IRC log for #javaee, 2013-10-08

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

All times shown according to UTC.

Time	Nick	Message
00:15		sfisque joined ##javaee
00:39		sfisque1 joined ##javaee
01:03		grug_ left ##javaee
05:04		sess joined ##javaee
05:45		dangertools joined ##javaee
06:00		Moogly2012 joined ##javaee
06:07		sess joined ##javaee
07:23		pdurbin joined ##javaee
09:38		Technodrome joined ##javaee
09:39		Technodrome left ##javaee
09:56		Quest joined ##javaee
10:00	Quest	whats usually the prefered tool to make jdbc pooled connections. is it org.apache.commons.dbcp.ConnectionFactory; import org.apache.commons.pool.ObjectPool; ? no built in libs in java?
10:08	sess	Quest: why not let the container handle it?
10:09	Quest	you dont know the advantages of pooling?
10:09	Faux	The container provides the pool, if you want to do stuff the legacy way.
10:09	sess	how is that relevant? The advantages are the same
10:10	Quest	creating connections is resource wanting
10:10	Quest	having a fix number of pool that expands or contracts. resuing connections already made.... thats what I was refereing when i said connection pool
10:11	Quest	the container dont do it by default i gues
10:11	sess	i know
10:11	sess	and my comment still stands
10:11	Quest	thats why libs like apache come int
10:11	Quest	in*
10:11	sess	im pretty sure containers pools connections by default
10:11	Quest	pool? in what way different from the apache libs pool?
10:11	sess	or it depends on what you inject perhaps
10:12	sess	im not an expert since i just use JPA where thats taken care of behind the scenes
10:12	sess	but its less work just injecting a connection
10:12	sess	¨from a pool that the container defines and takes care of
10:12	Quest	hm
10:13	sess	http://docs.oracle.com/javaee/6/tutorial/doc/bncjj.html
10:13	sess	read some of this
10:13	sess	you just define a data source
10:13	sess	then ask the container for a connection
10:13	sess	or an entitymanager in the case of jpa
10:14	Quest	hm
10:14	sess	or it seems like you inject a datasource
10:14	sess	but you can probably get a connection from that easily
10:17	Quest	how would the container manage the connection pool as stated in your stated link.? I mean how may connections are of default and how can i change things ?
10:17	sess	what container are you using? Jboss?
10:17	Quest	tomcat
10:18	sess	ah
10:18	sess	tomcat even uses apache commons pool
10:18	sess	you can configure it in context.xml
10:18	Quest	sess oh, in the background?
10:18	sess	both connection string etc and pooling config
10:19	Quest	so how many connections and how to change things
10:19	sess	http://tomcat.apache.org/tomcat-6.0-doc/jndi-datasource-examples-howto.html#Database_Connection_Pool_(DBCP)_Configurations
10:19	sess	you can set maxActive, maxIdle, maxWait
10:19	Quest	hm
10:22	Quest	I am just watching a prebuild app that has a class that uses those and the container is tomcat. now if tomcat already does such pooling why the app is doing it again
10:22	sess	perhaps it's old
10:23	sess	i really cant see any reason to handle connection pool by yourself these days
10:23	Quest	hm... it is old. yes
10:24	sess	http://tomcat.apache.org/tomcat-6.0-doc/jndi-resources-howto.html#JDBC_Data_Sources
10:24	sess	this seems like a good introduction
10:24	sess	also shows the code to get a connection
10:24	Quest	sess do you know when apache tomcate started using jdbc connection pooling by default?
10:24	sess	by default compared to what?
10:24	sess	i doubt they ever had a system without pooling
10:25	Quest	hm....... I wonder why it was made by custom means then
10:25	sess	perhaps tomcat 5 or older did not support datasources at all
10:25	sess	and you had to implement them by yourself
10:26	Quest	sess does that means that in that app, tomcat will also pool connections no matter the said class pools it or not. so that means there will be 2 pools but on main real pool (tomcats default
10:27	sess	tomcat only creates a pool if you configure it too
10:27	sess	it would be impossible for it otherwise since it wouldnt know anything about the database
10:27	Quest	oh. so tomcat only creates a pool if you configure it to make a pool. otherwise it does not creates a pool?
10:27	sess	correct
10:27	Quest	right.
10:28	Quest	so thats why I think they did it. so the app gets generic even if jboss or some other container replaces tomcate later on.
10:28	Quest	correct guess^?
10:29	sess	even using tomcats datasources, switching container would be extremly easy
10:29	sess	just configure the datasource in jboss instead
10:29	sess	no code changes needed
10:29	Quest	hm
10:29	Quest	which one would be better? I think custom made as we can introduce more custom code in it.
10:30	sess	personally, I dont see any reason to not let the container handle it
10:30	sess	less chance of doing things wrong
10:30	Quest	ya
10:30	sess	im not sure what kind of custom code you would want there
10:30	Quest	that too
10:30	sess	and if custom code is a good idea
10:30	Quest	e.g counting no. of connections used
10:31	sess	if bet you can configure the container for stuff like that
10:31	Quest	sysouting i mean
10:31	sess	or asking the pool
10:31	Quest	hm.
10:31	Quest	ok..
10:31	sess	also you still write the code that gets the connection from the data source
10:31	sess	¨so you can do stuff there
10:32	sess	the cool thing with containers is also that if you want to deploy a 2nd application using the same database
10:32	sess	they can share the same connection pool
10:33	Quest	hm
10:34	Quest	thanks for enlightening
10:35	sess	np :)
10:35		MegaMatt joined ##javaee
12:14		Quest joined ##javaee
12:20	Quest	silly question as static has no instance. but looking for tricks. -ChanServ- Flags +o were set on sess in ##javaee.
