| Time |
S |
Nick |
Message |
| 00:16 |
|
|
Naros joined ##javaee |
| 01:20 |
|
|
cem__ joined ##javaee |
| 02:53 |
|
|
kobain joined ##javaee |
| 03:03 |
|
|
jenue joined ##javaee |
| 03:38 |
|
pdurbin |
semiosis: so nothing is standard yet, as far as I can tell |
| 03:39 |
|
semiosis |
fwiw the jersey project (reference impl of jax-rs) is using MOXy for json |
| 03:39 |
|
pdurbin |
yep, exactly |
| 03:39 |
|
pdurbin |
a good starting point: The Java EE 7 Tutorial:JSON in Java EE RESTful Web Services | Java EE Documentation - http://docs.oracle.com/javaee/7/tutorial/doc/jsonp005.htm |
| 03:40 |
|
semiosis |
cool |
| 03:40 |
|
pdurbin |
"Jersey, the reference implementation for JAX-RS (JSR-311) included in GlassFish Server, provides support for binding JSON data from RESTful resource methods to Java objects using JAXB, as described in Using JAX-RS With JAXB in Chapter 31, "JAX-RS: Advanced Topics and Example". However, JSON support is not part of JAX-RS (JSR-311) or JAXB (JSR-222), so that procedure may not work for Java EE implementations |
| 03:40 |
|
pdurbin |
other than GlassFish Server. |
| 03:40 |
|
pdurbin |
The Java API for JSON Processing (JSR-353) does not explicitly support JSON binding in Java. A future JSR (JSON Binding) that is similar to JAXB for XML is under consideration for a future release of Java EE. |
| 03:40 |
|
pdurbin |
" |
| 03:40 |
|
semiosis |
oh interesting |
| 03:40 |
|
pdurbin |
which leads one to wonder about that future JSR... for JSON Binding |
| 03:41 |
|
pdurbin |
I'm not sure it exists yet |
| 03:41 |
|
pdurbin |
but here's a great blog post: JSON Binding with JAX-RS - Moxy, JSON-P, Jackson, Jettison (TOTD #214) (Arun Gupta, Miles to go ...) - https://blogs.oracle.com/arungupta/entry/json_binding_with_jax_rs |
| 03:41 |
|
pdurbin |
"MOXy is the default and preferred way of supporting JSON binding in your Jersey applications." |
| 03:42 |
|
pdurbin |
and here's some code: https://github.com/javaee-samples/javaee7-samples/tree/master/jaxrs/moxy |
| 03:42 |
|
semiosis |
pdurbin: you were at java-one right? looks like you missed it - https://oracleus.activeevents.com/2013/connect/sessionDetail.ww?SESSION_ID=3838 |
| 03:42 |
|
pdurbin |
and more on Moxy here: https://jersey.java.net/documentation/latest/media.html |
| 03:43 |
|
semiosis |
the slide deck from that java-one session: https://oracleus.activeevents.com/2013/connect/fileDownload/session/514F9D38E8ACECDE2A284EC311DDDD29/CON3838_Doughan.pdf |
| 03:43 |
|
pdurbin |
semiosis: d'oh! "Introducing the Java API for JSON-Binding (JSON-B) JSR [CON3838]" |
| 03:44 |
|
pdurbin |
"JSON-binding is converting Java objects to/from JSON" |
| 03:44 |
|
pdurbin |
yep. that's what I want. But what's the JSR number? |
| 03:45 |
|
pdurbin |
"JSR proposal is currently being prepared" |
| 03:45 |
|
semiosis |
whats in a number? ;) |
| 03:46 |
|
semiosis |
wow they're really calling it json-p? dont they know that name is taken? |
| 03:46 |
|
semiosis |
@lucky jsonp |
| 03:46 |
|
semiosis |
eh |
| 03:47 |
|
semiosis |
javaeebot: lucky json-p |
| 03:47 |
|
javaeebot |
semiosis: http://en.wikipedia.org/wiki/JSONP |
| 03:47 |
|
semiosis |
right |
| 03:47 |
|
pdurbin |
yeah, there's already a jsonp |
| 03:47 |
|
semiosis |
two bad naming collisions in here today... json-p & amqp |
| 03:48 |
|
semiosis |
naming things is hard |
| 03:48 |
|
pdurbin |
