Re: Wicket hot reloading - Proof of concept
Hi Cedric, As a first step I think it will be good if the user application has control whether to re-init the app or not. As we talked in IRC it would be good if the broadcasted event brings info what has changed. This way the user code can decide whether there is a need to restart the app. By default it should not be restarted. I'll use the library and give you more feedback. On Thu, Aug 1, 2013 at 10:35 PM, Cedric Gatay gata...@gmail.com wrote: Application reloading is one of the current bottleneck of hot reloading. We need to understand how things are working to allow good extension points in order to reload classes without reloading the whole application. As I said in my early post, this is a more a Proof Of Concept than a real usable solution. __ Cedric Gatay (@Cedric_Gatay http://twitter.com/Cedric_Gatay) http://code-troopers.com | http://www.bloggure.info | http://cedric.gatay.fr On Thu, Aug 1, 2013 at 7:29 PM, Dan Retzlaff dretzl...@gmail.com wrote: The subject of re-invoking Application#init() came up a couple of years ago in a related discussion. http://mail-archives.apache.org/mod_mbox/wicket-dev/201109.mbox/%3c4e721208.5030...@apache.org%3E Since the current design is not idempotent I think it would need some wicket-core refactoring to work nicely. On Thu, Aug 1, 2013 at 10:02 AM, Decebal Suiu decebal.s...@asf.ro wrote: Hi I'm the creator of wicket-plugin [1] and pf4j [2] and I'm very sensible at (wicket) hot swap. Play is stateless. So, it can unload the Controller, forget all the fields and load and initialize it again. True (generic) hot swap it hard to achieve but I will try to do incremental steps to see what I can do with my custom class loader (PluginClassLoader) with minimal memory leaks and minimal application restart:) It's interesting if I can use the same plugin (module) framework for the entire stack (web layer, service layer, persistence layer). My question is: I can monitor my wicket plugins to see modifications on class files, I can dispose my plugin class loader, I can create a new plugin class loader but can I init again my WicketApplication (programmatically)? It should be easy more for a stateless web application (framework). Make sense my idea with reinit wicket application? I see many problems with this approach and I don't know if it's possible but it's easy to put some questions :) PS I don't want to discuss in this post about OSGI [1] https://github.com/decebals/wicket-plugin [2] https://github.com/decebals/pf4j -- View this message in context: http://apache-wicket.1842946.n4.nabble.com/Wicket-hot-reloading-Proof-of-concept-tp4660575p4660627.html Sent from the Forum for Wicket Core developers mailing list archive at Nabble.com.
Re: Wicket hot reloading - Proof of concept
Am 01.08.13 09:55, schrieb Martin Grigorov: Michael, On Thu, Aug 1, 2013 at 9:48 AM, Michael Mosmann mich...@mosmann.de wrote: Only to be sure.. You know the run-jetty-run eclipse plugin? All I know about this project is that Eelco started it. I think this project is very usefull. But until its finished there are some solutions to take away the pain a little bit:) But looking at their page I see they also recommend DCEVM [1]. I use it since a while and I find it helpful. 1. http://blogs.jetbrains.com/idea/2013/07/get-true-hot-swap-in-java-with-dcevm-and-intellij-idea/ cool :) will try this..
Re: Wicket hot reloading - Proof of concept
Hi Martin, I will collect what has changed and log it out. I can add a property to auto restart the application as well. The current status of this POC is unbounded, I only tested it with small apps and I don't know if it will properly reload page classes for example if we don't reinit the application. I need to do further testing. If I have enough time this week end I will try to test it on a bigger application. Regards, __ Cedric Gatay (@Cedric_Gatay http://twitter.com/Cedric_Gatay) http://code-troopers.com | http://www.bloggure.info | http://cedric.gatay.fr On Fri, Aug 2, 2013 at 9:20 AM, Michael Mosmann mich...@mosmann.de wrote: Am 01.08.13 09:55, schrieb Martin Grigorov: Michael, On Thu, Aug 1, 2013 at 9:48 AM, Michael Mosmann mich...@mosmann.de wrote: Only to be sure.. You know the run-jetty-run eclipse plugin? All I know about this project is that Eelco started it. I think this project is very usefull. But until its finished there are some solutions to take away the pain a little bit:) But looking at their page I see they also recommend DCEVM [1]. I use it since a while and I find it helpful. 1. http://blogs.jetbrains.com/**idea/2013/07/get-true-hot-** swap-in-java-with-dcevm-and-**intellij-idea/http://blogs.jetbrains.com/idea/2013/07/get-true-hot-swap-in-java-with-dcevm-and-intellij-idea/ cool :) will try this..
Re: buildbot failure in ASF Buildbot on wicket-branch-6.x
You should commit with some delay between master and wicket-6.x because otherwise wicket-examples' test fail to use the same http port... On Fri, Aug 2, 2013 at 1:55 PM, build...@apache.org wrote: The Buildbot has detected a new failure on builder wicket-branch-6.x while building wicket. Full details are available at: http://ci.apache.org/builders/wicket-branch-6.x/builds/103 Buildbot URL: http://ci.apache.org/ Buildslave for this Build: hemera_ubuntu Build Reason: scheduler Build Source Stamp: [branch wicket-6.x] cdfa92ca3f8af8012a8f5c53076c1ffb7ce65944 Blamelist: svenmeier s...@meiers.net BUILD FAILED: failed compile sincerely, -The Buildbot
Re: buildbot failure in ASF Buildbot on wicket-branch-6.x
Should we do something to avoid that? Could the different branches use different ports? Or some buildbot-config? On Fri, Aug 2, 2013 at 7:58 AM, Martin Grigorov mgrigo...@apache.orgwrote: You should commit with some delay between master and wicket-6.x because otherwise wicket-examples' test fail to use the same http port... On Fri, Aug 2, 2013 at 1:55 PM, build...@apache.org wrote: The Buildbot has detected a new failure on builder wicket-branch-6.x while building wicket. Full details are available at: http://ci.apache.org/builders/wicket-branch-6.x/builds/103 Buildbot URL: http://ci.apache.org/ Buildslave for this Build: hemera_ubuntu Build Reason: scheduler Build Source Stamp: [branch wicket-6.x] cdfa92ca3f8af8012a8f5c53076c1ffb7ce65944 Blamelist: svenmeier s...@meiers.net BUILD FAILED: failed compile sincerely, -The Buildbot -- Jeremy Thomerson http://wickettraining.com