12:20	Quest	shit.
12:21	Quest	silly question as static has no instance. but looking for tricks. Is there a way to make this line of code generic for any class. ? public static Logger logger = Logger.getLogger(UserRegAction.class);
12:22	sess	not that i know of, i usually have that in each class
12:22	Quest	k
12:23	Quest	sess as a default behaviour of logger.debug("message"); only sys outs that to console and not wirte it on any file. correct ?
12:23	sess	not sure
12:23	sess	i dont remember log4js default config
12:24	Quest	k
12:26	tjsnell	depends on logging framework
12:28	Quest	why use it on each class. why not just declare in one class and inherit all other classes from it?
12:28	Quest	tjsnell log4j
12:28	Quest	is a framework
12:29	sess	Quest: because then its all the same logger
12:29	sess	you might aswell declare Log.java then and have a public static logger there
12:32	sess	you declare it in each class to get categories
12:32	Quest	sess why not the same logger?
12:33	sess	because you lose categories then
12:33	Quest	oh. so its only for catagorization
12:33	sess	lets say you want debug logging
12:33	sess	for a single class
12:33	Quest	k
12:33	sess	thats impossible if they share category
12:33	tjsnell	it's exactly for categorization
12:33	tjsnell	that's vital
12:33	Quest	sess do you have a log that you can show me made by log4j with many catagories?
12:33	Quest	as you said you use it a lot
12:33	Quest	tjsnell k
12:34	tjsnell	any log4j log file
12:34	Quest	ya. any
12:34	tjsnell	no
12:34	sess	Quest: use what a lot?
12:34	tjsnell	any log file written by anyone with a smidgen of a clue has many catagories
12:34	sess	yes
12:34	tjsnell	if one of my devs used the same thing in every logger instance I'd probably fire them
12:35	tjsnell	but then again I would've never hired them
12:36		Quest_ joined ##javaee
12:37	Quest_	disconeected
12:37	Quest_	Quest> can i see a sample
12:37	Quest_	* Disconnected
12:38	* Quest_	says by to Quest_
12:38	* Quest_	says by to Quest
12:38	* Quest_	says bye to Quest
12:38	tjsnell	start tomcat
12:38	tjsnell	look at log file
12:39	Quest_	ah.
12:40	* Quest_	forgot tomcat also uses log4j
12:42		tommmied joined ##javaee
12:44	acuzio	Quest: is tjsnell again creating problems ?
12:47		jieryn joined ##javaee
12:47		jieryn joined ##javaee
12:51	tjsnell	I'm awake
12:51	tjsnell	so yeah
12:59	Quest	acuzio :) I think you both should shake hands some day
13:06		kotza joined ##javaee
13:40		JAZ1976 joined ##javaee
13:40		Naros joined ##javaee
13:41		tommmied joined ##javaee
13:43		kobain joined ##javaee
14:09		kobain_ joined ##javaee
14:13	pdurbin	whartung: you've done https://wiki.shibboleth.net/confluence/display/SHIB2/NativeSPJavaInstall before?
14:46		Quest joined ##javaee
14:47	acuzio	shake hands with tjsnell never -
14:47	Quest	acuzio, :)
14:47	tjsnell	we tend to hug
14:48	Quest	iam getting 405 Method not allowed and 406 not acceptable error for getting a json object. http://pastebin.ca/2464130
14:48	acuzio	Quest: wow json
14:49	sess	Quest: try removing headers="Accept
14:49	Quest	did that before too
14:49	sess	also
14:49	sess	your urls looks wrong?
14:49	sess	your mapping specifies 4 parts
14:49	Quest	wrong. no i gues
14:50	Quest	which part?
14:50	sess	while the urls in the top specifies only 3
14:50	sess	create/board/3/vertical
14:50	sess	create/{parent}/{parentId}/{boxType}/{boxTitle}
14:50	Quest	no
14:50	Quest	create is static. all others are values
14:50	sess	i did not say it was
14:50	sess	count the /
14:51	Quest	4 /s
14:51	sess	on the bottom
14:51	sess	3 on the top
14:51	Quest	sorry, didnt got you
14:51	sess	so there doesnt seem to be any mapping for that url
14:52	Quest	url: "${pageContext.request.contextPath}/box/create/board/3/vertical/MyBoxTitle", justifies @RequestMapping (value = "/box/create/{parent}/{parentId}/{boxType}/{boxTitle}"
14:52	sess	on the other hand it looks like 4 in the javascript
14:52	sess	so whats up with the requests logs at the top?
14:53	acuzio	sess: are you Fubar^ ?
14:53	sess	uh no?
14:53	acuzio	are you the new op ?
14:54	sess	"the" is a strong word
14:54	acuzio	i thought "you" was strong - when did you become an op ?
14:54	sess	earlier today
14:54	Quest	sess, detailed error http://pastebin.ca/2464131
14:55	acuzio	sess: is it your birthday ?
14:55	sess	acuzio: no
14:55	acuzio	sess: so whats the occasion .,
14:55	sess	Quest: do you have jackson included?
14:55	sess	as a library
14:55	acuzio	Michael is a library /
14:55	sess	acuzio: none really
14:56	acuzio	sess: who gave you the op ?
14:56	sess	Quest did
14:56	Quest	<groupId>org.codehaus.jackson</groupId>
14:56	Quest	<artifactId>jackson-mapper-asl</artifactId>
14:57	Quest	only included in maven pom
14:57	acuzio	Quest: why is sess an op ?
14:57	sess	Quest: try removing @ResponseBody and replacing it with ModelAndView
14:57	sess	and see if you still get 406 errors
14:57	sess	acuzio: what's the big deal?
14:58	Quest	sess, hm.