here's that guy who gave the talk and posted those slides: https://twitter.com/bdoughan |
| 03:49 |
|
semiosis |
neat |
| 03:50 |
|
semiosis |
my advice is to not worry about a jsr. people got along fine for years & years without a jsr for dependency injection |
| 03:50 |
|
semiosis |
there's several great json binding libraries available today, pick one & dont look back |
| 03:50 |
|
pdurbin |
yeah, I'll probably give moxy a try |
| 03:51 |
|
semiosis |
i've been working with moxy this week and it's pretty nice |
| 03:51 |
|
pdurbin |
cool |
| 03:51 |
|
semiosis |
jackson is nice too |
| 03:52 |
|
semiosis |
havent used any of the others |
| 03:52 |
|
semiosis |
but they're probably also nice lol |
| 03:52 |
|
* semiosis |
wonders why there are so many |
| 03:52 |
|
pdurbin |
heh |
| 03:52 |
|
pdurbin |
this is why they standardize this stuff... eventually |
| 03:53 |
|
semiosis |
there's still three popular implementations of jpa iirc, and idk how many unpopular ones |
| 03:53 |
|
semiosis |
meh |
| 03:54 |
|
pdurbin |
MOXy is the New Default JSON-Binding Provider in GlassFish 4 (The Aquarium) - https://blogs.oracle.com/theaquarium/entry/moxy_is_the_new_default |
| 03:54 |
|
semiosis |
then go with that |
| 03:54 |
|
pdurbin |
"It is in fact correct that Java EE 7 has standardized JSON processing (via JSON-P) but not a JSON binding API yet (although such a standardized API is a very strong consideration for Java EE 8 and would build upon the JSON-P work)." |
| 03:55 |
|
pdurbin |
ah, and code at http://blog.bdoughan.com/2013/06/moxy-is-new-default-json-binding.html |
| 03:58 |
|
semiosis |
http://wiki.eclipse.org/EclipseLink/Examples/MOXy#MOXy.27s_JSON-Binding |
| 03:58 |
|
semiosis |
i found the twitter example there very helpful |
| 04:02 |
|
* pdurbin |
looks at http://wiki.eclipse.org/EclipseLink/Examples/MOXy/JSON_Twitter |
| 06:35 |
|
* sfisque |
is a sad panda |
| 06:36 |
|
sfisque |
GF 3.1.2.x does not support JNDI Events on its Custom Resource directory. :-( |
| 06:36 |
|
sfisque |
anyone know if this is fixed in the 4.x stream? |
| 06:51 |
|
|
dangertools joined ##javaee |
| 08:07 |
|
|
pleroma joined ##javaee |
| 08:43 |
|
|
weyer joined ##javaee |
| 09:38 |
|
|
AlexCzar joined ##javaee |
| 09:43 |
|
|
pleroma joined ##javaee |
| 10:21 |
|
neuro|sys |
morning |
| 10:37 |
|
acuzio |
morning ladies |
| 10:43 |
|
|
dangertools joined ##javaee |
| 11:26 |
|
|
cem__ joined ##javaee |
| 12:06 |
|
|
tangatools joined ##javaee |
| 12:06 |
|
|
tangatools joined ##javaee |
| 12:27 |
|
cem__ |
why we need to have servletconfig ? |
| 12:50 |
|
|
drspockbr joined ##javaee |
| 12:52 |
|
pdurbin |
javaeebot: lucky servletconfig |
| 12:52 |
|
javaeebot |
pdurbin: http://docs.oracle.com/javaee/5/api/javax/servlet/ServletConfig.html |
| 12:53 |
|
pdurbin |
cem__: here's a deceny example: http://www.mkyong.com/servlet/how-to-pass-parameters-to-a-servlet-servletconfig/ |
| 12:54 |
|
cem__ |
but there is no need of servletconfig right |
| 12:54 |
|
cem__ |
we can store in enum |
| 12:55 |
|
cem__ |
constant |
| 12:55 |
|
pdurbin |
javaeebot: lucky servletconfig vs. enum |
| 12:55 |
|
javaeebot |
pdurbin: http://tomcat.apache.org/tomcat-5.5-doc/servletapi/javax/servlet/ServletConfig.html |
| 13:00 |
|
pdurbin |
cem__: keep talking. I don't understand what you mean |
| 13:02 |
|
cem__ |
what do you mean |
| 13:02 |
|
cem__ |
you dont understand |
| 13:02 |
|
pdurbin |
right |
| 13:03 |
|
cem__ |
right ? |
