greptilian logo

IRC log for #javaee, 2013-12-03

Please see http://irclog.greptilian.com/javaee for which days have been logged.

| Channels | #javaee index | Today | | Search | Google Search | Plain-Text | plain, newest first | summary

All times shown according to UTC.

Time S Nick Message
00:48 kobain joined ##javaee
00:51 jieryn joined ##javaee
00:51 jieryn joined ##javaee
01:01 kobain joined ##javaee
01:03 SoniEx joined ##javaee
02:18 jenue joined ##javaee
03:23 cem_ joined ##javaee
03:23 cem_ k anyone familar in eclipse ?
03:28 pdurbin a bit
03:36 cem_ just for a change in html it is deploying the war :( , this is how it works ?
03:42 pdurbin not sure
03:43 pdurbin with netbeans and glassfish you don't have to redeploy for xhtml changes
04:41 cem_ noticed :set the folder :)
04:41 cem_ oops file
05:06 cem_ left ##javaee
05:40 spro joined ##javaee
05:46 jenue joined ##javaee
06:21 spro tomcat is just of configuration currently for me :/
06:22 spro set of configuration
06:54 dangertools joined ##javaee
06:54 dangertools joined ##javaee
06:55 SoniEx joined ##javaee
07:11 pleroma joined ##javaee
07:36 * spro tomcat scks
08:50 AartBluestoke joined ##javaee
09:37 balo joined ##javaee
11:21 r3m1 joined ##javaee
11:24 r3m1 left ##javaee
11:34 drspockbr joined ##javaee
14:17 kobain joined ##javaee
14:18 AartBluestoke1 joined ##javaee
14:41 Naros joined ##javaee
15:14 weyer joined ##javaee
15:35 pdurbin was getting "Caused an ERROR Provider org.glassfish.json.JsonProviderImpl not found" and this was helpful: http://blog.matthias-reining.com/testing-json-on-javaee--application-server-jsonproviderimpl-not-found/
15:36 pdurbin needed http://search.maven.org/#search|ga|1|a%3A%22javax.json%22
15:46 javier joined ##javaee
15:46 javier joined ##javaee
16:13 cem__ joined ##javaee
16:14 cem__ k after going thru , all can say its not :/
16:35 liq left ##javaee
16:40 sheenobu joined ##javaee
16:44 SoniEx joined ##javaee
17:01 liq joined ##javaee
17:02 * liq forever against tomcat
17:13 jenue joined ##javaee
17:34 neuro_sys liq: why
17:38 liq just for small things we need to [break;] and its not structured at all .
17:39 liq faq is bad too
17:40 jenue joined ##javaee
17:49 knoppix joined ##javaee
17:55 * liq tomcat is for people who wants to waste tons of time in configuration
18:01 * liq soon i'll how many other feature sucks in :P
18:11 jenue joined ##javaee
18:16 liq http://zeroturnaround.com/rebellabs/the-great-java-application-server-debate-with-tomcat-jboss-glassfish-jetty-and-liberty-profile/
18:17 liq 59% uses Tomcat :/
18:18 liq glassfish is 11% [just born reasonable]
18:22 sfisque joined ##javaee
18:22 * liq ppl go to other server bcuz u know application server :)
18:33 liq k doubt on sngletreadmdel
18:33 liq :P
18:34 liq here goes
18:34 liq is servlet instance re-instalized every now and then after serving the response
18:35 liq since i cant verify since using a host
18:50 jenue joined ##javaee
18:58 jenue joined ##javaee
19:29 sfisque anyone have experience with managing OIDs of things?
19:29 whartung like what sfisque
19:30 sfisque well i want to store things in my app in a hierarchy (probably ldap or similar) and i'd like to do it according to a standard for interoperability.  but i'm pretty ignorant of the whole OID framework
19:31 sfisque i'd like to know how it works (registration, etc.)
19:32 whartung Basically, OIDS are like domain names, right? "someone" owns some OID, and then they can do whatever they want down the tree. We work in health care, so we got our OIDs from the HL7 committee. As to who has dominion over which higher level OIDs, I can't say.
19:33 whartung but with our HL7 OIDs, we get it from them, and then we can do anything with that root, since it's "ours"
19:33 whartung so if they send us "1.2.3.4", we can do anything from there: "1.2.3.4.1", "1.2.3.4.1.2.4.3939393", etc.
