greptilian logo

IRC log for #javaee, 2013-10-25

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
00:01 pdurbin sfisque: can you run a separate application that talks Spring Remoting to the SSO and talks to your main application with RMI or whatever you want? A (hopefully smallish) service in between, I mean, a bridge of sorts.
00:03 pdurbin to avoid polluting the main app
00:03 whartung isn't Hesssian Springs "native" wire protocl?
00:03 sfisque yeah, i know there are "ways" to externalize spring.  i'm trying to do this in a way that does not complicate the deployment (we have enough issues having them use glassfish.  if "they" had their druthers, we'd be using spring inside of tomcat and there would be no glassfish).  i figure if they're talking RMI, we "should" be able to get a remote stub that answers to the interfaces they provide and it's all just vanilla RMI as far as we're concerned.  the o
00:03 sfisque i dunno whartung
00:03 sfisque hessian is caucho
00:04 sfisque whether spring uses it "also" natively, i do not know
00:04 whartung yea
00:04 whartung burlap is an old protocol
00:04 whartung so is hessian
00:05 sfisque oh your'e right.  brulap is caucho, hessian is spring
00:05 sfisque i just re-read that page closer
00:05 sfisque oh wait, i had it right, hessian is caucho.  so now i must google burlap
00:05 whartung yea
00:06 sfisque javaeebot lucky burlap java protocol remote
00:06 javaeebot sfisque: http://www.caucho.com/resin-3.0/protocols/burlap.xtp
00:06 whartung what, no google protocol buffers?!?!
00:06 sfisque that's just an encoding scheme, not a protocol
00:06 sfisque xml-lite
00:06 sfisque the local transit uses that in addition to json
00:06 whartung yea, this is all just RPC over Something
00:07 whartung so you're trying to get a compatible endpoint up?
00:07 whartung to receive messages?
00:07 sfisque i'm just hoping whichever proc they use, spring isnt doing "magic"
00:07 sfisque consume
00:07 whartung so to send messages?
00:08 sfisque central identity service exposes services via spring remoting.  i want to call and consume these services WITHOUT spring, if possible
00:08 sfisque from our ee-app in GF 3.1
00:08 whartung well in theory you should be able to do it with the Caucho hessian jar me thinks
00:08 sfisque IF they are using that proc
00:08 pdurbin sfisque: and it's all JSON?
00:08 sfisque depends on which protocol they have configured
00:08 whartung yea, well, that's the nut -- which are they publishing?
00:09 sfisque no, they have "re-exposed" some of the services to "external clients" via a json interface
00:09 sfisque dunno yet whartung
00:09 sfisque still awaiting "more info"
00:09 whartung then I wouldn't worry about it.
00:09 whartung they're NOT using RMI, cuz RMI is old school and hard to set up.
00:09 whartung they're not using JMS because its RPC
00:10 sfisque so i'm guessing, the "json interface" is probably just the spring web remoting
00:10 whartung They're not using JAX-WS because, you know, ewww, SOAP
00:10 sfisque hey, i like jax-ws
00:10 whartung I bet they're using Hessian CUZ IT'S FAST YO! Rod saysso
00:10 whartung and it's free
00:10 sfisque ROFL
00:10 whartung and it's default out of the box
00:10 sfisque if so, they yeah, we might be able to get away with just linking in the hessian libs and be done with it
00:11 sfisque i'm still putting my money that they are using spring web remoting
00:11 whartung when you view the problem through the Hip/Lazy filters, Hessian is the first choice
00:11 sfisque and all they probably did was document what it looks like "raw" for clients not using spring remoting natively
00:11 whartung oh, maybe, but they still need a wire protocol
00:12 sfisque it looks like hessian is http, the endpoint runs on 8080
00:12 sfisque so it's probably somethign like binary xml over http
00:12 whartung yea
00:13 whartung I mean, what else would it be lol
00:13 sfisque ip over carrier pigeon
00:13 sfisque :P
00:13 sfisque braille via ip over carrier pigeon
00:14 whartung here you go
00:14 whartung "Spring HTTP invokers use the standard Java serialization mechanism"
00:14 whartung yea, don't do that
00:14 whartung they're just pumping writeObject to HTTP streams
00:14 whartung that's just trouble, IMHO
00:15 pdurbin sfisque: when is the "more info" coming?
00:15 whartung "later", "soon"
00:16 sfisque hopefully by tomorrow afternoon.  likely monday or tuesday
00:16 sfisque you know… timely
00:16 pdurbin real soon now
00:16 whartung Real Soon Now(™)
00:17 sfisque whartung - yeah, unless they're base64'ing the payload, that could be REAL trouble
00:17 whartung "nothing to see here, move along"
00:17 whartung all right all, I must skedaddle
00:17 whartung ttyt
00:17 sfisque code strong!
00:17 sfisque me too actually.  time for me to pack up.  so much time and so little to do.....
