>> that History was explicitly about using the "hash" part of the URL and if
>> you wanted something else, or a pluggable implementation, you should use
>> another API (Places are pluggable, and you could easily build another API
>> similar to History but that would operate differently, whether pluggable or
>> not).
>> AFAICT the whole discussion is in the review you linked to (or possibly
>> another one, as there had been several tries IIRC)
>>
>  --
> You received this message because you are subscribed to the Google Groups
> "GWT Contributors" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/google-web-toolkit-contributors/d3babcef-5e2a-4a2e-a8a2-95a3dd88531f%40googlegroups.com
> 
> .
>
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/CAEqaEVh9viEQY8mCMoRUt7%2BJLPawFRqdJEtNG9pCiCgP83MH9Q%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread confile
It would be great if you describe a workaround not having HistoryImpl and 
using gwt-pushstate when not using places.


Am Freitag, 3. Oktober 2014 17:25:10 UTC+2 schrieb Richard Wallis:
>
> Hi Thomas,
>
> Please can you link to the discussion where you decided it was ok to break 
> compatibility with deferred binding implementations of the HistoryImpl.
>
> On Fri, Oct 3, 2014 at 5:15 PM, confile  > wrote:
>
>> Well if you do not use Places for example when you use GWTP you need the 
>> HistoryImpl to make GWT-pushstate working.
>>
>>
>>
>> Am Freitag, 3. Oktober 2014 17:11:35 UTC+2 schrieb Thomas Broyer:
>>>
>>>
>>>
>>> On Friday, October 3, 2014 4:09:33 PM UTC+2, Manuel Carrasco Moñino 
>>> wrote:



 On Fri, Oct 3, 2014 at 4:05 PM, confile  
 wrote:

> Well, I need to override HistoryImpl to enable HTML5 pushstate api.
>

 HistoryImpl is now a private class inside History, maybe we could 
 discuss whether to change its visibility. Maybe @Daniel could help here.

>>>
>>> When Daniel proposed the change, we decided it was OK to break 
>>> compatibility with, say, Johannes Barop's gwt-pushstate. The rationale was 
>>> that History was explicitly about using the "hash" part of the URL and if 
>>> you wanted something else, or a pluggable implementation, you should use 
>>> another API (Places are pluggable, and you could easily build another API 
>>> similar to History but that would operate differently, whether pluggable or 
>>> not).
>>> AFAICT the whole discussion is in the review you linked to (or possibly 
>>> another one, as there had been several tries IIRC)
>>>
>>  -- 
>> You received this message because you are subscribed to the Google Groups 
>> "GWT Contributors" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to google-web-toolkit-contributors+unsubscr...@googlegroups.com 
>> .
>> To view this discussion on the web visit 
>> https://groups.google.com/d/msgid/google-web-toolkit-contributors/d3babcef-5e2a-4a2e-a8a2-95a3dd88531f%40googlegroups.com
>>  
>> 
>> .
>>
>> For more options, visit https://groups.google.com/d/optout.
>>
>
>

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/9f813310-ff9a-4099-8d67-214e74cd2043%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread Jens

>
> Hi Thomas,
>
> Please can you link to the discussion where you decided it was ok to break 
> compatibility with deferred binding implementations of the HistoryImpl.
>

See comments in https://gwt-review.googlesource.com/#/c/6403/ 

-- J.

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/b2d70738-bb2d-4eed-a6ea-8b5da5f08a78%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread Richard Wallis
Ok so gwt 2.7 is going to provide pushstate functionality so we don't need
the 3rd party libraries?

On Fri, Oct 3, 2014 at 5:47 PM, Jens  wrote:

> Hi Thomas,
>>
>> Please can you link to the discussion where you decided it was ok to
>> break compatibility with deferred binding implementations of the
>> HistoryImpl.
>>
>
> See comments in https://gwt-review.googlesource.com/#/c/6403/
>
> -- J.
>
> --
> You received this message because you are subscribed to the Google Groups
> "GWT Contributors" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/google-web-toolkit-contributors/b2d70738-bb2d-4eed-a6ea-8b5da5f08a78%40googlegroups.com
> 
> .
>
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/CAEqaEVj_mNkvgqVmvp3SMeNjRH9mg%2BGq_%3DABtLiDSS93w3maTg%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread Jens

>
> It would be great if you describe a workaround not having HistoryImpl and 
> using gwt-pushstate when not using places.
>

There is no real workaround other than building a class on your own that 
either uses GWT History for hash fragments or pushState if available. With 
GWT Places that would be an implementation of Historian interface.