| 13:04 |
|
pdurbin |
cem__: keep talking. I don't understand what you mean |
| 13:05 |
|
cem__ |
neither i dont understand what do you mean |
| 13:05 |
|
cem__ |
do* |
| 13:06 |
|
pdurbin |
I'm not the one with the question :) |
| 13:06 |
|
pdurbin |
(for once) :) |
| 13:06 |
|
cem__ |
you never asked the question |
| 13:11 |
|
|
cem__ left ##javaee |
| 13:17 |
|
|
jieryn joined ##javaee |
| 13:17 |
|
|
jieryn joined ##javaee |
| 14:01 |
|
|
AlexCzar joined ##javaee |
| 14:10 |
|
|
acuzio joined ##javaee |
| 14:10 |
|
|
acuzio joined ##javaee |
| 14:19 |
|
|
sheenobu joined ##javaee |
| 14:29 |
|
AlexCzar |
hillarious conversation… |
| 14:31 |
|
pdurbin |
:) |
| 14:32 |
|
|
Naros joined ##javaee |
| 14:32 |
|
sheenobu |
Sounds like I barely missed something good |
| 14:32 |
|
|
sheenobu joined ##javaee |
| 15:53 |
|
|
helga joined ##javaee |
| 15:58 |
|
helga |
hi guys! could anybody explain, how to work with java Iterable? i need to implement function, that takes Iterable<Integer> as a parameter. i need to iterate through this sequence and mutate each element of it in a particular way. |
| 16:00 |
|
sess |
~ Iterable |
| 16:00 |
|
sess |
~Iterable |
| 16:00 |
|
sess |
bleh |
| 16:00 |
|
sess |
helga: http://docs.oracle.com/javase/7/docs/api/java/util/Iterator.html |
| 16:01 |
|
sess |
It should be fairly obvious from the API |
| 16:02 |
|
Fubar^ |
or use guava's Iterables.transform |
| 16:02 |
|
Fubar^ |
And there is no reason to look at the Iterator javadoc. Since you are doing it wrong if you use it that way |
| 16:03 |
|
sess |
huh? |
| 16:03 |
|
Fubar^ |
You can iterate over an iterable with enhanced-for |
| 16:03 |
|
Fubar^ |
So you don't need to know how an iterator works |
| 16:03 |
|
pdurbin |
helga: here's an example I happen to be staring at: http://wiki.apache.org/solr/Solrj#Highlighting |
| 16:04 |
|
pdurbin |
while (iter.hasNext()) {... |
| 16:04 |
|
Fubar^ |
for (Integer i : theIterable) { ... |
| 16:04 |
|
|
Losowski joined ##javaee |
| 16:04 |
|
sess |
I think "wrong" is a bit harsh, but sure a for loop is probably enough |
| 16:07 |
|
helga |
thank you, this: for (Integer i : theIterable) works!:) |
| 16:16 |
|
|
Losowski left ##javaee |
| 16:28 |
|
|
Losowski joined ##javaee |
| 16:34 |
|
pdurbin |
solr or elasticsearch? any preference? |
| 16:39 |
|
semiosis |
elasticsearch (the company) hired three people i know from the logstash community and one i know from redhat/gluster, so i'd go with them |
| 16:40 |
|
pdurbin |
interesting reason :) |
| 16:42 |
|
|
Losowski left ##javaee |
| 16:44 |
|
pdurbin |
bah! `@EJB MyBean myBean` was working a minute ago but now myBean is null. Why? |
| 16:45 |
|
sess |
is the class managed? |
| 16:47 |
|
pdurbin |
it's @Stateless and @Named |
| 16:47 |
|
Naros |
pdurbin: I use solr combined with Hibernate search |
| 16:47 |
|
Naros |
it works extremely well |
| 16:48 |
|
pdurbin |
Naros: not SolrJ? |
| 16:48 |
|
Naros |
the only caveat I am still dealing with is how to manage rebuilding the index due to non-ORM changes to the datastore. |
| 16:48 |
|
Naros |
I believe HSEARCH uses Lucene/SOLR under the hood |
| 16:49 |
|
Naros |
i could be wrong but i think you can have it use other impl too, not just lucene |
| 16:51 |
|
pdurbin |
we're just using ordinary JPA |
| 16:53 |
|
Naros |
Ah |
| 16:53 |
|
pdurbin |
sess: I restarted glassfish and now @EJB works again. #wtf |
| 16:55 |
|
sess |