14:58	acuzio	sess: do you not know ? in this channel being an op is a great honor , plus you can kick tjsnell whenever
14:58	Quest	lol
14:58	sess	sounds amazing
14:58		jieryn left ##javaee
14:59	sess	whoever that is
14:59	Quest	sess, so i should return ModelAndView and remove @ResponseBody?
14:59	acuzio	do you not know the history of this channel sess ?
14:59	sess	Quest: yeah, try to use it like it was a normal method for a normal page
14:59	sess	acuzio: i joined like a few days ago
15:00	Quest	sess, i wonder why on line 6 of http://pastebin.ca/2464131 . its post and not get
15:00	acuzio	sess: aaah , that explains it
15:00	sess	oh yeah i thought i saw POST somewhere, then looked at the js that was a GET and forgot about it
15:00	Quest	so whats wrong
15:00	sess	perhaps your "type" is wrong
15:00	Quest	type?
15:01	Quest	its type:"get"
15:01	sess	in jquery
15:01	sess	get is default so it doesnt have to be specified even
15:01	sess	otherwise it looks correct hmm
15:02	sess	Quest: if you just type the url in your browser
15:02	sess	does it work?
15:02	Quest	i would comment out type: get then and try
15:02	sess	i.e json should show up as the page
15:02	Quest	hm wait
15:03	tjsnell	unlike others bans and kicks don't get my panties wadded
15:03	Quest	haha. HTTP Status 500 - Could not write JSON: failed to lazily initialize a collection of role: web.entity.Boards.childBoxList, could not initialize proxy - no Session (through reference chain: web.entity.Boxes["parentBoard"]->web.entity.Boards["childBoxList"]); nested exception is org.codehaus.jackson.map.JsonMappingException: failed to lazily initialize a collection of role: web.entity.Boards.childBoxList, could not initialize proxy - no Session (throu
15:03	Quest	gh reference chain:
15:03	Quest	sess, i typed the link on url bar directly ^
15:03	sess	ok so if you fix that the problem just seems to be with jquery
15:04	acuzio	Quest: dont use JSON , its for the gays
15:04	Quest	sory?
15:05	Quest	sess, with jqery
15:05	acuzio	its for the gays
15:05	sess	Quest: something in your code causes you to post rather than GET
15:05	Quest	acuzio, i mean sess. what do you mean by "seems to be with jquery"
15:05	Quest	a minut
15:05	sess	as in youre using jquery to make the request
15:06	Quest	sess, but I have an other error 500 too
15:07	sess	didnt we go over lazy initialization a few days ago
15:07	Quest	yes. i got that but might it not be the reason only and not jquery
15:08	sess	no, because when doing the ajax you get a 506 error
15:08	sess	while this gives a 500 error = some random exception was thrown
15:08	sess	506 means something is wrong with the mapping
15:09	Quest	i commented out type: get in jquery and now I only get 500 (Internal Server Error)
15:09	Quest	I would fix it and return
15:09	Quest	sess, thanks!
15:10	sess	sounds like that was it then
15:10	sess	perhaps it is case sensitive
15:10	Quest	yes
15:10	sess	so type: "GET" was required
15:10	Quest	hm. may be
15:10	sess	or none at all
15:41	Quest	sess, this controller method should not give a lazy initialization error. I have done all in service transactional boundaries. http://pastebin.ca/2464157
15:42	sess	Quest: which line throws the exception?
15:43	Quest	dont see the line
15:43	Quest	in the error
15:43	sess	uh your service class does basicly nothing now
15:44	sess	why not set the parent in the controller
15:44	sess	and you skipped the important method: getBoardById
15:47	Quest	this will be a lot of code http://pastebin.ca/2464162
15:47	Quest	sess, why the service does nothing now?
15:47	sess	telling the line where the exception occured
15:47	sess	would help
15:47	Quest	I dont see the line in the error
15:47	sess	Quest: youre making a whole method for just assigning a simple variable
15:47	sess	seems unnecessary
15:47	sess	no line number in logs?
15:47	sess	in that case you are logging in a bad way
15:48	Quest	heres the full error http://pastebin.ca/2464163
15:48	Quest	sess, its tomcate logging. not mine
15:48	sess	oh its Jackson that throws it
15:49	sess	Jackson tries to convert all data
15:49	sess	to json
15:49	sess	so you have to make sure everything is initialized
15:49	Quest	or set to null
15:49	Quest	hm
15:49	sess	that would be a hack, but that would work in theory yes
15:50	sess	perhaps you can annotate for jackson that you do not wish to convert some lists
15:50	Quest	sess, but why jackson is worried about lazy init and sessions............ Could not write JSON: failed to lazily initialize a collection of role: web.entity.Boards.childBoxList, could not initialize proxy - no Session
15:50	sess	or you can make another class that Jackson can convert
15:50	sess	that does not contain the list
15:50	sess	because jackson is reading your entity
15:50	sess	and its not initialized
15:51	sess	calling getCollection will always throw a lazy exception if its not initialized or fetched in the current transaction
15:54	Quest	hm
15:54	sess	I think the proper design here would to make another class
15:54	sess	BoxBean or whatever
15:54	Quest	rebuilding
15:54	sess	that you populate in the service layer
15:55	sess	that way, lazy initialization exception cant occurs
15:55	sess	-r
15:55	sess	-s
15:59	Quest	changed the controller but still same error http://pastebin.ca/2464167
16:00	sess	Quest: you could also use something like this https://github.com/FasterXML/jackson-datatype-hibernate
16:00	sess	i think it can ignore non loaded collections
16:01	sess	im not sure the set null method actually works
16:01	Quest	ya but i wonder why its giving lazy init error now when all collections are null in the returned Boxes object?