Alternative would be to change GWT to make gwt-pushstate library work again.

-- J.

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/b124d76c-f6e0-4ca1-8255-9bb6772b7bd5%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread Jens

>
> Ok so gwt 2.7 is going to provide pushstate functionality so we don't need 
> the 3rd party libraries?
>

 It only uses pushState internally and does not expose any API for it. So 
no, you would still need a library or JSNI to use pushState.

-- J.

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/7e26d28b-15cf-497d-b8ae-1deaaf2d7c51%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread Richard Wallis
Can you not define an interface for HistoryImpl so that we can replace it
while you still keep your implementation private?

On Fri, Oct 3, 2014 at 5:52 PM, Jens  wrote:

> Ok so gwt 2.7 is going to provide pushstate functionality so we don't need
>> the 3rd party libraries?
>>
>
>  It only uses pushState internally and does not expose any API for it. So
> no, you would still need a library or JSNI to use pushState.
>
> -- J.
>
> --
> You received this message because you are subscribed to the Google Groups
> "GWT Contributors" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/google-web-toolkit-contributors/7e26d28b-15cf-497d-b8ae-1deaaf2d7c51%40googlegroups.com
> 
> .
>
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/CAEqaEViROMyjfNJSgR%3Ds4JPSNY5v0fz4Tot33ArTZf6gkA2bfA%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread confile
I do support Richard's idea.

Am Freitag, 3. Oktober 2014 18:19:30 UTC+2 schrieb Richard Wallis:
>
> Can you not define an interface for HistoryImpl so that we can replace it 
> while you still keep your implementation private?
>
> On Fri, Oct 3, 2014 at 5:52 PM, Jens > 
> wrote:
>
>> Ok so gwt 2.7 is going to provide pushstate functionality so we don't 
>>> need the 3rd party libraries?
>>>
>>
>>  It only uses pushState internally and does not expose any API for it. So 
>> no, you would still need a library or JSNI to use pushState.
>>
>> -- J.
>>
>> -- 
>> You received this message because you are subscribed to the Google Groups 
>> "GWT Contributors" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to google-web-toolkit-contributors+unsubscr...@googlegroups.com 
>> .
>> To view this discussion on the web visit 
>> https://groups.google.com/d/msgid/google-web-toolkit-contributors/7e26d28b-15cf-497d-b8ae-1deaaf2d7c51%40googlegroups.com
>>  
>> 
>> .
>>
>> For more options, visit https://groups.google.com/d/optout.
>>
>
>

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/306ab4f6-f0d8-4d56-b889-21f4531b6550%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread Thomas Broyer


On Friday, October 3, 2014 6:19:30 PM UTC+2, Richard Wallis wrote:
>
> Can you not define an interface for HistoryImpl so that we can replace it 
> while you still keep your implementation private?
>

No, because we specifically don't want to make it a "public API" where we 
have to maintain backwards compatibility.

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/e407e43b-7224-49be-81e7-7ce791130959%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread Thomas Broyer


On Friday, October 3, 2014 5:15:51 PM UTC+2, confile wrote:
>
> Well if you do not use Places for example when you use GWTP you need the 
> HistoryImpl to make GWT-pushstate working.
>

As far as I can see form the code, GWTP's PlaceManager is pluggable 
(sort-of) and could probably replaced with a version that uses pushstate 
(in JSNI) without using the URL's #hash
 

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/bc00da33-ccad-41f3-8d06-64169e9c0d8c%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread confile
Okay I get your point. Could you please provide some kind of workaround?

