Next 2.4 milestone?
348 views
Skip to first unread message
benmccann
Apr 3, 2015, 4:04:54 AM4/3/15
to play-fram...@googlegroups.com
I was curious what the plans were for cutting the next release whether that's another milestone or the first RC.
Some things that'd be good to get into the final version and potentially get in before an RC:
- I'd like to get a new version of sbteclipse into Play at which point I can kill off PlayEclipse entirely since all the changes have been upstreamed. I've just got one minor issue I'd like some SBT help on before we cut the sbteclipse release
- I'm really really hoping that someone can document how to build an application loader. It's super close, but I think we'll find in the process that it's not quite possible and a few code tweaks are needed.
- There's a CSRF bug fix PR open that we'll want to get merged
- Maybe we're waiting for Akka 2.4 to be released?
-Ben
Rich Dougherty
Apr 3, 2015, 8:19:37 AM4/3/15
to benmccann, play-fram...@googlegroups.com
Hi Ben
On Fri, Apr 3, 2015 at 9:04 PM, benmccann <benjamin...@gmail.com> wrote:
I was curious what the plans were for cutting the next release whether that's another milestone or the first RC.
I'm not sure when the next milestone is planned. James may be able to comment.
Some things that'd be good to get into the final version and potentially get in before an RC:
- I'd like to get a new version of sbteclipse into Play at which point I can kill off PlayEclipse entirely since all the changes have been upstreamed. I've just got one minor issue I'd like some SBT help on before we cut the sbteclipse release
Let me know if you need any help getting this looked at.
- I'm really really hoping that someone can document how to build an application loader. It's super close, but I think we'll find in the process that it's not quite possible and a few code tweaks are needed.
OK I can get ApplicationLoader documented.
- There's a CSRF bug fix PR open that we'll want to get merged
I reviewed that a few days ago and I have a comment about performance. It should be merged soon.
- Maybe we're waiting for Akka 2.4 to be released?
The plan is that Play 2.4 will use Akka 2.4. We'll probably have to wait for Akka 2.4 milestones before we can release a Play 2.4 RC. I think Akka 2.4 is due to be released sometime this month.
FYI here's my own list of tasks to get done for 2.4:
- I'm going to spend a few days looking at performance. At the moment Play 2.4 is slightly slower than Play 2.3 on some benchmarks. I'd like to address that. So far I've done some profiling and made a few minor tweaks that bring back a few % in performance. Next I'd like to reduce thread switching overhead by consolidating the ExecutionContexts down into a single ExecutionContext.
- Add support for Akka HTTP features at they become available. So far the big missing features are HTTPS and WebSockets.
Cheers
Rich
Ben McCann
Apr 6, 2015, 9:03:02 PM4/6/15
to Rich Dougherty, play-fram...@googlegroups.com
Let me know if you need any help getting this looked at.
- I'd like to get a new version of sbteclipse into Play at which point I can kill off PlayEclipse entirely since all the changes have been upstreamed. I've just got one minor issue I'd like some SBT help on before we cut the sbteclipse release
jsuereth has been helping me get unstuck on a lot, but we're both pretty busy, so it's tough for me to spend much time on it or catch him when we're both free. If you have any time to pick up my branch and get the tests to pass I'd be super happy. I've updated the issue with all the details.
Thanks!
Ben
James Roper
Apr 6, 2015, 9:08:05 PM4/6/15
to benmccann, play-fram...@googlegroups.com
Hi all,
The plan for the final release was/is to wait for Akka 2.4. Though, Akka 2.4 is delaying (was supposed to be released in March, then April, now May). I think it doesn't hurt to wait for Akka 2.4, otherwise we'll end up having to cut a Play 2.5 soon after Play 2.4 if we want Play users to be able to use Akka 2.4.
I'm pretty keen to start release candidates for Play 2.4 pretty soon.
Regards,
James
--
You received this message because you are subscribed to the Google Groups "Play framework dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to play-framework-...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Dominik Dorn
Apr 7, 2015, 12:10:33 PM4/7/15
to James Roper, benmccann, play-fram...@googlegroups.com
Its maybe just my personal opinion, but I'd rather have a 2.4 now and a 2.5 later then to wait 2 more months to move from the current snapshot build I have in my local repository, just for waiting for Akka.