19:34 whartung the fact that it's an HL7 root OID is really meaningless.
19:34 whartung You're basically quibbling over who made your ethernet card at that level.
19:35 whartung there's no real "authority" "Oh, that's an HL7 OID, that's better than an ASN.1 OID"
19:35 pdurbin I registered an OID once. My name is here: http://www.iana.org/assignments/enterprise-numbers/enterprise-numbers
19:36 whartung yea, that sounds about right
19:36 pdurbin sfisque: if they let me do it they'll let you :)
19:36 whartung 16789?
19:36 pdurbin such an awesome number
19:36 whartung well it gets even better.
19:37 whartung You can give sfisque ONE OF YOURS!
19:37 pdurbin !
19:37 pdurbin no way
19:37 pdurbin sfisque: take a hike
19:37 whartung so pdurbin is 1.3.6.1.4.1.16789
19:38 pdurbin well, berklee.edu
19:38 pdurbin where I used to work
19:38 whartung sfisque can 1.3.6.1.4.1.16789.1337
19:38 pdurbin :)
19:40 whartung here ya go sfisque
19:40 whartung http://pen.iana.org/pen/PenApplication.page
19:42 SoniEx joined ##javaee
19:42 whartung http://www.oid-info.com/
19:45 sheenobu joined ##javaee
19:46 pdurbin I'm dinking around with junit tests and one of them tests http://localhost:8080/api/search but I'm not sure how to make it execute only when the app has already been deployed
19:46 whartung pretty cool -- so that IANA number
19:46 pdurbin I mean, it passes if I've already deployed the app. But it fails if I undeploy and try to run my tests
19:47 pdurbin for now I'm just adding @Ignore so it doesn't run :)
19:47 sess pdurbin: it's called a bad unit test :V
19:47 whartung 1 is "ISO", 3 is "Identified organization", 6 is the Department of Defense, 1 is "the internet", 4 is private projects, 1 is enterprise,
19:47 sess possibly a mock could solve it
19:47 sess a web server mock
19:48 pdurbin sess: yeah, I'm using just straight junit right now... probably need to add a lot more to it. arquillian or whatever
19:49 sess note that it's not a unit test if it's not mocked
19:49 pdurbin ok
19:49 semiosis pdurbin: i use jenkins to orchestrate end-to-end tests.  i have projects set up to build automatically in this order, core -> rest api -> test client -> test driver.
19:50 semiosis when i push commits to one of the projects, jenkins runs the job & all the jobs downstream
19:50 pdurbin I guess this is more of an integration test... I'm doing url.openConnection()
19:50 sfisque cool thx for the link and info.  basically i just want to make sure i choose a tree node that isnt taken and "makes sense" when i start generating them internally
19:50 sfisque that way i don't collide
19:50 whartung well, register one with the IANA -- free and easy...
19:50 pdurbin semiosis: cool. for now I'm just working with netbeans and glassfish, no jenkins
19:50 whartung until then, use pdurbins :)
19:51 * pdurbin smacks whartung
19:51 * sfisque grins evilly
19:51 semiosis pdurbin: we have four environments for our apps: dev (local workstations), test (cluster where end to end integration tests run), staging & prod
19:51 whartung kids are neat -- unreadable, but neat
19:52 whartung oidss
19:52 whartung kids are the same way…but…
19:52 sfisque lolz
19:52 pdurbin semiosis: do the dev tests have mocks like sess said? to do the equivalent of curl commands?
19:53 semiosis we use mockito *extensively*
19:53 semiosis but it's not quite the equiv of curl
19:53 pdurbin semiosis: ok. but hopefully you know what I mean
19:53 sess pdurbin: for unit tests you would use mocks, otherwise it's some kind of integration / system test
19:54 sess otherwise you have to setup an actual web server
19:54 semiosis we use the mocks more on the DB side
19:54 sfisque aye.  unit test is for internal sanity.  IT is for external linkage testing
19:54 sess the tests with mocks would test logic requiring the url.connection, not the actual web parts
19:54 sfisque we leverage arquillian for IT so we get all the EJB goodies like injection and jndi lookup
19:54 pdurbin sess: right. not a unit test. but I do have an actual web server (glassfish) on port 8080 on my laptop (localhost)
19:55 sess pdurbin: tests should always be runnable from any computer without any external resources
19:55 sfisque ^^^ unit tests
19:56 sess sfisque: I tried arquillian but never got it to work
19:56 sfisque IT tests assume some form of "infrastructure" like dev/qa installs of remote services
19:56 sess a collegue of mine wrote his own test framework for EJB but it didn't handle transactions well so didn't get that working either...