Thank you
Michael

Am Freitag, 3. Oktober 2014 18:26:18 UTC+2 schrieb Thomas Broyer:
>
>
>
> On Friday, October 3, 2014 5:15:51 PM UTC+2, confile wrote:
>>
>> Well if you do not use Places for example when you use GWTP you need the 
>> HistoryImpl to make GWT-pushstate working.
>>
>
> As far as I can see form the code, GWTP's PlaceManager is pluggable 
> (sort-of) and could probably replaced with a version that uses pushstate 
> (in JSNI) without using the URL's #hash
>  
>

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/7574cb49-f169-431a-ad26-789e0544edda%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread Thomas Broyer


On Friday, October 3, 2014 4:06:01 PM UTC+2, Manuel Carrasco Moñino wrote:
>
> There is no plans to upgrade Jetty. It implies some changes in 
> JettyLauncher and WebServer, I have already a patch for that but not 
> committed yet because I have not tested enough. If someone else demands it 
> we could consider to upgrade it.
>

AFAIK, we can't update Jetty to 9.x because HTMLUnit (used for 
JUnitShell/GWTTestCase) depends on Jetty 8.x APIs (WebSockets)
There's been a patch proposed already and it has been abandoned because of 
that.
 

> Anyway I think it should be nice not to bundle jetty in the .jar deployed 
> in maven repos, but set it as a dependency so as it's easier to experiment 
> other versions.
>

It could possibly help with some setups but I doubt it'd be that useful.

If we had a modularized GWT then you could avoid depending on gwt-junit, 
gwt-devmode-jettylauncher and gwt-codeserver (e.g. when compiling) and live 
with Jetty 9.x dependencies in your classpath/buildpath, or we could 
possibly provide a gwt-codeserver and/or gwt-devmode-jettylauncher using 
using Jetty 9.x; so you'd use Jetty 8.x with gwt-junit to run your 
unittests, and you could use Jetty 9.x in DevMode/superDevMode (and benefit 
from Jetty 9.x perf improvements for gwt-codeserver?)
It's not possible with our current monolithic artifacts (and build) though.

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/d22e2d8b-0bf1-4f17-9805-9fdb944076f5%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread Richard Wallis
Ok the Historian interface looks like it will solve our issues.  I'll
rewrite GWTP's PlaceManagerImpl to use it and update my pushstate
implementation.

On Fri, Oct 3, 2014 at 6:34 PM, Thomas Broyer  wrote:

>
>
> On Friday, October 3, 2014 4:06:01 PM UTC+2, Manuel Carrasco Moñino wrote:
>>
>> There is no plans to upgrade Jetty. It implies some changes in
>> JettyLauncher and WebServer, I have already a patch for that but not
>> committed yet because I have not tested enough. If someone else demands it
>> we could consider to upgrade it.
>>
>
> AFAIK, we can't update Jetty to 9.x because HTMLUnit (used for
> JUnitShell/GWTTestCase) depends on Jetty 8.x APIs (WebSockets)
> There's been a patch proposed already and it has been abandoned because of
> that.
>
>
>> Anyway I think it should be nice not to bundle jetty in the .jar deployed
>> in maven repos, but set it as a dependency so as it's easier to experiment
>> other versions.
>>
>
> It could possibly help with some setups but I doubt it'd be that useful.
>
> If we had a modularized GWT then you could avoid depending on gwt-junit,
> gwt-devmode-jettylauncher and gwt-codeserver (e.g. when compiling) and live
> with Jetty 9.x dependencies in your classpath/buildpath, or we could
> possibly provide a gwt-codeserver and/or gwt-devmode-jettylauncher using
> using Jetty 9.x; so you'd use Jetty 8.x with gwt-junit to run your
> unittests, and you could use Jetty 9.x in DevMode/superDevMode (and benefit
> from Jetty 9.x perf improvements for gwt-codeserver?)
> It's not possible with our current monolithic artifacts (and build) though.
>
> --
> You received this message because you are subscribed to the Google Groups
> "GWT Contributors" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/google-web-toolkit-contributors/d22e2d8b-0bf1-4f17-9805-9fdb944076f5%40googlegroups.com
> 
> .
>
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/CAEqaEVhctL2_UevaFdJAXHjomfqyYQLK7k1zTREWZStLzdv6vg%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread Manuel Carrasco Moñino
Thanks for the info Thomas, I din't  realise that patch.