classic random error |
| 16:55 |
|
sess |
containers fail everyonce in a while |
| 16:55 |
|
sess |
for whatever reason |
| 16:58 |
|
pdurbin |
stupid container |
| 17:04 |
|
pdurbin |
(glassfish 4) |
| 17:06 |
|
pdurbin |
I know I can put a comment above a <dependency> in maven but is there an element I can use within the <dependency> block? something like <whyImIncludingThisDependency>? |
| 17:07 |
|
acuzio |
sess: Did you get to the bottom of the Transaction stuff ? |
| 17:08 |
|
sess |
acuzio: I removed the exception translator instead, and left transaction settings as they were |
| 17:08 |
|
sess |
haven't got time to dig deep into it and don't want to change any code |
| 17:12 |
|
semiosis |
pdurbin: it's just xml, you can put <!-- --> pretty much anywhere |
| 17:13 |
|
semiosis |
except within another tag. this is not allowed: <depen<!-- foo -->dency> |
| 17:14 |
|
pdurbin |
semiosis: yeah, just put the comment within, I guess |
| 17:14 |
|
pdurbin |
I just didn't know if there was a standard way to do this |
| 17:14 |
|
pdurbin |
even <description> or something would be fine |
| 17:56 |
|
|
kinabalu joined ##javaee |
| 18:16 |
|
|
sfisque joined ##javaee |
| 18:25 |
|
sfisque |
anyone here ever wired in a custom JNDI Factory into glassfish directly? it seems Custom Resources do not implement the Event API that jndi supports. |
| 18:56 |
|
pdurbin |
no but this new thread on the glassfish dev list is getting interesting: About adding modules/features for GlassFish dynamiclly (sic) https://java.net/projects/glassfish/lists/dev/archive/2013-12/message/0 |
| 19:09 |
|
whartung |
I thought this was already done with all of the OSGi work |
| 19:13 |
|
sfisque |
i'm guessing it involves "non EE" features or features "outside" the osgi spec |
| 19:48 |
|
pdurbin |
not sure |
| 19:56 |
|
sheenobu |
someone mentioned OSGI. yay? |
| 20:00 |
|
pdurbin |
yay |
| 20:01 |
|
whartung |
more like "vey" as in "oi vey" |
| 20:01 |
|
sfisque |
lolz |
| 20:04 |
|
sfisque |
so i realized that i don't need to do dynamic uri's for dispatching data. i must have been having a brain fart, because it occurred to me that i could just wrap the link around a f:setPropertyActionListener |
| 20:05 |
|
sfisque |
prevents me from supporting deep linking for those linkages, but still doable in the near term for now |
| 20:06 |
|
sfisque |
having the REST->redirect->facelets would allow me to deep link for data dynamic views in a jsf context |
| 20:49 |
|
|
knoppix joined ##javaee |
| 21:04 |
|
sheenobu |
oh! I forgot to continue that line of thought |
| 21:04 |
|
sheenobu |
I want to like osgi |
| 21:04 |
|
sheenobu |
but I haven't used it enough |
| 21:04 |
|
sheenobu |
and I don't know enough people who use it |
| 21:08 |
|
sfisque |
if you're doing EE, you "use it". bascially unless you're writing a container level facility, you don't actually write osgi code |
| 21:13 |
|
sheenobu |
right, I understand that. |
| 21:14 |
|
sheenobu |
If I'm going to 'use' osgi, it would be with an osgi container directly, not specifically with an EE container that is osgi-based. |
| 21:14 |
|
sfisque |
OR |
| 21:15 |
|
sheenobu |
not that I prefer one over the other (this /is/ ##javaee) but that i would consider that to be 'using'. |
| 21:15 |
|
sfisque |
you'd write an implementation for a facility that the container exposes regardless of whether it's an EE container or an osgi container |
| 21:15 |
|
sheenobu |
right |
| 21:15 |
|
sfisque |
like publishing a directory server that exposes a JNDI interface, and bootstraps in a way to be osgi compliant so it can plug in or be hot swapped according to the spec |
| 21:16 |
|
sheenobu |
right |
| 21:17 |
|
sheenobu |
project i'm on right now pre-dates spring, modern javaee and has NO modularity. I guess I just want that, regardless of the tech used |
| 21:17 |
|
sfisque |
"modern javaee"… as in pre ejb3? |
| 21:18 |
|
sheenobu |
I assume as much. IIRC spring predates ejb3 by quite a bit |
| 21:18 |
|
sfisque |
only a couple of years actually |
| 21:18 |
|
sheenobu |
i see |
| 21:18 |
|
sfisque |
ejb3 rolled off the assembly line about 5ish years ago |
| 21:19 |
|
sfisque |
and ejb1 spec came out about 14 years ago |
| 21:19 |
|
sheenobu |
woo |
| 21:20 |
|
sfisque |
javaeebot lucky spring framework java timeline |
| 21:20 |
|
sheenobu |
i'm a bit young to remember that >_> |
| 21:20 |
|
javaeebot |
sfisque: http://en.wikipedia.org/wiki/Spring_Framework |
| 21:20 |
|
sfisque |
oh noze… is javaeebot sick? |
| 21:20 |
|
sfisque |
oh there we go |
| 21:20 |
|
sfisque |
spring 1.0 march 2004 |
| 21:20 |
|
sheenobu |
i see that |
| 21:21 |
|
sfisque |
so v1.0 predates ejb3 by about 4 years, but what everyone "considers" spring dates back to 2006 |
| 21:21 |
|
sfisque |
so about 2 years prior to ejb3 |
| 21:21 |
|
sheenobu |
i see |
| 21:26 |
|
sfisque |
i'm starting to look at prettyfaces. i forget who linked it to the channel, but thank you if you're listening |
| 21:26 |
|
|
weyer joined ##javaee |
| 21:26 |
|
sfisque |
it looks like a perfect solution to what i want |
| 21:26 |
|
sheenobu |
prettyfaces is nice. |
| 21:27 |
|
sfisque |
yeah. i miss having controller dispatch via url (like struts/tiles) with jsf. pretty faces looks like it will give that back to me |
| 21:28 |
|
sfisque |
i'm kind of intrigued why jsf punted on controller dispatch via url, given that some of the struts/tiles dudes were on the JSF committee in the early stages |
| 21:29 |
|
sfisque |
you'd think they'd wire in support for it parallel to the component handling that jsf does |
| 21:29 |
|
semiosis |
my guess would be in those days configuration was preferred over convention |
| 21:30 |
|
sfisque |
i guess, but controller dispatch allows you to deep link on dynamic pages, which is a weak spot for jsf |
| 21:31 |
|
sfisque |
maybe jsf-3 will finally give us best of both worlds :-) |
| 21:33 |
|
sheenobu |
we use spring mvc and i don't think we use controller dispatch via URL. not 100% how that happened |
| 21:33 |
|
sfisque |
you have no uri/{some_dynamic_key} url binding? |
| 21:34 |
|
sfisque |
surprising if so, that's one of the few winning aspects of spring IMO |
| 21:36 |
|
sheenobu |
yeah we have a custom handlermapper/controller resolver |
| 21:42 |
|
sess |
sfisque: cool, i should try it out aswell |
| 21:44 |
|
pdurbin |
the spec leads are very happy to receive feedback |
| 21:46 |
|
sfisque |
sheenobu nifty |
| 21:46 |
|
sfisque |
pdurbin spec leads … on JSF? |
| 21:47 |
|
pdurbin |
sfisque: yeah. Ed Burns and company |
| 21:47 |
|
* sfisque |
picks up the phone... |
| 21:47 |
|
pdurbin |
all the Java EE spec leads really |
| 21:48 |
|
|
Losowski joined ##javaee |
| 21:48 |
|
pdurbin |
that's what they said at this talk anyway: Meet the Java EE Specification Leads [BOF2795] https://oracleus.activeevents.com/2013/connect/sessionDetail.ww?SESSION_ID=2795 |