The changes made in regard to dependency injection are worth a release on their own in my opinion.
In short: I'd go for a 2.4 now and a 2.5 when Akka is ready.
just my 2c
cheers,
Dominik
Ben McCann
Apr 7, 2015, 12:22:44 PM4/7/15
to Dominik Dorn, James Roper, play-fram...@googlegroups.com
I'll throw out another option... The plan for Play 3.0 I believe includes moving the package that most classes reside in among other possible changes. When we go a long time between releases it makes upgrades tough as an end user. This is mostly because there are more changes needed in a project to upgrade. Something that would be nice would be to have a release that contains a lot of API changes, but not many other changes. This would make the API changes easier to deal with as an end user. One reason is that more changes also means there's more likely to be bugs or changes in functionality that weren't expected. When we run into something like that we'd typically like to rollback until we can address the issue, but that gets really complicated when the upgrade is a big one with lots of changes.
I think it could be pretty nice to do a Play 2.4 soon. And then a 2.5 or 3.0, which contains not only the Akka upgrade, but the package changes and maybe some other larger but easy to implement API changes currently planned for 3.0.
-Ben
James Roper
Apr 8, 2015, 3:09:21 AM4/8/15
to Ben McCann, Dominik Dorn, play-fram...@googlegroups.com
Hey all,
I've just had a discussion with Roland (akka lead). Akka 2.4 may not be available till July, and there's every chance it will slip beyond that. So we're going to go ahead with a Play 2.4 release. It may also be possible that Akka 2.4 will be binary compatible enough™ with Akka 2.3 that we can safely upgrade to it in a Play 2.4.x patch release. But we'll deal with that when it comes.
With that in mind, I'd like to release Play 2.4.0-RC1 this week. Ben, do you think you can have your stuff ready in time for that? Rich, do you think you can have your execution context stuff done in time for that? Other tasks:
* Merge a few more pull requests (Greg Methvins JsObject one perhaps - though not sure, what do people think of it?)
* Complete the Play 2.4 highlights doc
Regards,
James
Ben McCann
Apr 8, 2015, 3:53:13 AM4/8/15
to James Roper, Dominik Dorn, play-fram...@googlegroups.com
I'm super super close with the sbteclipse stuff. I need a bit of help understanding an SBT thing I got stuck on. If someone can help unstick me we can probably cut a release potentially as soon as tomorrow and get Play upgraded to use it. Josh has been super helpful to me and I'll keep bugging him. If anyone else wants to take a peek I wouldn't mind
I also really want to make sure that we can provide our own ApplicationLoader otherwise I won't be able to migrate my app to Play 2.4 since all of our Guice modules take constructor parameters which Play 2.4 doesn't support yet as far as I can tell. Rich, said he'd take a look in his last email.
Super excited for 2.4!!
-Ben
Ben McCann
Apr 8, 2015, 5:49:57 PM4/8/15
to James Roper, Dominik Dorn, play-fram...@googlegroups.com
Josh said my new sbteclipse test is failing because there's a bug in the sbteclipse scripted setup. He's going to push a fix for that tonight. I'll do a bit more testing on it after that and then I think we're good to go
Rich Dougherty
Apr 8, 2015, 6:17:29 PM4/8/15
to Ben McCann, play-fram...@googlegroups.com
On Tue, Apr 7, 2015 at 1:02 PM, Ben McCann <b...@benmccann.com> wrote:
Let me know if you need any help getting this looked at.
- I'd like to get a new version of sbteclipse into Play at which point I can kill off PlayEclipse entirely since all the changes have been upstreamed. I've just got one minor issue I'd like some SBT help on before we cut the sbteclipse release
jsuereth has been helping me get unstuck on a lot, but we're both pretty busy, so it's tough for me to spend much time on it or catch him when we're both free. If you have any time to pick up my branch and get the tests to pass I'd be super happy. I've updated the issue with all the details.
From a later email it sounds like you've got some help from Josh and you're OK with this now. If not, I may be able to help out next week.