On Fri, Oct 3, 2014 at 6:34 PM, Thomas Broyer  wrote:

>
>
> On Friday, October 3, 2014 4:06:01 PM UTC+2, Manuel Carrasco Moñino wrote:
>>
>> There is no plans to upgrade Jetty. It implies some changes in
>> JettyLauncher and WebServer, I have already a patch for that but not
>> committed yet because I have not tested enough. If someone else demands it
>> we could consider to upgrade it.
>>
>
> AFAIK, we can't update Jetty to 9.x because HTMLUnit (used for
> JUnitShell/GWTTestCase) depends on Jetty 8.x APIs (WebSockets)
> There's been a patch proposed already and it has been abandoned because of
> that.
>
>
>> Anyway I think it should be nice not to bundle jetty in the .jar deployed
>> in maven repos, but set it as a dependency so as it's easier to experiment
>> other versions.
>>
>
> It could possibly help with some setups but I doubt it'd be that useful.
>
> If we had a modularized GWT then you could avoid depending on gwt-junit,
> gwt-devmode-jettylauncher and gwt-codeserver (e.g. when compiling) and live
> with Jetty 9.x dependencies in your classpath/buildpath, or we could
> possibly provide a gwt-codeserver and/or gwt-devmode-jettylauncher using
> using Jetty 9.x; so you'd use Jetty 8.x with gwt-junit to run your
> unittests, and you could use Jetty 9.x in DevMode/superDevMode (and benefit
> from Jetty 9.x perf improvements for gwt-codeserver?)
> It's not possible with our current monolithic artifacts (and build) though.
>
> --
> You received this message because you are subscribed to the Google Groups
> "GWT Contributors" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/google-web-toolkit-contributors/d22e2d8b-0bf1-4f17-9805-9fdb944076f5%40googlegroups.com
> 
> .
>
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/CAM28XAvKYKh5CYZGrU%2B7NgC8TR1Lnd%2BikpiBZjq1NMv13M%2BDGQ%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread confile
@Richard Wallis Could you please publish your updated pushstate 
implementation?

Thank you 
Michael

Am Freitag, 3. Oktober 2014 18:51:51 UTC+2 schrieb Richard Wallis:
>
> Ok the Historian interface looks like it will solve our issues.  I'll 
> rewrite GWTP's PlaceManagerImpl to use it and update my pushstate 
> implementation.
>
> On Fri, Oct 3, 2014 at 6:34 PM, Thomas Broyer  > wrote:
>
>>
>>
>> On Friday, October 3, 2014 4:06:01 PM UTC+2, Manuel Carrasco Moñino wrote:
>>>
>>> There is no plans to upgrade Jetty. It implies some changes in 
>>> JettyLauncher and WebServer, I have already a patch for that but not 
>>> committed yet because I have not tested enough. If someone else demands it 
>>> we could consider to upgrade it.
>>>
>>
>> AFAIK, we can't update Jetty to 9.x because HTMLUnit (used for 
>> JUnitShell/GWTTestCase) depends on Jetty 8.x APIs (WebSockets)
>> There's been a patch proposed already and it has been abandoned because 
>> of that.
>>  
>>
>>> Anyway I think it should be nice not to bundle jetty in the .jar 
>>> deployed in maven repos, but set it as a dependency so as it's easier to 
>>> experiment other versions.
>>>
>>
>> It could possibly help with some setups but I doubt it'd be that useful.
>>
>> If we had a modularized GWT then you could avoid depending on gwt-junit, 
>> gwt-devmode-jettylauncher and gwt-codeserver (e.g. when compiling) and live 
>> with Jetty 9.x dependencies in your classpath/buildpath, or we could 
>> possibly provide a gwt-codeserver and/or gwt-devmode-jettylauncher using 
>> using Jetty 9.x; so you'd use Jetty 8.x with gwt-junit to run your 
>> unittests, and you could use Jetty 9.x in DevMode/superDevMode (and benefit 
>> from Jetty 9.x perf improvements for gwt-codeserver?)
>> It's not possible with our current monolithic artifacts (and build) 
>> though.
>>
>> -- 
>> You received this message because you are subscribed to the Google Groups 
>> "GWT Contributors" group.
>> To unsubscribe from this group and stop receiving emails from it, send an 
>> email to google-web-toolkit-contributors+unsubscr...@googlegroups.com 
>> .
>> To view this discussion on the web visit 
>> https://groups.google.com/d/msgid/google-web-toolkit-contributors/d22e2d8b-0bf1-4f17-9805-9fdb944076f5%40googlegroups.com
>>  
>> 
>> .
>>
>> For more options, visit https://groups.google.com/d/optout.
>>
>
>

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/2560e5b0-6f75-4580-b75b-435bd04296b1%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Re: GWT 2.7 release plan

2014-10-03 Thread Richard Wallis
Ok, I'll be publishing it here: https://github.com/rdwallis/gwt-pushstate
probably early next week.