16:02	Quest	it should work. why not
16:02	sess	i think its the implementation of Boxes that decides when the exception should be thrown
16:02		sfisque joined ##javaee
16:02	sess	rather than the value of the collection
16:02	sess	or something
16:02	sess	not sure of the specifics
16:02	whartung	pdurbin: I don't think I even installed the Shib servlet handler, and if I did it was 4-5 years ago
16:02	Quest	hm
16:02	sess	either way it's an ugly solution
16:02	Quest	sess, last resort is to try a bean now
16:02	Quest	right?
16:03	sess	you could also check out my link
16:03	sess	that disables json for uninitialized collections
16:03	Quest	which class in it
16:03	sess	http://stackoverflow.com/questions/7440613/is-there-a-way-to-make-jackson-jsonize-only-fetched-lazily-loaded-objects
16:03	sess	https://github.com/FasterXML/jackson-datatype-hibernate
16:04	Quest	@JsonIgnore might help?
16:04	sess	dunno, never saw that annotation
16:05	Quest	public @ResponseBody @JsonIgnore Boxes createBox
16:05	Quest	k
16:06	sess	that doesnt look right
16:06	sfisque	so here's an interesting question because i cannot find definitive documentation. so, with the recent JSRs for EE, you can dual purpose a class as both @Stateless(ful) AND @Namee with a CDI scope (e.g. @SessionScoped). i'm curious how that plays out since EJB "state" is not coupled to the web session. i'm probably going to build a test war and check it out, but i'm curious what others have found doing such a thing
16:06	whartung	I haven't got that far in to EE7 to explore that sfisque
16:06	pdurbin	whartung: no worries... I'm getting a little bit of help at http://irclog.perlgeek.de/shibboleth/2013-10-08
16:06	pdurbin	sfisque: you mean @Named, right?
16:06	sfisque	yes
16:06	sfisque	typo
16:07	whartung	Would think that even though they're the same "bean", they'd be two different instances in this case.
16:07	sfisque	that's my conceptual take
16:07	sess	probably no JTA activated?
16:07	sess	on the CDI one
16:07	sfisque	because we have a dev here who wants to tag all the EJBs as @Named to work around something he's working on.
16:07	Quest	sess, hm. will try the bean resort after dinner break. thanks again!
16:08	sfisque	and i'm very resistant to it
16:08	sfisque	i see it playing huge havoc
16:08	sess	never heard of anyone using @Named on and EJB
16:08	sess	an*
16:08	sfisque	me either
16:08	whartung	yea I just don't have the CDI experience to really explore that. what problem is he trying to solve?
16:08	sess	either way it sounds horribly wrong
16:08	sfisque	but they do offer an example of it in the JSR discussion exampe on oracle.com
16:09	sfisque	he's trying to create an anotation that does some AOP stuff, and it doesn't work properly on EJBs (because reflection is dicey since its' all pooled and heavy proxied)
16:10	sfisque	basically (imo) misusing a platform feature
16:10	sess	well if it doesnt work with EJBs, using @Named will probably fuck it up badly
16:10	sfisque	as a workaround
16:10	sfisque	that's my thought
16:10	sfisque	newbies…. /sigh
16:10	sess	but wait, AOP doesnt work on EJBs?
16:10	sfisque	:P
16:10	sfisque	it woudl depend on how/what you're doing
16:11	sess	ah you mean he's doing something very specific
16:11	sfisque	reflection is "tricky" when you cross the boundary into ejb
16:11	sfisque	yes
16:11	sfisque	and aop relies heavily (generally) on reflection
16:11	whartung	yea, I can see AOP having kittens with EJBs, but then we have Interceptors, which are EE blessed AOP for EJBs
16:11	sess	I assume he was using interceptors
16:12	whartung	I should say, having kittens with INSTANCES of EJBS
16:12	whartung	but he may be doing some dynamic on the fly shenanigans with AOP, interceptors tend to be a bit static
16:12	sfisque	he's not using the "mandated" lifecycle interceptor. basically trying to wire up his own
16:12	whartung	ok. why?
16:12	sess	oh that sounds
16:12	sess	lovely
16:12	whartung	haha
16:13	sfisque	because he's young and tries to do "cute" stuff
16:13	sfisque	you know the profile
16:13	whartung	o
16:13	whartung	k
16:13	sess	stuff made by the nice java people is obviously bad
16:13	sess	so you should make your own implementations
16:13	sess	of everything
16:13	* sfisque	nods solemnly to sess
16:14	whartung	that behavior can be fixed… http://www.amazon.com/Remote-Dog-Training-Shock-Collar/dp/B003YLETZY
16:14	sfisque	ROFL
16:14	sfisque	i was thinking more along the lines of a LART
16:15	sess	:D
16:15	sfisque	like… say… a toaster oven to the forehead
16:15	sfisque	it gives such a satisfying >thud<
16:15	whartung	or an old candy colored iMac
16:15	sfisque	:-)
16:43	pdurbin	just moved from ant to maven: https://github.com/IQSS/dvn/commit/8ca34ad
16:44	sfisque	irc, there is a betty ford program for that
16:44	sfisque	:P
16:44	sfisque	ant. the gateway drug
16:46	pdurbin	:)
16:47	pdurbin	death to ant
16:47	pdurbin	;)
16:47	whartung	I like ant
16:47	pdurbin	i know i know
16:47	sfisque	i'm ok with maven, but i do perfer ant
16:47	sess	maven should have been named ant eater
16:47	whartung	for the most part I like imperative systems over declarative ones
16:47	sfisque	maven automates better, ant gives more power
16:47	sfisque	ROFL sess
16:47	sess	with great power
16:47	sess	comes great idiots
16:48	sfisque	hey, i resemble that remark!!