19:56 sfisque we had someone do the heavy lifting on getting it bootstrapped.  took him a few months but it's VERY NICE now that it runs
19:56 sfisque arq bootstraps embedded GF and everything links up nicely in IT harness
19:57 sfisque whartung - confirmed my iana pen app.  waiting ..  ….  …… ...........
19:57 sess it's supposed to yeah, but never worked for me. Was a few years ago so don't remember the problem :p
19:57 sess perhaps I should give it another shot
19:58 sfisque if you have a spare resource, it's worth it, but definitely takes a bit of sweat and elbow grease
19:58 sess or i'd just use spring that has everything out of the box :)
19:58 sess writing tests for an existing application is crap anyways
20:01 sfisque oooh tasty… openDJ  - ldap in java
20:01 sfisque blogs.forgerock.com/opendj
20:04 sfisque anyone have experience doing the following mech (or suggestions why it would not be good or an alternative) [cont]
20:04 knoppix joined ##javaee
20:04 sfisque use REST endpont to accept a url like    host/{unique_id} -> redirect to deliver a JSF facelet constructed page
20:04 sfisque where the content is driven by the unique_id
20:05 sfisque that way pages can be templated but the url drives "what" is being rendered as content
20:06 sess to what end
20:06 pdurbin sfisque: forgerock? you should join me in #openam :)
20:06 sfisque done
20:07 sfisque sess - i want to be able to do multi-tenancy.  have the url indicate "what", but still decorate with facelets in a templated manner
20:07 sess how would an url not already indicate what
20:08 sess What makes your method different from host/index.jsf
20:08 sfisque index.jsf does not tell me "what' i want to look at, just how to render it
20:09 sfisque lets say i have a site that has dogs, cats, and birds as "what".  and i want to be able to dynamically add more "what" so its not hard coded in the face-config.xml or java code.  so i need a url that indicates the what, and redirects to the "how"
20:09 sfisque without the application knowing the "what" until request time
20:10 sess how would you add more "what" without adding code?
20:11 sfisque for instance, if i have a rest endpont defined as  "category/{unique_category_id}.  the user (or link) can point to  host/category/22394958 and that would redirect to index.jsf after setting some session level vars, that then get rendered in the index.jsf.  since the id allows lookup into some repo, it's not hard coded
20:11 sess the design sounds more fitting for spring mvc either way
20:11 sfisque aye, but i want to try doing it vanilla EE
20:12 sess how would the server know what 2239blah maps to?
20:12 sfisque the bean that handles the rest request can look it up
20:14 sess also it would get a little weird since JSF is stateful
20:14 sfisque the other way would be to hang request params off the url  (?unique_cat_id=234283420) but i'd rather have cleaner urls that do not have to process params
20:15 sess prettyfaces could be of interest perhaps
20:15 sfisque does prettyfaces have a "url dispatch" mech?  afaik, faces does not support that out of the box and primefaces doesnt offer it either from what i see
20:16 sfisque one of those cases where i wish jsf was more like struts/tiles
20:16 sess not sure, never used
20:16 pdurbin I'm definitely interested in prettyfaces. I think the newer version is called rewrite
20:17 sfisque but yah, i remember spring mvc having url->method linkage which was nice
20:17 sess Now with dynamic view ID mapping for URLs (through El method expressions,) you can show different views based on run-time conditions, effectively de-coupling the URL from the JSF view.
20:17 sfisque but i'd rather not go down that route if possible.  :-P
20:17 sess this was perhaps what you were looking for?
20:17 sfisque yeah, but thats EE7, no?