– Rich
Rich Dougherty
Apr 8, 2015, 6:39:24 PM4/8/15
to James Roper, Ben McCann, Dominik Dorn, play-fram...@googlegroups.com
On Wed, Apr 8, 2015 at 7:08 PM, James Roper <ja...@typesafe.com> wrote:
Rich, do you think you can have your execution context stuff done in time for that?
I'll do my best, but I expect to have a PR for review on Monday morning. I'm pretty sure I know what needs doing now, it's just a matter of doing it. :)
Ben McCann
Apr 8, 2015, 10:43:14 PM4/8/15
to Rich Dougherty, James Roper, Dominik Dorn, play-fram...@googlegroups.com
Josh fixed the problem. It was a weird one. Possibly a bug in Scala. I'm working on a last bit of cleanup now
--
You received this message because you are subscribed to the Google Groups "Play framework dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to play-framework-...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Ben McCann
Apr 8, 2015, 11:07:53 PM4/8/15
to Rich Dougherty, James Roper, Dominik Dorn, play-fram...@googlegroups.com
Ready for an sbteclipse release now. After it's cut, I'll upgrade Play to use it and make the corresponding improvements there and then I'm all set with what I wanted to get done.
Filed an issue asking for a new release:
Thanks,
Ben
Dale Wijnand
Apr 9, 2015, 2:55:20 AM4/9/15
to Ben McCann, Rich Dougherty, James Roper, Dominik Dorn, play-fram...@googlegroups.com
Would Play 2.4 without Akka 2.4 still require Java 8?
Sorry only recently joined the mailing group and don't know what 2.4 brings and requires.
Dale
Ben McCann
Apr 9, 2015, 10:40:42 AM4/9/15
to Dale Wijnand, Rich Dougherty, James Roper, Dominik Dorn, play-fram...@googlegroups.com
Yep, Play 2.4 requires Java 8 independently of Akka
-Ben
Ben McCann
Apr 13, 2015, 11:03:04 PM4/13/15
to Dale Wijnand, Rich Dougherty, James Roper, Dominik Dorn, play-fram...@googlegroups.com
Here's the PR to finish up my eclipse work (note that that it requires a release to be cut for sbteclipse).
Rich, you still willing to take a look at https://github.com/playframework/playframework/issues/4086?
-Ben
Rich Dougherty
Apr 14, 2015, 12:00:22 AM4/14/15
to Ben McCann, Dale Wijnand, James Roper, Dominik Dorn, play-fram...@googlegroups.com
Hi Ben
Thanks for the reminder. Yes I'll take a look before the RC.
– Rich
--
Ben McCann
Apr 18, 2015, 2:22:05 AM4/18/15
to Rich Dougherty, Dale Wijnand, James Roper, Dominik Dorn, play-fram...@googlegroups.com
Only 4 issues left assigned to the 2.4 milestone!
Rich Dougherty
Apr 18, 2015, 6:59:54 PM4/18/15
to Ben McCann, Dale Wijnand, James Roper, Dominik Dorn, play-fram...@googlegroups.com
I'll get onto documenting play.application.loade (#4086) tomorrow. I talked to James about the dev mode memory leak (#3799) and I think it's because of a Guice thread leak. The leak is reasonaby slow—it happens on each run command, not on each reload—so we may be able to live this with for a while then fix it after the RC.
naturallog
Apr 22, 2015, 3:28:10 AM4/22/15
to play-fram...@googlegroups.com, b...@benmccann.com, ja...@typesafe.com, dom...@dominikdorn.com, dale.w...@gmail.com
It would be great to have a milestone of play-slick to go along with the RC1.
To unsubscribe from this group and stop receiving emails from it, send an email to play-framework-dev+unsub...@googlegroups.com.
--
You received this message because you are subscribed to the Google Groups "Play framework dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to play-framework-dev+unsub...@googlegroups.com.
--
You received this message because you are subscribed to the Google Groups "Play framework dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to play-framework-dev+unsub...@googlegroups.com.
--
--
You received this message because you are subscribed to the Google Groups "Play framework dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to play-framework-dev+unsub...@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages