Play 2.1.2-RC2 released

1,354 views
Skip to first unread message

James Roper

unread,
Jun 28, 2013, 6:57:55 PM6/28/13
to play-framework
We are pleased to announce the release of Play 2.1.2-RC2.

This release should be the final RC for 2.1.2, and contains a number of fixes for regressions that were introduced in Play 2.1.2-RC2, as well as a few other bugs.  Specifically:

#1248 Fixed regression in Akka configuration loading 
#1229 #820 Accept-Language header fixes 
#1152 Fixed broken 100 continue support 
#984 Honour config.resource property in dev mode 
#959 Results stream/feed now automatically sends EOF 
#924 Quoted charset in content types now works 
#851 Context classloader in Java actions now set correctly 

These fixes are in addition to the issues fixed in 2.1.2-RC1: 

#1110 #1090 #1050 Performance improvements 
#978 URL path encoding fixes 
#945 HTTP pipelining fixes 
#839 Ensured Play initialisation happens in a non netty thread in dev mode 
#810 XPath regression fixes 


The release can be downloaded from:


Thankyou particularly to all those that contributed to this release.

Enjoy!

James

--
James Roper
Software Engineer

Typesafe – Empowering professional developers to build amazing apps.
Twitter: @jroper

Georgee

unread,
Jun 30, 2013, 5:51:23 AM6/30/13
to play-fr...@googlegroups.com
Is it use Scala 2.10.0? 2.10.2 is released already :)

суббота, 29 июня 2013 г., 5:57:55 UTC+7 пользователь James Roper написал:

James Roper

unread,
Jun 30, 2013, 7:36:11 PM6/30/13
to play-framework
Hi Georgee,

It's very simple to use Scala 2.10.2 with a Play 2.1.x project, playframework.com is using Play 2.1.2-RC2 and running on Scala 2.10.2 so it definitely works.  Just set:

scalaVersion := "2.10.2"

in your Build.scala file.  Unless we find Play is critically impacted by a bug in a library in Scala, we generally won't upgrade it in a stable version, but there's nothing stopping you from upgrading it in your project.

Cheers,

James


--
You received this message because you are subscribed to the Google Groups "play-framework" group.
To unsubscribe from this group and stop receiving emails from it, send an email to play-framewor...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

Steven Marcus

unread,
Jun 30, 2013, 8:25:48 PM6/30/13
to play-fr...@googlegroups.com

Setting scalaVersion in your own builds results in "mixed" org.scala-lang dependencies.

That is, scala-library 2.10.2 is included in your project, but scala-actors and scala-reflect, pulled in by Play, are still 2.10 unless you take more steps to explicitly update those deps.

Is running with mixed org.scala-lang deps recommended?


Thx

Simon Ochsenreither

unread,
Jul 2, 2013, 7:00:31 AM7/2/13
to play-fr...@googlegroups.com

Unless we find Play is critically impacted by a bug in a library in Scala, we generally won't upgrade it in a stable version, but there's nothing stopping you from upgrading it in your project.

Considering that Scala is now both backward and forward compatible in minor/update versions, wouldn't it be reasonable to update the Scala dependencies once, instead of letting every user of Play do it?
Imho, looking at all the fixed compiler crashes in 2.10.1 and 2.10.2, it would be a worthwhile update.

Eduardo M. Cavalcanti

unread,
Jul 2, 2013, 1:46:43 PM7/2/13
to play-fr...@googlegroups.com


Unless we find Play is critically impacted by a bug in a library in Scala, we generally won't upgrade it in a stable version, but there's nothing stopping you from upgrading it in your project.

Considering that Scala is now both backward and forward compatible in minor/update versions, wouldn't it be reasonable to update the Scala dependencies once, instead of letting every user of Play do it?
Imho, looking at all the fixed compiler crashes in 2.10.1 and 2.10.2, it would be a worthwhile update.

+1
Thanks for the great work.