On Fri, Oct 3, 2014 at 6:53 PM, confile 
wrote:

> @Richard Wallis Could you please publish your updated pushstate
> implementation?
>
> Thank you
> Michael
>
> Am Freitag, 3. Oktober 2014 18:51:51 UTC+2 schrieb Richard Wallis:
>>
>> Ok the Historian interface looks like it will solve our issues.  I'll
>> rewrite GWTP's PlaceManagerImpl to use it and update my pushstate
>> implementation.
>>
>> On Fri, Oct 3, 2014 at 6:34 PM, Thomas Broyer  wrote:
>>
>>>
>>>
>>> On Friday, October 3, 2014 4:06:01 PM UTC+2, Manuel Carrasco Moñino
>>> wrote:

 There is no plans to upgrade Jetty. It implies some changes in
 JettyLauncher and WebServer, I have already a patch for that but not
 committed yet because I have not tested enough. If someone else demands it
 we could consider to upgrade it.

>>>
>>> AFAIK, we can't update Jetty to 9.x because HTMLUnit (used for
>>> JUnitShell/GWTTestCase) depends on Jetty 8.x APIs (WebSockets)
>>> There's been a patch proposed already and it has been abandoned because
>>> of that.
>>>
>>>
 Anyway I think it should be nice not to bundle jetty in the .jar
 deployed in maven repos, but set it as a dependency so as it's easier to
 experiment other versions.

>>>
>>> It could possibly help with some setups but I doubt it'd be that useful.
>>>
>>> If we had a modularized GWT then you could avoid depending on gwt-junit,
>>> gwt-devmode-jettylauncher and gwt-codeserver (e.g. when compiling) and live
>>> with Jetty 9.x dependencies in your classpath/buildpath, or we could
>>> possibly provide a gwt-codeserver and/or gwt-devmode-jettylauncher using
>>> using Jetty 9.x; so you'd use Jetty 8.x with gwt-junit to run your
>>> unittests, and you could use Jetty 9.x in DevMode/superDevMode (and benefit
>>> from Jetty 9.x perf improvements for gwt-codeserver?)
>>> It's not possible with our current monolithic artifacts (and build)
>>> though.
>>>
>>> --
>>> You received this message because you are subscribed to the Google
>>> Groups "GWT Contributors" group.
>>> To unsubscribe from this group and stop receiving emails from it, send
>>> an email to google-web-toolkit-contributors+unsubscr...@googlegroups.com
>>> .
>>> To view this discussion on the web visit https://groups.google.com/d/
>>> msgid/google-web-toolkit-contributors/d22e2d8b-0bf1-
>>> 4f17-9805-9fdb944076f5%40googlegroups.com
>>> 
>>> .
>>>
>>> For more options, visit https://groups.google.com/d/optout.
>>>
>>
>>  --
> You received this message because you are subscribed to the Google Groups
> "GWT Contributors" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/google-web-toolkit-contributors/2560e5b0-6f75-4580-b75b-435bd04296b1%40googlegroups.com
> 
> .
>
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/CAEqaEVjcfPpgeg5uCOkz__TTFux8z2JVQpyN2eVCX_W%2BYfOaCA%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] Removal of IE6/7 specific code from the code base

2014-10-03 Thread Michael Vogt
Sorry for being away so long. I will have some time next week to work on this.

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/CAAfA4FzE900SfZVU9pVK-0XJ6ccojYubQpohmPYBcTzEJd6c9A%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.


[gwt-contrib] class cast exception in ff 17

2014-10-03 Thread Stephen Haberman
Hey,

I know FF17 is pretty old, but do we care about breakages in it? We
noticed some client-side ClassCastExceptions in our logs that are
from a user on FF17. The app works fine in FF latest.

The exception is odd, because, according to the de-obfuscated stack
trace, given a class hierarchy like:

   class Base
   class Foo extends Base
   class Bar extends Base

We are in "new Foo()", from the Foo cstr we call the Base cstr, which
calls Base.someMethod(). Base.someMethod() calls Base.anotherMethod().

Now the bizarre thing is that we end up in *Bar*.anotherMethod(), and
get a class cast exception.

E.g. somehow we jumped subclasses, and went from executing code in
Foo/Foo's base class to code that is in Bar.

Which seems really odd. Which is why I mention it. If we don't care,
that's fine, but it made me scratch my head for awhile.