16:49		tommmied1 joined ##javaee
16:55	pdurbin	now to re-write our dev guide. should be shorter :)
17:02	MegaMatt	pdurbin, wonder if you know my friend Vanessa Venti ..
17:03	pdurbin	javaeebot: lucky Vanessa Venti
17:03	javaeebot	pdurbin: http://www.linkedin.com/in/vanessaventi
17:03	pdurbin	MegaMatt: nope
17:04	MegaMatt	you must have common linked in connections via harvard, I'd assume
17:04	pdurbin	MegaMatt: "3rd" it says
17:04	MegaMatt	right on
18:01	Quest	sess, it works with an entity bean but not with the real entity. http://pastebin.ca/2464227
18:01	sfisque	entity bean vs real entity
18:01	sfisque	what does that mean?
18:01	sess	isnt entity beans ejb2 stuff
18:01	sess	or even older
18:01	Quest	well.
18:02	sfisque	aye, but what is a "real entity" then?
18:02	sfisque	vs an entity bean
18:02	Quest	sfisque, my entity has some lists in it. even if i make them null and not fetch them. it gives the stated 500 error. but if i return an identical bean instead. it works fine.
18:02	Quest	any help for this spring jackson, and lazy initialization error? http://pastebin.ca/2464227
18:02		bruce-zu joined ##javaee
18:03	sess	i already told you the reason
18:03	sess	and the solution
18:03	Quest	sfisque, real would have @Entity and other would be plain pojo
18:03	Quest	sess, but there should be some better solution
18:03	sess	you could load the data
18:03	sess	or use the plugin i linked
18:04	sess	3 solutions, all of them fine
18:04	Quest	sess, load data> what if theres no data in the child lists
18:05	sess	doesnt matter
18:05	sess	try to load it
18:05	Quest	load? how
18:05	Quest	in service methods?
18:05	Quest	what do you mean by load?
18:05	sess	yes
18:05	sess	using the getter
18:05	Quest	hm
18:06	sess	before closing the transaction
18:06	Quest	why not just make the childLists box.setChildList(null) ; in service method?
18:06	Quest	that would suffice too i gues?
18:06	sess	no
18:07	sess	that might delete data
18:07	sess	from the db
18:08	sess	Object o = entitymanager.find(myClass.class, id)
18:08	sess	Hibernate.initialize(o.getLazyCollection)
18:08	sess	return o
18:08	sess	done
18:10		syncsys_ joined ##javaee
18:11	Quest	disconnected after <sess> no
18:11	Quest	<Quest> am?
18:11	sess	[20:08] <@ sess> \| that might delete data \|
18:11	sess	[20:08] <@ sess> \| from the db \|
18:11	sess	[20:09] <@ sess> \| Object o = entitymanager.find(myClass.class, id) \|
18:11	sess	[20:09] <@ sess> \| Hibernate.initialize(o.getLazyCollection) \|
18:11	sess	[20:09] <@ sess> \| return o \|
18:11	sess	[20:09] <@ sess> \| done
18:11	Quest	hm
18:14	Quest	one question:. my app is of such a strategy that one box may have more boxes. and those may have more. and it goes on and on. Now if i get the box object for json and if its not getLazyCollection . I would need to get ALL levels of child box list(S) in order for jackson not to make any errors?
18:17	sfisque	so you're calling a VO a "real entity".
18:18	sess	Quest: then initialization might not be the best solution
18:18	Quest	VO?
18:18	sess	checkout that jackson hibernate plugin
18:19	sfisque	VO == value object. a pojo that is just for holding data (no business methods)
18:20	Quest	sfisque, VO != entity . VO.setVar(entity.getVar);
18:21	sess	sfisque: he never claimed a vo was an entity
18:21	sess	i think
18:21	sfisque	a VO can be an entity (if you bear in mind that the only thing that makes an entity special is the annotations, and all there really is data + mutator/accessors) but VO's generally are more "context driven" and don't necessarily map to a schema
18:21	Quest	Vo is same identical to entity but without @Entity
18:22	sfisque	you may have a RestPerson bean that is a VO that subsets your Person entity with just the fields the Rest endpoint is exposing
18:22	sfisque	and a ViewPerson that is used in a jsf page that has a different subset of fields
18:23	sfisque	and a MessagePerson that is a JMS payload with a different subset of fields
18:23	sfisque	they're all pojos but the entity maps to your schema, and the VO's probably do not
18:24	sfisque	but in a simplified app, you can get away with treating your entity as a VO
18:24	sfisque	because that's really all it is + some annotations to make it special
18:24	Quest	sfisque, thats what was creating problem if you see my code. I dont want double classes and hasle
18:25	sfisque	you have to be careful though. some EE annotations can play havoc with other thirdparty endpoint frameworks.
18:25	sfisque	that's why VO's are generally better. it strips off the JPA/JTA stuff and just hands a pojo to whatever facility is doing the rendering/export/etc.