20:18 sess it stands as a prettyfaces feature
20:18 sess and I dont see why they wouldnt be able to do it without EE7
20:18 sfisque oh… then i shall have to look at it
20:18 sess http://ocpsoft.org/prettyfaces/#features
20:18 sess the features does sound temptating
20:19 sess though I love being able to instantly see what xhtml file i should edit when having an url
20:19 sfisque hopefully it does what i want.  which is the opposite of what they indicate.  i dont want to dispatch to templates dynamically, i want to dispatch to data dynamically and decorate in a standard way
20:19 sess ...dispatch to data?
20:20 sfisque basically do the standard struts model of   url -> bean -> render.  which is opposite of jsf which is url -> render  for gets (not posts which does its own magic)
20:21 sess it sounds like you REALLY shouldn't use JSF
20:21 sess if you want to force spring mvc behavior into it
20:21 sfisque e.g.     host/78886766 dispatches to index.xhtml but uses the data that maps to 78886766 whereas host/77777 dispatches to index.xhtml but renders the data that maps to 77777
20:21 Naros yah the action frameworks are nice to map a uri to an action or even a specific method in an action bean
20:22 sfisque personally i would use struts/tiles for this (it's baked into it's model) but i find struts/tiles ajax klunky compared to jsf
20:22 Naros Using the ModelDriven interface in Struts makes decoupling reusable parameter cases to the action dispatch really clean
20:22 Naros sfisque: it isn't too bad if you leverage the json plugin for your ajax-stuff
20:23 Naros There is even a conversation plugin for struts that gives you JSF backing bean functionality
20:24 sfisque oh?
20:24 sfisque i must look into this too
20:24 Naros Yep, look up struts2-conversation
20:24 sfisque nifty
20:25 Naros You basically decorate your action classes with designated annotations that describe the conversation boundaries and what fields in the action class are part of the conversation.  On the start of a convo, it stores it in a session scope and then retrieves it on future requests that use the same conversation id.
20:25 sfisque do i have to pass the conv id around or does it handle that transparently for me?
20:25 Naros so basically a convo id + session id -> lookup map get objects -> rehydrate action -> invoke action
20:26 Naros there is a sc:conversations tag that just embeds it into your forms for you
20:26 sfisque N I C E
20:26 Naros it even has a sc:form that builds atop of struts's form tag to do it too
20:26 Naros and even has a url tag that does the same
20:26 Naros so its just as simple as changing from s:form to sc:form :P
20:27 sfisque h:form
20:27 sfisque :P
20:27 Naros I reworked it in-house for some changes we needed but it has worked really nicely.
20:27 sfisque oh struts.. right
20:27 sfisque s:form
20:27 sfisque lolz
20:27 Naros Yep
20:27 sfisque TY naros.  i will look into this
20:28 Naros Anytime :P
20:28 Naros It certainly bridges the gap between JSF and Action frameworks iimo
20:28 sfisque i really like how jsf streamlines the backing bean linkage, so having that with struts/tiles would be great
20:28 Naros Altho I do like the prettyfaces & primefaces features
20:28 sfisque cool
20:28 sfisque aye.  i'm leveraging prime so i need "jsf" in some capacity under the hood
20:29 Naros Yah, its very similar to annotating the backing bean like in JavaEE with annotations.
20:29 Naros I've tried playing with the Struts2-JSF plugin but haven't gotten it to work too well yet
20:30 Naros Trying to merge an action framework with JSF feels wrong :P
20:30 sfisque i got it to work, but it does not decouple the individual fields very well in the back end
20:30 Naros decouple how exactly?
20:31 sfisque oh wait, not that plugin, it was the struts-ajax plugin i got working
20:31 sfisque basically the ajax requests back to the original rendering method, which can do too much heavy lifting if youre just updating a single field
20:31 Naros If you want ajax/jquery widgets, take a sneak peak at struts2-jquery-plugin
20:32 Naros the taglib for it is pretty slick but there are a fair number of things I've uncovered that are a tad buggy but its open-source and can be tweaked iirc.
20:33 Naros as to your particular test case, I don't see how that would be too much heavy lifting if you want to change a single field
20:33 Naros send ajax request with convo id or record pk plus the field
20:35 sfisque if the method before dispatching to the render layer marshals alot of data, it would remarshall even if we only need a single field repopulated
20:35 Naros json result type?