Ben McCann

unread,
Jul 2, 2013, 11:00:05 PM7/2/13
to play-fr...@googlegroups.com
I think I've found a regression. In 2.1.0 if I visited a page like locahost:9000/test/my test then "my test" would be passed as the parameter to the controller and now it's "my%20test". Would be great if we could get a fix for that for the next release. Should I file an issue on GitHub?

Super excited for a couple of the fixes in 2.1.2. Thanks for working on it!!

Marconi Lanna

unread,
Jul 4, 2013, 2:01:50 AM7/4/13
to play-fr...@googlegroups.com
I understand one wouldn't want to upgrade to a major Scala release (say, 2.10 to 2.11) in a minor Play release, but it makes little sense not to upgrade to a minor release (2.10.2), especially now that Scala (and Typesafe) have guaranteed cross compatibility.

It doesn't inspire much confidence that a company like Typesafe would allow such inconsistencies across their core products. You are sending mixed messages here: "2.10.2 is fully compatible with 2.10, we guarantee you. We are only too afraid to upgrade our own products ourselves, though."

Eat the latest version of your own dog food.

James Roper

unread,
Jul 4, 2013, 3:20:23 AM7/4/13
to play-framework
On Tue, Jul 2, 2013 at 8:00 PM, Ben McCann <benjamin...@gmail.com> wrote:
I think I've found a regression. In 2.1.0 if I visited a page like locahost:9000/test/my test then "my test" would be passed as the parameter to the controller and now it's "my%20test". Would be great if we could get a fix for that for the next release. Should I file an issue on GitHub?

Is this for :param style params or *param style params?  If it's for *param style params, my%20test is expected, we can't decode it otherwise there is no way for a controller to distinguish between / and %2F.
 

Super excited for a couple of the fixes in 2.1.2. Thanks for working on it!!


On Tuesday, July 2, 2013 10:46:43 AM UTC-7, Eduardo M. Cavalcanti wrote:


Unless we find Play is critically impacted by a bug in a library in Scala, we generally won't upgrade it in a stable version, but there's nothing stopping you from upgrading it in your project.

Considering that Scala is now both backward and forward compatible in minor/update versions, wouldn't it be reasonable to update the Scala dependencies once, instead of letting every user of Play do it?
Imho, looking at all the fixed compiler crashes in 2.10.1 and 2.10.2, it would be a worthwhile update.

+1
Thanks for the great work.

--
You received this message because you are subscribed to the Google Groups "play-framework" group.
To unsubscribe from this group and stop receiving emails from it, send an email to play-framewor...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

James Roper

unread,
Jul 4, 2013, 3:29:27 AM7/4/13
to play-framework
It's got nothing to do with being afraid of upgrading, and everything to do with drawing a line to limit the amount of time that we have to spend maintaining old releases and get on with moving Play forward.

Improvements, upgrades etc happen in the master branch.  Only bugfixes and other changes that are having a serious impact on users get backported to stable releases.  Scala is not treated specially here, why should we treat it specially?  It can be upgraded by the user just as easily as any other library.
 

--
You received this message because you are subscribed to the Google Groups "play-framework" group.
To unsubscribe from this group and stop receiving emails from it, send an email to play-framewor...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.


Ben McCann

unread,
Jul 4, 2013, 3:32:21 AM7/4/13
to play-fr...@googlegroups.com

It was for the :param style

You received this message because you are subscribed to a topic in the Google Groups "play-framework" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/play-framework/esgvgNcJfl8/unsubscribe.
To unsubscribe from this group and all its topics, send an email to play-framewor...@googlegroups.com.

James Roper

unread,
Jul 4, 2013, 8:59:24 PM7/4/13
to play-framework
Hi Ben,

I wrote a test to confirm this, and couldn't reproduce.  The following route:

GET     /foo/:bar                   controllers.Application.foo(bar)

And the following action:

def foo(bar: String) = Action {
  Ok("Decoded: " + bar + " Encoded: " + routes.Application.foo(bar))
}

Show that space is being correctly decoded.  Interestingly, it shows that space is being incorrectly encoded as a + (this is only valid for query strings).  + is, correctly, not being decoded to a space.