| 21:48 |
|
pdurbin |
:) |
| 21:48 |
|
sfisque |
i definitely have a bone to pick with the GF people who did the JNDI repo. i was going to wire up our jndi configuration fields with event handlers, and then, got all sorts of ClassCastExceptions because they didnt implement EventContext. blech |
| 21:49 |
|
sfisque |
i wanted to remove the need for a JMX endpoint for runtime configuration, by having JNDI callbacks when values changed on the fly. boy was i miffed by that roadblock |
| 21:49 |
|
sess |
I only wish the "official" web framework wasn't such a weird one |
| 21:50 |
|
sess |
as in JSF |
| 21:50 |
|
sfisque |
aye. at first jsf is very wierd. it took me sometime to "unlearn" controller dispatch from struts/tiles and embrace component dispatch. talk about a paradigm shift |
| 21:51 |
|
sfisque |
it finally hit me… had the "oh…. now i get it" moment, and i felt better after that |
| 21:51 |
|
semiosis |
and this is why i put up a rest api barrier between core java & frontend code |
| 21:51 |
|
semiosis |
jax-rs FTW! |
| 21:51 |
|
sess |
sfisque: it works, but it's quirky and lacks many features that feels obvious that they are needed |
| 21:52 |
|
sheenobu |
yeah i hated life in the JSF 1.1 days, afraid to go back |
| 21:52 |
|
sfisque |
what mech do you use to maintain "state/session" with that |
| 21:52 |
|
sess |
such as running some code on load |
| 21:52 |
|
sess |
doing it is possibly, but feels very hacky and unsafe |
| 21:52 |
|
sess |
the JSF templates are pretty good though |
| 21:52 |
|
sheenobu |
our sessions were giant though, I mean like hibernate entities pushed into the session, relationships and all. |
| 21:52 |
|
sess |
wish I could use JSF templates with spring mvc |
| 21:53 |
|
sess |
since JSP is a giant pain in the butt |
| 21:53 |
|
sfisque |
oh my…. [in george takei voice] |
| 21:53 |
|
sess |
sheenobu: I just abuse viewScope these days |
| 21:53 |
|
sess |
put everything in it |
| 21:53 |
|
sfisque |
why not use struts2/tiles2 then? if you want good templating with url dispatch? |
| 21:53 |
|
sess |
and work against it with ajax |
| 21:53 |
|
sess |
sfisque: struts2 feels unstandard, and doesnt it use jsp? |
| 21:53 |
|
sheenobu |
imo i don't mind jsp(x), providing we use decent taglibs, dont' embed code, etc |
| 21:54 |
|
sfisque |
struts uses whatever you wire in, jsp, velocity, tiles, whatever |
| 21:54 |
|
sess |
just the fact that it lacks ui:composition drives me mad |
| 21:54 |
|
sfisque |
and i LOVE tiles… always have going back to 1.2 |
| 21:54 |
|
sess |
I hate tiles |
| 21:54 |
|
sfisque |
yes, no decoration/compositing |
| 21:54 |
|
sess |
having to look in a configure file every single time im looking for something simple? No thanks |
| 21:55 |
|
sess |
I guess it could be alright with simple templating, just header/footer/sidebar |
| 21:55 |
|
sfisque |
looser decoupling. controller doesnt have to know where the file is, just uses some agreed on uri |
| 21:55 |
|
|
Losowski left ##javaee |
| 21:56 |
|
sfisque |
tiles works great for that. you build a boilerplate layout and then punch in the tiles in the various areas |
| 21:56 |
|
sfisque |
give the layout subclasses meaningful uri's and the controller is then decoupled |
| 21:57 |
|
sfisque |
but yes, tiles does require more xml than i'd like |
| 21:57 |
|
sfisque |
though with v2 it's "better" than v1 was |
| 21:58 |
|
sfisque |