(Our production build is using a GWT snapshot from ~April, but I also
reproduced it in our app with a GWT snapshot build from ~last week.)

- Stephen

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/20141003135259.76c16354%40sh9.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] class cast exception in ff 17

2014-10-03 Thread 'Roberto Lublinerman' via GWT Contributors
Can you isolate the bug and open an issue for it? If you upload a repro
case I'll take a look.

On Fri, Oct 3, 2014 at 11:52 AM, Stephen Haberman <
stephen.haber...@gmail.com> wrote:

> Hey,
>
> I know FF17 is pretty old, but do we care about breakages in it? We
> noticed some client-side ClassCastExceptions in our logs that are
> from a user on FF17. The app works fine in FF latest.
>
> The exception is odd, because, according to the de-obfuscated stack
> trace, given a class hierarchy like:
>
>class Base
>class Foo extends Base
>class Bar extends Base
>
> We are in "new Foo()", from the Foo cstr we call the Base cstr, which
> calls Base.someMethod(). Base.someMethod() calls Base.anotherMethod().
>
> Now the bizarre thing is that we end up in *Bar*.anotherMethod(), and
> get a class cast exception.
>
> E.g. somehow we jumped subclasses, and went from executing code in
> Foo/Foo's base class to code that is in Bar.
>
> Which seems really odd. Which is why I mention it. If we don't care,
> that's fine, but it made me scratch my head for awhile.
>
> (Our production build is using a GWT snapshot from ~April, but I also
> reproduced it in our app with a GWT snapshot build from ~last week.)
>
> - Stephen
>
> --
> You received this message because you are subscribed to the Google Groups
> "GWT Contributors" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/google-web-toolkit-contributors/20141003135259.76c16354%40sh9
> .
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/CAC7T7gkXtB3Ky3%3DFY11M0_E8MERpU0tKhkd2-C71wncG4RpTsg%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] class cast exception in ff 17

2014-10-03 Thread 'Goktug Gokdogan' via GWT Contributors
It might be related to the bug[1] that mdempsky@ filed to Firefox earlier.
The Firefox's VM was causing an incorrect polymorphic dispatch sometimes.

[1] https://bugzilla.mozilla.org/show_bug.cgi?id=872796

On Fri, Oct 3, 2014 at 11:52 AM, Stephen Haberman <
stephen.haber...@gmail.com> wrote:

> Hey,
>
> I know FF17 is pretty old, but do we care about breakages in it? We
> noticed some client-side ClassCastExceptions in our logs that are
> from a user on FF17. The app works fine in FF latest.
>
> The exception is odd, because, according to the de-obfuscated stack
> trace, given a class hierarchy like:
>
>class Base
>class Foo extends Base
>class Bar extends Base
>
> We are in "new Foo()", from the Foo cstr we call the Base cstr, which
> calls Base.someMethod(). Base.someMethod() calls Base.anotherMethod().
>
> Now the bizarre thing is that we end up in *Bar*.anotherMethod(), and
> get a class cast exception.
>
> E.g. somehow we jumped subclasses, and went from executing code in
> Foo/Foo's base class to code that is in Bar.
>
> Which seems really odd. Which is why I mention it. If we don't care,
> that's fine, but it made me scratch my head for awhile.
>
> (Our production build is using a GWT snapshot from ~April, but I also
> reproduced it in our app with a GWT snapshot build from ~last week.)
>
> - Stephen
>
> --
> You received this message because you are subscribed to the Google Groups
> "GWT Contributors" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/google-web-toolkit-contributors/20141003135259.76c16354%40sh9
> .
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/CAN%3DyUA008%2B%3DwTYRP3gU4f4hWOBLFXEz-HUrK%3DGxiYgUtKGXNxg%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.


[gwt-contrib] Re: class cast exception in ff 17

2014-10-03 Thread Jens
I have also seen this behavior once in our app. If I remember correctly the 
generated JavaScript looked correct, yet FireFox called the wrong method. 

IMHO it is a JIT bug in older Firefox versions.

-- J.

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/383ce5eb-770d-4b6a-994c-257989fa98de%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] class cast exception in ff 17

2014-10-03 Thread Stephen Haberman

> It might be related to the bug[1] that mdempsky@ filed to Firefox
> earlier. The Firefox's VM was causing an incorrect polymorphic
> dispatch sometimes.
> 
> [1] https://bugzilla.mozilla.org/show_bug.cgi?id=872796

Ah, nice!