So there is a bug, but it also existed in 2.1.1.  The bug didn't exist in 2.1.0, but another bug did exist there, where + was being decoded as a space.

At any rate, since this isn't a regression from 2.1.1, and there are no other issues reported with 2.1.2-RC2, we'll release the RC as 2.1.2, and schedule this other bug to be fixed in 2.1.3.

Regards,

James
Message has been deleted

rudy

unread,
Jul 4, 2013, 10:23:36 PM7/4/13
to play-fr...@googlegroups.com
HI James, Ben, 

I can confirm that this bug exists as I am hitting it as well. You are right about it not happening in 2.1.0 but I am able to reproduce it in 2.1.2-RC2. 

In 2.1.0 however, the character + is omitted when the param reaches the action. Not sure if this is a known issue.

James Roper

unread,
Jul 5, 2013, 1:56:25 AM7/5/13
to play-framework
On Thu, Jul 4, 2013 at 6:53 PM, rudy <ru...@greenfossil.com> wrote:
HI James, Ben, 

I can confirm that this bug exists as I am hitting it as well. You are right about it not happening in 2.1.0 but I am able to reproduce it in 2.1.2-RC2. 

Which bug?  The + encoding bug?  Or the %20 decoding bug?  The latter I couldn't reproduce.  Could you provide some sample code that reproduces?
 


On Friday, July 5, 2013 8:59:24 AM UTC+8, James Roper wrote:

Yann Simon

unread,
Jul 5, 2013, 12:08:16 PM7/5/13
to play-fr...@googlegroups.com

Could you please consider the fix for issue 1057 (parsing of invalid accept headers)

https://github.com/playframework/Play20/pull/1057

Thanks!

Ben McCann

unread,
Jul 5, 2013, 4:23:33 PM7/5/13
to play-fr...@googlegroups.com
Hi James,

Thanks for attempting to reproduce. The problem is no longer occurring for me. I'm guessing that it was some of SBT's wonky caching behavior once again. I appreciated you taking the time and am sorry it didn't turn out to be a real issue.

There's a bug that Nilanjan and I worked together to get fixed many months ago that's still present in 2.1.2-RC2. Would we be able to get the fix ported? It's fixed in master, but that unfortunately doesn't do us a lot of good until a new release is cut with the bug fix.

Thanks,
Ben

James Roper

unread,
Jul 5, 2013, 8:58:11 PM7/5/13
to play-framework
Hi Ben,

Ok, we'll consider that change for 2.1.3.  Do you know what the work around is to get it working today?  You should be able to add the dependency to your plugins.sbt file, and that will pull in the later version of junit-interface.

Cheers,

James

Rudresh Jariwala

unread,
Jul 6, 2013, 10:28:14 PM7/6/13
to play-fr...@googlegroups.com
Hi James, 

I confirm that the %20 encoding bug exists in 2.1.2-RC2 and you are able to reproduce it

But in 2.1.0, When I pass + as a part of my query string e.g. : "test + string" , My Controller action is not seeing the character after being parsed. The value I see in my action is "test  string"



Regards, 
Rudresh J.

Ben McCann

unread,
Jul 22, 2013, 3:09:12 PM7/22/13
to play-fr...@googlegroups.com
Hi James,

I see Play 2.1.3-RC1 in the Maven repo today. Did this change make it in? I haven't yet had a chance to test whether I can override the dependencies and pull in a newer version of junit-interface myself, but I think it'd be great to have this bug fix in 2.1.3 regardless so that others don't run into it. Any chance we can add it to 2.1.3 if it's not there already?

Thanks,
Ben

Ben McCann

unread,
Jul 28, 2013, 1:35:00 PM7/28/13
to play-fr...@googlegroups.com
Hi James, 

I tried overriding the dependency in my project's dependencies in Build.scala and it did not work.

You suggested that I should add it to my plugins.sbt file. I'm not sure how to do that. Can you let me know so that I can test it?

Thanks,
Ben
Reply all
Reply to author
Forward
0 new messages