OK - deployed on stax and it works Thanks
D. On Tue, Oct 19, 2010 at 1:25 PM, Vassil Dichev <[email protected]> wrote: > Now my conversation messages are not duplicated *and* the conversation > is recorded. Can you check if it works? > > On Tue, Oct 19, 2010 at 12:08 PM, Richard Hirsch <[email protected]> > wrote: >> In the timeline, the link for the conversation is missing - therefore >> it is tough to see if the reply is really part of the conversation or >> not. >> >> On Tue, Oct 19, 2010 at 11:04 AM, Vassil Dichev <[email protected]> wrote: >>> Ah, you mean the link with the conversation? Got it. Let me try to fix this. >>> >>> On Tue, Oct 19, 2010 at 11:30 AM, Richard Hirsch <[email protected]> >>> wrote: >>>> I just deployed vassil's change on stax ( >>>> http://esmecloudserverapache.dickhirsch.staxapps.net/) and the >>>> conversations don't work any more - they don't show up in the timeline >>>> :-< >>>> >>>> >>>> On Tue, Oct 19, 2010 at 9:43 AM, Ethan Jewett <[email protected]> wrote: >>>>> Far more sophisticated than the println "instrumentation" I was using. >>>>> I'll >>>>> keep that in mind for the future. I hadn't paid much attention to the >>>>> after-hook methods in the MetaMapper until now. >>>>> >>>>> Ethan >>>>> >>>>> On Tue, Oct 19, 2010 at 9:20 AM, Vassil Dichev <[email protected]> wrote: >>>>> >>>>>> Heh, I added an afterSave handler with Thread.dumpStack to show me >>>>>> where does the double save occur. >>>>>> >>>>>> On Tue, Oct 19, 2010 at 9:49 AM, Ethan Jewett <[email protected]> wrote: >>>>>> > Hi Vassil, >>>>>> > >>>>>> > I got as far as confirming the the message ids are different, so it >>>>>> > looks to me like two messages are created and saved. I then started >>>>>> > looking for where the second message is created and didn't make any >>>>>> > progress. My tests seemed to show the the method creation code in >>>>>> > UserActor was only called once. >>>>>> > >>>>>> > If you can find some time to investigate, that would be excellent. >>>>>> > >>>>>> > Ethan >>>>>> > >>>>>> > On Tuesday, October 19, 2010, Vassil Dichev <[email protected]> wrote: >>>>>> >> An action is not likely to create a duplicate message, as even when it >>>>>> >> resends, a new message is never created and the mailbox is checked if >>>>>> >> such a message exists. I've solved at least 2 duplication bugs so far >>>>>> >> and the problem has always been that the same message was visualized >>>>>> >> twice. >>>>>> >> >>>>>> >> The first thing to look for is the id of the message in the page >>>>>> >> source- if the id is the same, then it's the same message shown twice >>>>>> >> by the UI. >>>>>> >> >>>>>> >> I'd like to take a look at the problem and check if I see an easy >>>>>> solution. >>>>>> >> >>>>>> >> Vassil >>>>>> >> >>>>>> >> >>>>>> >> On Tue, Oct 19, 2010 at 5:32 AM, Richard Hirsch >>>>>> >> <[email protected]> >>>>>> wrote: >>>>>> >>> On Mon, Oct 18, 2010 at 7:30 PM, Ethan Jewett <[email protected]> >>>>>> wrote: >>>>>> >>>> No, I don't think so. I did a clean install locally last week and I >>>>>> only >>>>>> >>>> have the RSS-pull action set up. It's a ghost in the machine >>>>>> >>>> somewhere >>>>>> ;-) I >>>>>> >>>> spent some time tracing through actor messaging paths and I guess >>>>>> >>>> I'll >>>>>> just >>>>>> >>>> need to spend a little more time on it. >>>>>> >>>> >>>>>> >>>> An alternative would be to find an SVN check-in that doesn't have >>>>>> >>>> the >>>>>> issue >>>>>> >>>> and then track down the exact change that caused the problem. Not >>>>>> >>>> sure >>>>>> which >>>>>> >>>> will be more difficult ... >>>>>> >>> >>>>>> >>> This would probably be a challenge - especially if the problem has >>>>>> >>> been around for a while. >>>>>> >>> >>>>>> >>> D. >>>>>> >>>> >>>>>> >>>> Ethan >>>>>> >>>> >>>>>> >>>> On Mon, Oct 18, 2010 at 5:51 PM, Richard Hirsch < >>>>>> [email protected]>wrote: >>>>>> >>>> >>>>>> >>>>> On Mon, Oct 18, 2010 at 5:41 PM, Ethan Jewett <[email protected]> >>>>>> wrote: >>>>>> >>>>> > Hi all, >>>>>> >>>>> > >>>>>> >>>>> > Dick did some testing after my commits today and found a couple >>>>>> >>>>> > of >>>>>> >>>>> issues. >>>>>> >>>>> > The first (ESME-290) was that I broke the public timeline. That >>>>>> >>>>> > is >>>>>> now >>>>>> >>>>> > fixed. >>>>>> >>>>> > >>>>>> >>>>> > The second issue is ESME-291 - >>>>>> >>>>> > https://issues.apache.org/jira/browse/ESME-291. The problem is >>>>>> that >>>>>> >>>>> replies >>>>>> >>>>> > show up duplicated in the streams view and in the public timeline >>>>>> view >>>>>> >>>>> > (though in the public timeline view it is *only* for messages >>>>>> >>>>> > that >>>>>> were >>>>>> >>>>> > created before the public timeline actor started up, so if you >>>>>> create new >>>>>> >>>>> > replies you need to restart the server to see them duplicated in >>>>>> the >>>>>> >>>>> public >>>>>> >>>>> > timeline). This is actually an old issue. It exists at least in >>>>>> release >>>>>> >>>>> 1.1 >>>>>> >>>>> > and probably well before that. >>>>>> >>>>> > >>>>>> >>>>> > The reason this is happening is that when a reply is created >>>>>> >>>>> > there >>>>>> are >>>>>> >>>>> > actually two messages created and persisted to the database. One >>>>>> >>>>> > is >>>>>> >>>>> created >>>>>> >>>>> > in the UserActor, which is what is supposed to happen. But >>>>>> >>>>> > another >>>>>> >>>>> message >>>>>> >>>>> > is created somewhere else and I can't figure out where! Any >>>>>> >>>>> > ideas? >>>>>> >>>>> >>>>>> >>>>> Some action maybe? >>>>>> >>>>> >>>>>> >>>>> Maybe Vassil has an idea. >>>>>> >>>>> > >>>>>> >>>>> > Ethan >>>>>> >>>>> > >>>>>> >>>>> >>>>>> >>>> >>>>>> >>> >>>>>> >> >>>>>> >> >>>>>> >> >>>>>> >> -- >>>>>> >> Twitter: http://twitter.com/vdichev >>>>>> >> Blog: http://speaking-my-language.blogspot.com >>>>>> >> >>>>>> > >>>>>> >>>>>> >>>>>> >>>>>> -- >>>>>> Twitter: http://twitter.com/vdichev >>>>>> Blog: http://speaking-my-language.blogspot.com >>>>>> >>>>> >>>> >>> >>> >>> >>> -- >>> Twitter: http://twitter.com/vdichev >>> Blog: http://speaking-my-language.blogspot.com >>> >> > > > > -- > Twitter: http://twitter.com/vdichev > Blog: http://speaking-my-language.blogspot.com >