i will admit though i do prefer jsf ajax over S2/T2 ajax |
| 21:58 |
|
sess |
I use struts 1 with tiles in a project, that's very confusing |
| 21:58 |
|
sess |
dont remember if it was tiles 1 or 2 |
| 21:58 |
|
sfisque |
probably 1 on both. i do not know if S1 has a T2 plugin |
| 21:58 |
|
sfisque |
but yeah S1/T1 is "good" but does get heavy handed with the xml configuration |
| 21:58 |
|
sess |
If I have an url, i have to search around for like 1-2 minutes to find the correct JSPs and controllers |
| 21:59 |
|
Naros |
o.O |
| 21:59 |
|
sess |
reused content JSPs combined with controller inheritance |
| 21:59 |
|
sess |
ugh |
| 22:00 |
|
sfisque |
i try to avoid controller inheritence. i prefer controllers be idempotent pojos, if possible unless it's a gross transgression against DRY |
| 22:00 |
|
Naros |
Yah we have quite a number of controllers that subclass more abstract types but the inheritance tree is generally no more than 3 deep |
| 22:00 |
|
sess |
idempotent? |
| 22:00 |
|
sess |
is that even a word |
| 22:00 |
|
sfisque |
javaeebot definition idempotent dictionary |
| 22:00 |
|
javaeebot |
sfisque: Error: "definition" is not a valid command. |
| 22:00 |
|
sfisque |
javaeebot lucky idempotent dictionary entry |
| 22:01 |
|
javaeebot |
sfisque: http://www.merriam-webster.com/dictionary/idempotent |
| 22:01 |
|
sfisque |
in this context, self contained, no sub/super classes |
| 22:01 |
|
sfisque |
one of them thar mathy words |
| 22:01 |
|
sess |
that word made no sense whatsoever in this context lol |
| 22:01 |
|
* sfisque |
is a mathematician |
| 22:01 |
|
sess |
but sure i get your point |
| 22:02 |
|
sess |
keeping it DRY is very hard without inheritence |
| 22:02 |
|
sess |
especially when fetching data is often only possible from the controller itself |
| 22:02 |
|
sess |
rather than utility classes |
| 22:02 |
|
sfisque |
there are other ways than inheritence. many dev's under utilize composition over realization |
| 22:03 |
|
sfisque |
fragment the utility and inject/instantiate it |
| 22:03 |
|
sfisque |
like legos |
| 22:03 |
|
sess |
the composed instances are usually less powerful than the controllers though |
| 22:03 |
|
sfisque |
inheritence is powerful but there are mis-uses of it i feel |
| 22:03 |
|
sess |
with DI frameworks |
| 22:04 |
|
sess |
yeah inheritence is very misused, I do it myself :x |
| 22:04 |
|
Naros |
if you're talking struts, you can often avoid inheritance in favor of ModelDriven<> |
| 22:04 |
|
sfisque |
right, but merely inheriting an object just to get a method is not "proper" in a (self admittedly way) pedantic sense |
| 22:04 |
|
whartung |
I was inherited... |
| 22:04 |
|
sfisque |
lolz |
| 22:04 |
|
sess |
sfisque: exactly |
| 22:05 |
|
sess |
I made some stuff for JSF that im not quite sure im alright with |
| 22:05 |
|
sess |
such as AbstractEntityController<T> that stores an ID and an instance of an entity |
| 22:05 |
|
sess |
for those pages where you edit some object |
| 22:05 |
|
sfisque |
we had someone here previously who ran rampant with inheritence. lets just say, he managed to reinvent spaghetti code without "goto" which is quite an accomplishment |
| 22:05 |
|
sess |
often it introduces more complexity than help |
| 22:06 |
|
whartung |
yea, to a point |
| 22:06 |
|
sfisque |
oh templated controllers… |
| 22:06 |
|
* sfisque |
runs away screaming |
| 22:06 |
|
Naros |