Thanks (Jens too) for saving me the trouble of trying to reproduce
that. I was really not looking forward to it. :-)

- Stephen

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/20141003141714.51464061%40sh9.
For more options, visit https://groups.google.com/d/optout.


Re: [gwt-contrib] GWT 2.7 release plan

2014-10-03 Thread Colin Alworth
https://code.google.com/p/google-web-toolkit/issues/detail?id=8716 is still 
open (and is a regression over GWT 2.6 and previous). Is this a reasonable 
item to get fixed?

On Wednesday, October 1, 2014 2:23:43 PM UTC-5, Brian Slesinsky wrote:
>
> - Make sure sample apps work with DevMode -superdevmode
> - I think we're waiting on a patch to CLDR 25 
>
>
> On Wed, Oct 1, 2014 at 12:15 PM, 'Daniel Kurka' via GWT Contributors  
> wrote:
>
>> Hi all,
>>
>> we just settled on a GWT 2.7 release plan:
>>
>>  - We *code freeze* on *October 7th* and branch for GWT 2.7.
>>  - As soon as we have the *remaining patches submitted*, we put out a 
>> beta1 build, this should be no later than *October 7th.*
>>  - Putting out a *beta1 externally* allows us to collect feedback on the 
>> new super dev mode integration externally as well.
>>  - We are going to *flip incremental to default* tomorrow and *wait for 
>> 1-2 weeks* for google internal feedback, if there is no serious issues 
>> we are going to *put out RC1*
>>  - GWT 2.7 will still be compatible with Java 6.
>>
>> Patches / Fixes that need to go in:
>>  - Recompile on reload: https://gwt-review.googlesource.com/#/c/9323/
>>  (dankurka)
>>  - Sending the wrong permutation to the client in SDM, if no files have 
>> changed (dankurka).
>>  - Investigate why some people are seeing errors with incremental  not 
>> restricting to one permutation (dankurka).
>>  - Public directories are not copied o the war directory when using SDM 
>> (skybrian).
>>  - Restore Java 6 compatibility (skybrian).
>>  - Document limitations of JsonUtils.safeEval and discourage usage 
>> (goktug) (promote Json.parse)
>>
>> Patches that are nice to have:
>>  - Improve exception logging in SDM (goktug).
>>
>> *If you have any outstanding patches that you thing need to go into GWT 
>> 2.7, please bring them to our attention, by replying to this thread or 
>> adding me as a reviewer on Gerrit and setting the topic to GWT2.7.*
>>
>> -Daniel
>>
>>  
>

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/0b3edc35-cf55-4096-837c-fb4bcbbc4d32%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


[gwt-contrib] Running TestCase from IDEA

2014-10-03 Thread Danilo Reinert
I've already set-up the gwt project in Intellij, excluding the super 
sources, and I can compile with success.
But some testcases I can run from the IDE without any problem. Others, 
there's a classpath problem.
See this log:

Tracing compile failure path for type 
'org.hibernate.jsr303.tck.tests.constraints.application.ValidationRequirementGwtTest'
   [ERROR] Errors in 
'file:/home/reinert/gwt/trunk/user/test/org/hibernate/jsr303/tck/tests/constraints/application/ValidationRequirementGwtTest.java'
  [ERROR] 
org.hibernate.jsr303.tck.tests.constraints.application.ValidationRequirementTest
 
cannot be resolved to a type
   [ERROR] Errors in 
'org/hibernate/jsr303/tck/tests/constraints/application/ValidationRequirementTest.java'
  [ERROR] Line 45: No source code is available for type 
org.jboss.testharness.AbstractTest; did you forget to inherit a required 
module?
  [ERROR] Line 59: No source code is available for type 
org.hibernate.jsr303.tck.util.TestUtil; did you forget to inherit a 
required module?

com.google.gwt.junit.JUnitFatalLaunchException: The test class 
'org.hibernate.jsr303.tck.tests.constraints.application.ValidationRequirementGwtTest'
 
had compile errors; check log for details
...

I've already tried many things to add these missing classes to the path of 
the running test, but I had no success.
Has anyone solved a problem like that?

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/ae462ba7-fb22-4e47-9167-6cc74d5d4006%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.