18:26	sess	´for filling them with data, there are mapping frameworks
18:26	sess	that can copy via reflection
18:26	sess	because copying values to a VO is a bitch
18:27	sfisque	aye, apache digester is one such lib
18:27	sess	that said i enjoy letting my true domain classes flow throughout the application as much as possible
18:27	sfisque	jackson is another
18:27	sess	jackson does bean to bean mapping?=
18:27	sess	thought it was only bean to json
18:27	Quest	sfisque, I can just do Entity e = getEntityBySomeServiceMethod () Entity VOe = e; return e;
18:27	sfisque	oh, you meant bean to bean
18:27	sfisque	i misunderstood
18:27	Quest	return VOe; *
18:28	sess	uh
18:28	sess	that does nothing
18:28	sfisque	aye, that just assigns the pointer
18:28	sess	ive seen you try something like that before, i think you missunderstand something basic in java
18:28	sfisque	you'd have to do a deep clone
18:28	sess	VOe is still e
18:28	sess	the same instance
18:28	sess	changes will apply to both of them
18:29	sfisque	that's the annoying part of java, it hides the danger of pointers
18:29	sess	not really annoying
18:29	sess	and having pointers here wouldnt help
18:29	sfisque	well, it would be more obvious to him what that assignment did
18:30	sess	it would do the same thing in c
18:30	sess	unless it was a primitive
18:30	sfisque	in C yes, but C isnt oo. in c++ you can override the operator to do deep clone if you wanted to
18:31	Quest	sfisque, I can just do Entity e = getEntityBySomeServiceMethod () Entity VOe = e.clone(); return VOe;
18:31	sess	overloading = sounds like an extremly bad idea
18:31	sfisque	depends.. clone() is protected
18:31	sess	clone returns an instance of the same class
18:31	sess	so that wont help
18:31	sess	or well you could return an instance of the same class with null in the collections i guess
18:32	sess	could work in theory
18:32	sfisque	it woul still fail unless they share an interface or one extends the other
18:32	sfisque	class cast exceptin
18:33	sess	what would be casted?
18:33	sfisque	VO could extend entity and make certain accessor/mutators private
18:33	sess	in Entity e = e.clone()
18:33	sess	if it returns Entity
18:33	sfisque	clone() returns an object
18:33	sess	i mean skipping the VO class
18:33	sfisque	it's defined in Object
18:33	sfisque	unless they changed it to be <T> in recent iterations
18:33	sfisque	or <?>
18:33	sfisque	<T> would still have classcast issue
18:34	sess	probably not
18:34	sess	but just cast it then
18:34	sess	Entity e2 (Entity)e.clone()
18:34	sfisque	still throws excetpion unless they are in the same inheritence change
18:34	sfisque	***chain
18:34	sess	though its a bad solution to null the collections anyways
18:34	sess	its the same class :*V
18:34	sess	thats pretty close in the inheritance chain id say
18:34	sfisque	you cant do Point p = (Point) r.clone // if r is a Rect
18:34	sess	no
18:35	sess	im speaking of cloning the SAME class
18:35	sfisque	right but then doing an assignment would fail
18:35	sfisque	if the result is not in the same chain
18:35	sess	god
18:35	sess	the result is the same class aswell
18:36	sess	Entity e1 = entitymanager.find(Entity.class, 1);
18:36	Quest	I think in order to clone an object i need to implement clonable interface. I think iam better of with other solutions
18:36	sfisque	i was referring back to his VO v = entity.clone();
18:36	sess	Entity e2 = (Entity)e1.clone();
18:36	sess	VO would not work with clone no
18:36	sfisque	which woudl only work if VO extended En or a shared interface
18:36	sess	yeah
18:37	sfisque	ok. now we understand each other :-)
18:37	sfisque	we had a disconnect
18:37	sfisque	i wasnt disagreeing with your example, i was referring back to Q's ex
18:42	* Quest	realizes he have many concepts to understand as a fresher
18:43	Quest	it is making very clear sense now by the way of concepts learned here
18:44	sfisque	:-)
18:48	Quest	If on line 25 of http://pastebin.ca/2464227 i did box.setType(boxType); and it did worked fine with no errors. so i could also do box.setChildBoxList(null); on line 26 also ?
18:49		tommmied joined ##javaee
18:49	sess	well it's a newly created instance
18:49	sess	its already null
18:49	sess	unless you create one in its constructor
18:50	Quest	if it was already null, why lazy init error then? iam have confused again :)
18:52	sess	because you are putting objects from hibernate
18:52	sess	in it
18:52	sess	you can only get lazy init after that
18:52	Quest	oh
18:52	sess	boxService.setParent(box, parentBox);
18:52	sess	jackson starts making json of box
18:52	sess	finds parentbox
18:52	sess	traverses it
18:52	Quest	once i touched the new object by some service/hibernate method. its then an attached entity
18:52	sess	find uninitialized collection
18:52	sess	bam, exception
18:52	sess	no
18:53	sess	nothing is attached in the controller
18:53	Quest	got it
18:53	sess	if by attached you mean managed
18:53	Quest	managed. yes
18:53	Quest	but i got your point
18:53	sess	it stops being managed when you leave the service
18:53	sess	if it was managed, lazy loading would work :)
18:53	Quest	oh
18:54	Quest	so jackson scans all the code
18:54	Quest	in the controller method
18:54	Quest	<sess> boxService.setParent(box, parentBox);
18:54	Quest	<sess> jackson starts making json of box
18:55	sess	no
18:55	sess	jackson scans all fields in b
18:56	sess	in this case, most of your data is thrown away since you return your VO instead
18:56	sess	with only ID and title populated
18:56	Quest	but if one field is null and its a list. why jackson makes problem
18:56	sess	in this case it shouldnt
18:57	sess	since b is not an entity even
18:57	Quest	ya this case works fine
18:57	Quest	but iam in context to the error case
18:58	sess	i dont have the code for the error case
18:58	Quest	this one http://pastebin.ca/2464167
19:01		kobain joined ##javaee
19:01	sess	as we saw before, setting null doesn't remove the initialization error
19:02	Quest	why so
19:02	sess	its complicated
19:02	sess	its controlled by the parent of the collection
19:02	sess	rather than the collection
19:02	sess	so the value of the collection doesnt matter
19:02		scripty joined ##javaee
19:02	Quest	you said because jackson is looking at the entity. the managed/attached entity. well thats not in controller
19:02	sess	anyways
19:03	sess	https://github.com/FasterXML/jackson-datatype-hibernate
19:03	sess	just get this
19:03	sess	and get it over with it
19:03	scripty	Hey Quest
19:03	sess	it should solve your problem perfectly
19:03	Quest	sess, just conceptulizing.