00:17 * sfisque waves
00:17 pdurbin o/
01:12 sfisque joined ##javaee
03:23 sfisque1 joined ##javaee
03:58 scripty joined ##javaee
03:58 scripty Howdy
03:58 scripty All
04:42 dangertools joined ##javaee
04:42 dangertools joined ##javaee
06:00 scripty ??
06:59 cheater_2 joined ##javaee
08:19 AlexCzar joined ##javaee
09:33 zoot so...
10:08 * pdurbin reviews a few more talks: https://github.com/pdurbin/wiki/commits/master/java/javaone/2012.mdwn
11:23 MegaMatt joined ##javaee
12:29 tommmied joined ##javaee
13:20 Naros joined ##javaee
13:44 kobain joined ##javaee
13:45 AlexCzar joined ##javaee
14:08 SLovenberg joined ##javaee
15:29 SLovenberg joined ##javaee
15:32 cheturvedi joined ##javaee
15:53 acuzio joined ##javaee
15:54 SLovenberg joined ##javaee
16:00 AlexCzar joined ##javaee
16:22 acuzio joined ##javaee
16:31 pdurbin new! https://jersey.java.net/release-notes/2.4.html
16:45 semiosis jersey \m/
16:45 acuzio java.net is a terrible bloody domain
16:45 acuzio it hosts jersey which i love - but 60%+ of the time the site is down -
16:53 whartung I know, java.net is awful
17:39 pdurbin it would be cool if there were a http://blogs.perl.org for Java. a central place where anyone could have a Java blog if they want one
18:14 sfisque joined ##javaee
18:27 cem_ joined ##javaee
18:27 cem_ hi pdurbin
18:28 pdurbin cem_: hey
18:30 cem_ the last post sess and quest talked about caching but i still dont get it how to do caching , how to store the data in browser ?
18:32 cem_ can someone help me
18:33 whartung javabot lucky http caching
18:33 whartung bah
18:33 whartung javabot: lucky http caching
18:33 whartung bah
18:33 whartung I suckk
18:33 sfisque javaeebot lucky http caching
18:33 javaeebot sfisque: http://www.w3.org/Protocols/rfc2616/rfc2616-sec13.html
18:33 whartung ah!
18:33 whartung CLOSE!
18:34 sfisque yet so far...
18:34 sfisque :-)
18:34 sfisque g'day
18:34 whartung g'day sir
18:35 sfisque so i was successful in testing "bypassing" the client side camel linkages and sending a template/vars into AMQ and having camel process it and send the desired email out.  so i can begin removing camel client side from our build :-)
18:35 sfisque which should drop about 20m from our war
18:35 sfisque about 40% shrinkage
18:35 whartung *cough* omg
18:36 whartung "lightweight"
18:36 sfisque yeah.  i think that will speed up our build and deploy steps :-)
18:36 whartung a little
18:36 sfisque next step, hooking up to a remote spring remoted service without consuming spring on the client side.
18:36 whartung get an SSD and quitcherwhining
18:36 sfisque ROFL
18:37 sfisque ssd's are over-rated atm.  once i can by a 2.5 inch 2tb drive and not have it cost more than the notebook it's going in, i'm sold
18:37 whartung no no no, hardly omg… (well, it depends...)
18:37 cem_ what is the use of ipad ?
18:38 whartung if you mean just raw storage, yea, but SSDs are stupid fast….stupid stupid stupid fast for writing.
18:38 sfisque my big fear is the "half life" of flash.  i'd hate to replace platters with ssd and find that the ssd dies too fast
18:38 whartung IOPS is crazy
18:38 whartung I think that's all hooey today, frankly.
18:38 sfisque *nix OS's are hard on drives
18:38 whartung I have an  SSD in my mac here
18:39 whartung hasn't spot it out yet
18:39 whartung cem_: they're great for leveling tables
18:39 sfisque ROFL
18:39 cem_ LOL
18:39 sfisque they're great consumption devices.  my wife loves hers, but i could never handle that formfactor for authoring
18:40 sfisque unless point and click programming were to really experience a break through
18:40 whartung thats a granularity problem
18:40 sfisque what the tablet market needs is the 21st century equivalent of hypercard
18:40 whartung programing is too fine grained for "guis"
18:41 whartung save light weight scripting....
18:41 sfisque depends.  HC was phenominal for quickly building an app, but there's no current equivalent
18:41 whartung or rewriting General Motors back end SOA ...
18:45 pdurbin whartung: overheard: "adfs is probably the saml worlds nightmare"
18:45 whartung adfs?
18:45 whartung isn't that the Common Lisp package mangler?
18:45 pdurbin heh. no
18:45 sfisque javaeebot lucky adfs java
18:45 pdurbin javaeebot: lucky adfs microsoft
18:45 javaeebot sfisque: http://stackoverflow.com/a/7015244
18:45 javaeebot pdurbin: http://technet.microsoft.com/en-us/library/cc736690(v=ws.10).aspx
18:45 pdurbin Active Directory Federation Services (ADFS)
18:46 whartung oh, yea, well, that's AD
18:46 whartung everythings a nightmare
18:46 sfisque "hey, lets leverage kerberos, and add all sorts of non standard extensions…"  YAY!!!!
