Displaying #apache-syncope/2016-04-06.log:

Wed Apr 6 05:25:46 2016  jbonofre:Joined the channel
Wed Apr 6 06:26:21 2016  ilgrosso:Joined the channel
Wed Apr 6 06:38:34 2016  ilgrosso:Joined the channel
Wed Apr 6 06:43:11 2016  ilgrosso_:Joined the channel
Wed Apr 6 06:57:00 2016  fmartelli:Joined the channel
Wed Apr 6 07:16:47 2016  _massi_:Joined the channel
Wed Apr 6 08:56:30 2016  sberyozkin:Joined the channel
Wed Apr 6 09:12:04 2016  coheigea:Joined the channel
Wed Apr 6 09:29:57 2016  jbonofre:Joined the channel
Wed Apr 6 10:52:53 2016  coheigea:ilgrosso_: ping
Wed Apr 6 10:53:03 2016  ilgrosso_:coheigea: ilgrosso is alive
Wed Apr 6 10:54:52 2016  coheigea:ilgrosso_: Am wondering why there is no "war" file here: http://repo1.maven.org/maven2/org/apache/syncope/syncope-fit/2.0.0-M2/ ?
Wed Apr 6 10:57:40 2016  coheigea:With 2.0.0-SNAPSHOT, there is a org/apache/syncope/fit, but not for M2?
Wed Apr 6 11:01:37 2016  coheigea:There is no syncope-standalone for M2 here either: http://repo1.maven.org/maven2/org/apache/syncope/syncope-standalone/
Wed Apr 6 11:14:13 2016  ilgrosso_:The snapshot repo is populated by Jenkins, which makes some kind of copy, so everything is there
Wed Apr 6 11:14:30 2016  ilgrosso_:The release repo is populated by mvn release:prepare / release:perform which uses deploy
Wed Apr 6 11:14:51 2016  ilgrosso_:all artifacts that can be downlaoded from the download page are skipped during deploy, by config
Wed Apr 6 11:18:02 2016  ilgrosso_:with 2.0 we are not deplying to Maven any artifact that cannot used by Maven, including WAR files, since we are not generating WAR overlays any more, via archetype
Wed Apr 6 11:18:52 2016  coheigea:ilgrosso_: Ok thanks! One drawback to this approach is that there are no "wars" available to download from the non-snapshot repo. So for example, I have a test in github that downloads the fit-core-reference war + unpacks + configures it + launches it to do some testing on it, but I can't then use this approach with the released version
Wed Apr 6 11:19:25 2016  ilgrosso_:use standalone
Wed Apr 6 11:19:48 2016  ilgrosso_:under webapps/syncope there is essentially the exploded WAR
Wed Apr 6 11:20:16 2016  coheigea:ilgrosso_: Ok, so use standalone from the apache dist rather than maven? Yeah that could work
Wed Apr 6 11:20:27 2016  ilgrosso_:coheigea: yep
Wed Apr 6 11:22:15 2016  coheigea:ilgrosso_: Grazie! While I have you, another quick question. When adding a new domain to Syncope via the EntityManagerFactory thing, is it possible to change the configuration dynamically?
Wed Apr 6 11:22:37 2016  ilgrosso_:what do you want to change?
Wed Apr 6 11:23:16 2016  ilgrosso_:adding a new domain requires restart (and redeploy, in case)
Wed Apr 6 11:23:24 2016  coheigea:Let's say we have users in domain A and everything is working fine. Now we want to add in users in domain B, but don't want to have to restart the core to do this.
Wed Apr 6 11:23:31 2016  coheigea:Hmm ok that answers my question then
Wed Apr 6 11:26:06 2016  coheigea:Is there any scope for investigation on avoiding a redeploy or is this final?
Wed Apr 6 11:28:13 2016  ilgrosso_:for the moment, no more work on domains is planned
Wed Apr 6 11:29:22 2016  ilgrosso_:to be clear, if you have domain A and B configured upfront, they are both available, so you can add users to A and B
Wed Apr 6 11:34:03 2016  coheigea:yep understood.
Wed Apr 6 11:35:11 2016  coheigea:Let me rephrase, is there a fundamental obstacle towards making this process dynamic that you know of? Or could it be done with some investigation?
Wed Apr 6 11:43:13 2016  ilgrosso_:the point is that a domain requires several components - including DataSource - that are supposed to be available at the beginning of the whole core init process
Wed Apr 6 11:43:22 2016  ilgrosso_:(which is Spring-based, as you know)
Wed Apr 6 11:43:49 2016  ilgrosso_:one should think of reconsidering the whole init to allow each domain to be something that can be specified at runtime
Wed Apr 6 11:44:13 2016  ilgrosso_:but this would have a huge impact on the way how the core app is thought
Wed Apr 6 11:44:38 2016  coheigea:ok thanks again
Wed Apr 6 11:47:08 2016  ilgrosso_:nothing impossible for 2.1 / 2.2 / 3.0 / ... of course, but needs some solid use case requirement behind
Wed Apr 6 12:06:15 2016  fmartelli:Joined the channel
Wed Apr 6 12:06:42 2016  ilgrosso:Joined the channel
Wed Apr 6 12:39:09 2016  jbonofre:Joined the channel
Wed Apr 6 13:49:19 2016  _massi_1:Joined the channel
Wed Apr 6 13:50:38 2016  fmartelli:Joined the channel
Wed Apr 6 14:05:27 2016  fmartelli1:Joined the channel
Wed Apr 6 14:21:46 2016  svizzero81:Joined the channel
Wed Apr 6 18:30:34 2016  _massi_:Joined the channel
Wed Apr 6 18:35:05 2016  coheigea:Left the channel
Wed Apr 6 19:02:34 2016  jbonofre:Joined the channel
Wed Apr 6 19:17:13 2016  jbonofre:Joined the channel
Wed Apr 6 19:22:14 2016  _massi_:Joined the channel

Comments