19:03	Quest	sess, ok.
19:03	Quest	scripty, hi
19:03	Quest	sess, real big thanks for a lot of your time!
19:04	sess	cp
19:04	Quest	cp?
19:04	scripty	My have started java recently completed app in node.js
19:04	scripty	I*
19:04	Quest	scripty, ok.
19:05	Quest	thats nice
19:05	scripty	Yes
19:05	Quest	sess, this also works fine. Boxes boxToBeReturned = new Boxes();
19:05	Quest	boxToBeReturned.setTitle(savedBox.getTitle());
19:05	scripty	Updated my Git ;)
19:05	Quest	hm..
19:06	Quest	sess, now the only choice i have to make is either do all that in controller or make some recursive method in service class or use hibernate.getLazyInit
19:06	Quest	sess, or ofcourse the utility you refered
19:10	Quest	scripty, are you selling your app
19:11	Quest	i mean making any kind of money?
19:11	scripty	no i do only open source
19:11	Quest	whats it about?
19:12	scripty	video chat real time app
19:12	Quest	which requires no middle server?
19:12	Quest	correct?
19:12	scripty	middle server ?
19:13	scripty	It requires a server
19:13	Quest	client 1 - server - client 2
19:13	Quest	?
19:13	scripty	so i built a server for it and xchat kind client for it
19:13	Quest	if 2 people need to chat. would any third machine be involved?
19:14	scripty	client 1 - server- client 2
19:15	scripty	It works on LAN and WAN
19:16	Quest	so you do need a middle server for 2 clients to communicate with each other?
19:16	scripty	Yes
19:17	scripty	I just built for office use
19:17	Quest	why not make the initiating client as client and the other as server. both would / can have server/client capabilities. eliminat the need of middle server? i think your answer would be "problems with nat / firewalls ?"
19:18	scripty	huh ?
19:18	scripty	firewalls must be taken care by users
19:19	scripty	i provide SSL
19:19	Quest	ok . so why not just eliminate the need of middle server?
19:19	Quest	each client can also act as a server. cant it
19:19	scripty	no way
19:20	Quest	why
19:20	scripty	please explain your theory
19:20	Quest	why not make the initiating client as client and the other as server. both would / can have server/client capabilities. eliminat the need of middle server?
19:20	scripty	i dont want to decentralize
19:20	scripty	no way
19:21	Quest	ever so called client would be listening for connections in a serversocket and at the same time scanning the network for others clients.
19:22	scripty	yea
19:23	Quest	so, need of server is gone. isnt it
19:23	Quest	so, need of one dedicated server is gone. isnt it
19:23	Quest	no need to route all trafice through one server
19:23	scripty	server to centralize things isnt it ?
19:24	MegaMatt	How would client 1 know how to find client 2 with out a server to tell it where it lives?
19:24	MegaMatt	But after that, DCC is fine by me
19:25	scripty	server to centralize after that apply sockets , for example finding client 2 remote ip ; MegaMatt said
19:25	Quest	scripty, yes. if you want to centralize. then server is ok. do you just want all the traffice data to go through server or is there any other purpose of server?
19:26	Quest	MegaMatt, every so called client would be listening for connections in a serversocket and at the same time scanning the network for others clients.
19:26	scripty	currently it goes to the server
19:26	MegaMatt	Scanning a WAN for other clients is not a good idea, imo
19:27	Quest	wan Is an other story
19:27	MegaMatt	How about IPv6?
19:27	Quest	MegaMatt, scripty no. the clients ping the server . the server cant find clients on wan
19:28	scripty	WebSocket is a protocol providing full-duplex communications channels over a single TCP connection
19:28	MegaMatt	scripty:
19:28	MegaMatt	It works on LAN and WAN
19:28	scripty	yes
19:28	MegaMatt	Sorry, that was a quote
19:28	scripty	you need to set it up
19:28	MegaMatt	I'm telling Quest -- having a client that is just scanning the network for servers … yuck.
19:29	MegaMatt	Why bother doing that? Why not just have a server who does all the housekeeping for you?
19:29	MegaMatt	And then just do Direct Client to Client after that?
19:29	MegaMatt	I think that's much cleaner.
19:29	MegaMatt	Anyhow
19:29	MegaMatt	Not my discussion , and I have other things to do :)
19:30	MegaMatt	Let you to have at it again
19:30	MegaMatt	*two
19:30	scripty	how you will provide SSL security then ?
19:30	MegaMatt	What's any of that have to do with lack of or not of SSL?
19:30	MegaMatt	you can do SSL from client to server… and from client to client
19:31	Quest	MegaMatt, yes. direct after that is better choice. like skype does for file transfers
19:31	Quest	if iam correct
19:31	scripty	as of now in server i check ip address , log it , user data
19:31	scripty	SSL , all the interaction takes place using socket
19:31	Quest	MegaMatt, but a point is "firewall issues between clients"
19:31	scripty	rest all**
19:32	MegaMatt	Firewall between the clients is an issue even in client -> server -> client ;)
19:32	MegaMatt	But since both clients would know their network config in connecting to server
19:32	MegaMatt	You can figure out the best way to do the client to client
19:34	MegaMatt	anyhow.. :) .. afk
19:34	Quest	MegaMatt, how skkype solves it then when skype makes direct link between 2 clients?