18:46 whartung YAY!
18:47 pdurbin let's start #sso on freenode
18:47 sfisque ####sso
18:47 pdurbin heh
18:47 sfisque ######################sso
18:48 sfisque ##javaPwnzWindowzCanS_ckIt
18:51 pdurbin let's burn all the computers
18:52 sfisque dewd, how can i plaze my g/\m3zzz?
18:52 sfisque :P
18:52 cem_ but how to update cache
19:06 * pdurbin looks at SAML for dummies | SURFnet Blog - https://blog.surfnet.nl/?p=1417
19:07 pdurbin "You may wonder where exactly SURFconext should be positioned in al this. In fact, SURFconext acts as a proxy between the IDP and the SP. Although this slightly complicates matters when relaying messages between IDPs and SPs, the same basic idea as sketched here applies."
19:08 pdurbin https://www.dataverse.nl/dvn/faces/login/SamlLoginPage.xhtml takes you to http://surfnet.nl
19:13 pdurbin whartung: does that mean surfnet.nl is a "discovery" service?
19:15 pdurbin javaeebot: lucky saml discovery service
19:15 javaeebot pdurbin: https://wiki.oasis-open.org/security/IdpDiscoSvcProtonProfile
19:34 syncsys_ joined ##javaee
19:49 kobain joined ##javaee
19:49 Guest41539 joined ##javaee
19:50 kobain_ joined ##javaee
19:50 Guest15203 joined ##javaee
20:09 kobain99 joined ##javaee
20:20 whartung no idea what that is pdurbin
20:21 tommmied joined ##javaee
20:21 pdurbin whartung: heh. no worries
20:21 pdurbin it's from some other country or something ;)
20:22 whartung looks like a normal profile workflow to me
20:26 whartung they don't really speak as to where they are in the workflow
20:26 whartung there are lots of places they can interject
20:28 pdurbin how dare they interject
20:30 pdurbin I see <file>${file.reference.oiosaml.​java-discovery-1.2.war}</file> in nbproject/project.xml
20:30 pdurbin so some discovery must be going on
20:30 whartung I've not looked at any discovery profiles for saml I duynno what they might be discovering
20:34 pdurbin whartung: well, I want to be able to support multiple IdPs per installation of our java ee app
20:34 whartung I mean, that's fine, you don;t  need an intermediary for that
20:34 whartung (assuming standard compliance, bla bla bla)
20:35 pdurbin hmm. ok. I was hoping not. that I don't need that surf thing or equivalent
20:37 whartung In the end, to log in, you just need to push the AuthResponse to your app. Where this comes from is pretty much irrelevant
20:38 pdurbin ok. makes sense
20:38 pdurbin right now OIOSAML is doing that for me
20:39 whartung right, you just need to decode the payload, authenticate it, react to the assertions, and move on. The payload source is secondary.
20:48 pdurbin right. I'll work on that next
20:48 pdurbin first I'm verifying that OIOSAML works with all the IdPs I care about
20:48 whartung see the issue is
20:48 whartung there's two issues
20:48 pdurbin as I wrote at [dvn-auth] Is your Shibboleth IdP (Identity Provider) ready for OIOSAML testing? - https://lists.iq.harvard.edu/pipermail/dvn-auth/2013-October/000002.html
20:48 whartung 1) is the payload coming from someone you trust.
20:49 whartung that's pretty straight forward (you can just try all of the certs you have, for example)
20:49 whartung the harder problem
20:49 whartung is that when someone hits your site the first time, is where do you redirect them to?
20:49 whartung that's a different issue completely
20:49 pdurbin like... "which institution are you from?"
20:49 whartung since you may have "no idea" who this person is in order to "choose" the correct IDP
20:49 whartung now, you you could prompt them with a list
20:49 whartung yea
20:49 whartung and then forward, and be done
20:50 pdurbin well, that's what that surf page is
20:50 sfisque assumedly, wouldnt u look them up in a directory (like ldap?)  and get their ou=xxxx node?
20:50 pdurbin people click their home institution
20:50 whartung right, see that's another way of doing it
20:50 pdurbin whartung: actually three are THREE issues
20:50 whartung the "IdP" can do it.
20:51 pdurbin 3. pdurbin has to pick up the kids
20:51 whartung if you send them to school/soccer/whatever with jerky, you don't have to rush off cuz they can live on their own for awhile...
20:51 pdurbin whartung: but let's please pick this up later :)
20:51 sfisque and make sure when they log in, the button is blue, otherwise, it's back to QA for you
20:52 pdurbin or devops. eep!
20:53 * pdurbin reassures the devops: http://irclog.perlgeek.de/crimsonfu/2013-10-25#i_7765455
21:02 syncsys_ joined ##javaee
23:28 sfisque 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.