lol, I believe I has one of those too.... DocumentAttachmentAction<T> |
| 22:07 |
|
sfisque |
we have…. |
| 22:07 |
|
* sfisque |
starts counting |
| 22:07 |
|
sfisque |
…. more than this…. |
| 22:07 |
|
* sfisque |
holds up both hands and feet |
| 22:07 |
|
sfisque |
…. too many |
| 22:07 |
|
Naros |
haha |
| 22:07 |
|
sfisque |
an easter egg from a previous member of the team |
| 22:07 |
|
* sfisque |
begins to cry |
| 22:08 |
|
Naros |
We have things like WorkOrderAction and then a WorkOrderCreateAction and a WorkOrderSaveAction and a WorkOrderApproveAction, etc. |
| 22:08 |
|
Naros |
where the create, save, and approve subclass the work order action |
| 22:08 |
|
sfisque |
and they are all subclasses of WorkOrderAction<T>? |
| 22:08 |
|
sfisque |
nailed it |
| 22:08 |
|
sfisque |
!!! |
| 22:08 |
|
Naros |
Nah, it isn't templated. |
| 22:09 |
|
sfisque |
may the baby java bless you |
| 22:09 |
|
sfisque |
:-D |
| 22:09 |
|
Naros |
The parent class basically holds all common parameters, methods, and refactored api |
| 22:09 |
|
sfisque |
aye and they are semantically related, which makes sense for inheritence |
| 22:09 |
|
* sfisque |
does a sincere golf clap |
| 22:09 |
|
Naros |
hah |
| 22:09 |
|
sfisque |
:-D |
| 22:10 |
|
Naros |
One could argue tho in struts, just create a single action and leverage different methods mapped to urls, but i dislike that approach. |
| 22:10 |
|
sfisque |
for a small site that would be fine. |
| 22:10 |
|
sfisque |
if you're serving up a simple workflow with a couple renderings, why not |
| 22:10 |
|
Naros |
prefer my action class to be specific to a certain control flow |
| 22:10 |
|
sess |
single action is pretty awrful |
| 22:10 |
|
sess |
awful |
| 22:10 |
|
sfisque |
but not for something enterprise driven |
| 22:11 |
|
sess |
"works for small sites" is just another word for bad |
| 22:11 |
|
sess |
anything works for small sites |
| 22:11 |
|
sfisque |
not really. if you're just building a toy site for some friends, why over architect |
| 22:11 |
|
sess |
It's not even more work really |
| 22:12 |
|
sess |
I assume you meant using like querystring to point out an action |
| 22:12 |
|
sfisque |
that's one mech that for a small site (like 5 users?) is a good fit, as long as you still sanitize the input |
| 22:13 |
|
sfisque |
security concerns dont go away with scale, once it's public cloud facing |
| 22:13 |
|
sess |
just seperating the actions to different methods would improve design greatly with almost no effort |
| 22:13 |
|
sess |
:d |
| 22:13 |
|
sess |
aaanyways im off to bad :> |
| 22:13 |
|
sess |
bed even* |
| 22:14 |
|
Naros |
lol |
| 22:14 |
|
* sfisque |
waves to sess |
| 22:14 |
|
sfisque |
with modern IDEs sure. |
| 22:14 |
|
sfisque |
right click —> refactor —> move up |
| 22:15 |
|
whartung |
right click -> rewrite -> move out |
| 22:15 |
|
sfisque |
eclipse? |
| 22:15 |
|
whartung |
life |
| 22:16 |
|
sfisque |
oh, lolz |
| 22:23 |
|
|
kobain joined ##javaee |
| 23:20 |
|
|
pdurbin_m joined ##javaee |
| 23:21 |
|
pdurbin_m |
where did sess go? Ed Burns replied right away: http://www.evanchooly.com/logs/%2523%2523jsf/2013-12-05 |
| 23:23 |
|
sfisque |
oh my icefaces |
| 23:29 |
|
sfisque |
i see you dropped a hint about using rest to serve up facelets |
| 23:30 |
|
sfisque |
or was that ed burns offering that as a hint? |
| 23:30 |
|
sfisque |
How about if you could use JSF templates with JAX-RS MVC in Jersey? |