19:35	scripty	~java
19:37	MegaMatt	Quest: or IRC when doing a /dcc ;)
19:38	MegaMatt	the nice thing about having a server in the middles is that .. as client1 I don't have to ever tell client2 my IP .. I can stay hidden
19:38	MegaMatt	If I don't know/trust you, I might want to keep my IP away from you - and only show it to the trusted server guy :)
19:39	scripty	more of security
19:39	Quest	MegaMatt, yes. but in dcc in irc and file transfer on skype. ip is knows
19:39	scripty	but in my app you can use whois :D
19:39	Quest	known
19:42	MegaMatt	A server can always lie about whois information ;)
19:42	MegaMatt	But yes, DCC is how you find the truth ;)
19:42	MegaMatt	So there's always a trade-off
19:43	MegaMatt	It's up to you to make the best decisions for your situation/environment/etc
19:43	MegaMatt	Now I'm really AFK :)
19:43	scripty	whois is a lie if you mask it or implemented to lie ;)
19:44	Quest	how skype eliminates the firewall isue
19:45	scripty	like a trusted user ?
20:10	Quest	no. i meant while 2 users transfer files
20:10	Quest	they get direct with each other that moment. if there are firewalls in the middle. that would be problematic
20:31	scripty	I dont believe they get a direct connection
20:31	scripty	its thru server
20:32	scripty	i gave def above about soc
20:36	tjsnell	skype uses hole punching
20:36	tjsnell	http://www.h-online.com/security/features/How-Skype-Co-get-round-firewalls-747197.html
20:37	Faux	I meant to learn how STUN works properly at some point.
20:42	pdurbin	whartung: still fighting shib. day 2: http://irclog.perlgeek.de/shibboleth/2013-10-08 :)
20:43	whartung	yay
20:43	MegaMatt	pdurbin - you use shib and openAM?
20:44	pdurbin	MegaMatt: I tried OpenAM a bit but you have to remove packages from glassfish, which weirds me out
20:45	MegaMatt	Ahhh.. I think you posted that on the OpenAM alias, didn't you?
20:45	MegaMatt	Try TomCat :)
20:45	* pdurbin	sighs
20:45	pdurbin	MegaMatt: from a high level, this is what I'm up to: https://lists.iq.harvard.edu/pipermail/dvn-auth/2013-September/000000.html and https://lists.iq.harvard.edu/pipermail/dvn-auth/2013-October/000001.html :)
20:46	MegaMatt	Ah cool
20:54	whartung	he's been fighting this for about 20 years now MegaMatt
20:55	MegaMatt	:)
20:57	whartung	Shiboleth was a great motivator in why we wrote our own IdP
20:57	MegaMatt	.. You wrote your own IdP???
20:57	whartung	yup
20:58	whartung	It's actually pretty cool, if I do say so myself
20:58	MegaMatt	Too many protocols to support ;)
20:58	whartung	meh, hardly.
20:59	whartung	HTTP and some SAML
20:59	sfisque	IdP … id provider?
20:59	whartung	Identity Provider -- term of art in the SAML spaec
21:00	whartung	space
21:00	MegaMatt	What about OAUTH and OpenID?
21:00	whartung	SAML is like French -- they have a different word for everything
21:00	sfisque	wasnt oauth defeated last year?
21:00	whartung	we don't support OAuth or OpenID
21:01	whartung	THere's a curious issue with OAuth folks don't necessarily think through
21:01	whartung	(dunno about OpenID, I haven;t looked at it)
21:01	whartung	but, as with any hash based system
21:01	whartung	you have to store password in plaintext for them to work (or have access to them in PT)
21:02	whartung	which is sucky
21:02	MegaMatt	mm
21:03	MegaMatt	Just looking at all the (weird) stuff you can do with OpenAM .. means that I'm in awe of anyone who'd want to write their own IdP ;)
21:04	whartung	but our IdP support SAML 2.0 Web Profile, for both SSO and SLO, support Two Factor Authentication, has user provisioning and replication to client apps (through a non-standard protocol as there really isn't one for this), user activity tracking, you can log out a user from the entire site at the console, see what apps they're using, user templates, role mapping..,
21:04	MegaMatt	nice
21:05	whartung	but, let me tell you, this stuff is all so fiddly
21:05	whartung	so many moving parts
21:05	whartung	we also have an adhoc 3rd party token facility (sort of like what OAuth does)
21:05	MegaMatt	CDSSO is a bit of black magic imo ;)
21:06	whartung	CDSSO?
21:06	MegaMatt	Cross Domain Single Sign-on
21:06	whartung	nah, it's simple
21:06	MegaMatt	with all that token passing and crap
21:07	whartung	ha, my answer is the first hit on Google /flex
21:07	whartung	http://stackoverflow.com/questions/342378/cross-domain-login-how-to-login-a-user-automatically-when-transferred-from-one
21:07	whartung	SAML Web 2.0 is more black magic, however
21:08	MegaMatt	heh
21:08	whartung	most of the implementations rely on a self submitting form via JS to do the work
21:10	whartung	But SAML Web Profile is pretty nice, you can do some neat stuff with it. We've done some cool use cases with it.
21:18	pdurbin	keep up the SAML chatter! biked home. off to pick up the kids :)
21:20	MegaMatt	Heh
21:20	MegaMatt	Well, I'll admit - I'm newb to all this stuff … trying to pick it up and learn it
21:21	MegaMatt	In your opinion, what's the big deal about REST .. any ways?
21:25		cem_ joined ##javaee
22:01		pdurbin_m joined ##javaee
22:02	pdurbin_m	MegaMatt: rest is better than soap
22:02	pdurbin_m	whartung: not quite 20 years. Nice answer
22:03		quesa joined ##javaee
22:03	quesa	hi
22:04	pdurbin_m	quesa: hey
22:12		kobain_ joined ##javaee

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