20:36 Naros you can have json on serialize specific properties of a pojo
20:36 Naros *only
20:36 Naros Many of our grids contain a plethora of fields but we only serialize the 5 or 6 columns of data via json
20:37 pdurbin speaking of JSON, do folks like http://docs.oracle.com/javaee/7/api/javax/json/JsonObject.html as opposed to older JSON libraries? pre-spec libraries
20:37 Naros Haven't looked at it as of yet but anything that lets me get away from third-party apis make me happy :P
20:38 sfisque aye.  i'm happy they finally folded it in
20:38 pdurbin yeah, that's what I was thinking... try to use what's in the spec. it seems fine. there's even a decent example on that page
20:38 Naros I think the jqGrid sends a marshalled JSON string when users want to do searches and I had to exploit the JSON plugin's API to deserialize it into something I could manage in the action framework :E
20:40 Naros anyway sfisque, the key is inside your result type="json" tag, you can use <param name="includeProperties">some comma delimited list or regex of included proeprties</param>
20:40 Naros you can even use name="excludeProperties" to exclude things
20:40 Naros that way your json result to the UI is bloated
20:41 Naros *isnt
20:42 Naros the only issue I've seen with the json plugin is when I serialize out an entity that has references which were not loaded and those nasty lazily load issues happen :P
20:44 sfisque right but that is what gets serialized, not what gets executed.  when i say marshal, i mean  method->repo not UI->method
20:44 sfisque but anyway, this new avenue looks promising to pursue
20:45 Naros Not sure I necessarily follow, but ok :)
20:45 Naros I would presume your talking more about ajax action mapping -> method invocation
20:46 Naros which contains your calls to whatever service/business api you need
20:46 sfisque i believe so.  it's been a few months since i abandoned struts2/tiles2 and went with jsf2
20:46 sfisque i just remember not liking the ability to bind "alternate" methods to components like jsf does
20:47 Naros Well action mappings can include method mappings too.  <action name="someAction" method="mySpecialMethod" class="com.company.packageA.MyAction">...</action>
20:47 Naros you could get fancy and use wildcards too :P
20:48 Naros and you can use wildcards in conjunction with hardcoded mappings.   e.g.: name="editAction" vs name="*Action"
20:49 sfisque anywho, i'm going to look at prettyfaces
20:49 sfisque for now
20:49 Naros ciao
20:49 sfisque laters naros
20:49 Naros you could write this base cost api for me :P
20:49 Naros so I dont have to
20:49 sfisque anywho, on a side note, anyone ever bridged jabber or irc with JMS
20:49 sfisque ?
20:50 Naros That'd be a fun experiment
20:51 sfisque i know AMQ has a bridge, but i was hoping i could find a driver (sar/rar) that would bridge with jbossmq embedded in jboss-as
20:56 sfisque if it's a paying gig naros, i'd be game.  my contract ends in 2 weeks, so i'm actively looking for a new gig
20:56 sfisque :-D
21:10 Naros :)
21:19 cem__ joined ##javaee
21:25 pdurbin sfisque: good luck. I'm sure you'll find something
21:27 sfisque TY pdurbin
21:27 sfisque slow time of the year, but i can spend time working on crossroads (my fb replacement) during the down time ;-D
21:32 pdurbin sfisque: let us know when we can look at it
21:33 sfisque of course
21:40 * cem__ getting scoldings :(
21:50 neuro_sys stab them
21:54 cem__ lol its my elder bro :/
21:57 sfisque even more reason!
21:57 sfisque :P
21:58 cem__ he scolds me often
21:58 cem__ no big deal used to it
22:06 cem__ k 'm back with my computer
22:06 * cem__ yappy!
22:06 Naros hehehe
22:07 Naros Joys of sibling disputes.
22:08 cem__ well i changed something and it was not working so he scolded me not my fault
22:08 * cem__ how can i know :(
22:09 Naros always the lil siblings fault :P
22:09 Naros did that to my sister growing up all the time
22:30 cem__ gn
22:38 knoppix_ joined ##javaee
22:56 sfisque left ##javaee
22:58 sfisque joined ##javaee
23:04 sess joined ##javaee
23:09 knoppix_ joined ##javaee
23:14 knoppix__ joined ##javaee

| Channels | #javaee index | Today | | Search | Google Search | Plain-Text | plain, newest first | summary

Please see http://irclog.greptilian.com/javaee for which days have been logged.