[00:01:46] *** aslak has quit IRC
[00:20:46] *** emmanuel has quit IRC
[00:43:51] *** cbrock has quit IRC
[01:03:36] *** jbossbot has quit IRC
[01:03:43] *** dmlloyd has quit IRC
[01:49:19] *** jbossbot has joined #weld-dev
[01:50:09] *** dmlloyd has joined #weld-dev
[06:41:14] *** magesh has joined #weld-dev
[07:31:23] *** jharting has joined #weld-dev
[08:11:08] *** oskutka has joined #weld-dev
[08:35:18] *** aslak has joined #weld-dev
[09:16:24] *** maschmid has joined #weld-dev
[09:36:13] *** epbernard has joined #weld-dev
[09:36:13] *** epbernard is now known as emmanuel
[09:37:47] *** emmanuel has quit IRC
[09:54:18] *** emmanuel has joined #weld-dev
[10:33:23] *** alesj has joined #weld-dev
[11:39:36] *** stuartdouglas has quit IRC
[11:40:34] *** stuartdouglas has joined #weld-dev
[11:43:10] *** alesj has quit IRC
[11:56:29] *** alesj has joined #weld-dev
[12:11:25] *** emmanuel has quit IRC
[12:12:22] <nickarls> alesj: ping
[12:15:41] <alesj> nickarls: hey
[12:16:50] <nickarls> alesj: was wondering about WELD-697, how are they treated currently? does the "foreground conversation" time out like all the others?
[12:16:51] <jbossbot> jira [WELD-697] Foreground conversations [Open (Unresolved) Feature Request, Major, Unassigned] https://issues.jboss.org/browse/WELD-697
[12:18:05] <alesj> nickarls: i would say so
[12:18:11] <nickarls> the touch() happens at deactivate()
[12:18:23] <nickarls> and the clean at invalidate()
[12:18:39] <nickarls> so as long as the validity isn't checked at start, the foreground conversation should live, I think
[12:19:08] <nickarls> hmm, which order were they called in, now?
[12:20:27] <nickarls> the WeldPhaseListener does it in the reverse order
[12:20:51] <alesj> elt me check
[12:21:01] <nickarls> so if the conversation times out in 10 minutes, and you resume one after 11, the "foreground" conversation is cleaned out before it's touched?
[12:23:13] <nickarls> that's my theory
[12:24:18] <nickarls> so to work around it, you would have to reverse the order, touch at resume of conversation or pass the cid to isExpired as an exception
[12:24:19] <alesj> see AbstractConversationContext
[12:24:47] <nickarls> yep
[12:24:54] <alesj> if it's not transient
[12:25:01] <alesj> it looks like it gets touched before cleanup
[12:25:49] <nickarls> Look at WeldPhaseListener.deactivateConversations
[12:25:59] <nickarls> it calls invalidate and the deactivate
[12:26:11] <alesj> yup
[12:26:31] <alesj> invalidate will first clean it, and then deactivate will touch it
[12:27:02] <nickarls> invalidate will mark it short-running
[12:27:08] <nickarls> transient
[12:27:30] <nickarls> after that, touching it will do it little good, right?
[12:27:37] <nickarls> or am I completely misreading the code
[12:28:00] <alesj> invalidate just ends it
[12:28:30] <alesj> which makes it traniesnt, right
[12:29:42] <nickarls> actually, it won't even be touched as the transients are killed off before that
[12:30:12] <alesj>        if (getCurrentConversation().isTransient())
[12:30:12] <alesj>          {
[12:30:12] <alesj>             destroy();
[12:30:22] <alesj> so, what's trhe issue you're having?
[12:30:41] <alesj> ** off to lunch **
[12:31:12] <nickarls> I would like the foreground conversation to live as long as the session as it did in Seam2
[12:31:53] <nickarls> I don't think it's a spec violation either as the timeout is only a hint to CDI, if I don't remember incorrectly
[12:35:05] <nickarls> It would perhaps be possible to make a Seam extension that would tickle the conversation at the beginning of the request but the code is quite deep in the AbsConvCtx.
[12:35:18] <nickarls> Haven't look what seam conversation exposes for this.
[12:40:58] *** magesh1 has joined #weld-dev
[12:43:01] *** magesh has quit IRC
[13:10:33] <alesj> nickarls: seam conversation is basically similar to conversation-context
[13:11:59] <alesj> perhaps ask this on CDI 1.1. mailing list
[13:12:10] <alesj> and we can then adjust Seam conversation
[13:12:15] <alesj> nickarls: ^
[13:23:31] *** epbernard has joined #weld-dev
[13:23:31] *** epbernard is now known as emmanuel
[13:28:20] <nickarls> alesj: ok, until then I'll just make my own Weld that does what I want ;-P
[13:28:47] <alesj> nickarls: not again ? :-)
[13:29:08] <nickarls> new! improved!
[13:56:16] *** magesh1 has quit IRC
[14:00:15] *** rmartinelli has joined #weld-dev
[14:20:37] *** magesh has joined #weld-dev
[15:18:54] *** rmartinelli has quit IRC
[15:19:48] *** rmartinelli has joined #weld-dev
[15:30:41] *** aslak has quit IRC
[16:20:56] *** magesh has left #weld-dev
[16:25:51] *** mbg has joined #weld-dev
[17:02:46] *** jharting has quit IRC
[17:07:45] *** mbg has quit IRC
[17:08:12] *** mbg has joined #weld-dev
[17:16:24] *** alesj has quit IRC
[17:21:45] *** oskutka has quit IRC
[17:44:58] *** emmanuel has quit IRC
[17:48:20] *** kevinpollet has joined #weld-dev
[18:02:15] *** epbernard has joined #weld-dev
[18:02:15] *** epbernard is now known as emmanuel
[18:18:13] *** maschmid has quit IRC
[18:44:22] *** mbg|away has joined #weld-dev
[19:27:26] *** kevinpollet has quit IRC
[19:28:26] *** kevinpollet has joined #weld-dev
[19:38:25] *** maschmid has joined #weld-dev
[20:03:20] *** kevinpollet has quit IRC
[20:46:52] *** alesj has joined #weld-dev
[20:48:40] *** kevinpollet has joined #weld-dev
[21:06:11] *** mbg|away has quit IRC
[21:34:57] *** maschmid has quit IRC
[21:35:28] *** maschmid has joined #weld-dev
[21:55:12] *** rmartinelli has quit IRC
[22:01:41] *** mbg|wfh has joined #weld-dev
[22:07:04] *** mbg has quit IRC
[22:07:49] *** mbg has joined #weld-dev
[22:19:39] *** kevinpollet has quit IRC
[22:56:52] *** rruss has joined #weld-dev
[23:15:07] *** mbg|wfh has quit IRC





top