<< 29-August-2008 : bevinbot on #glassfish at freenode [download] [back] >>
 
 
time nick message

13:12

<hani>

are they any jersey people here?

13:12

<hani>

the jsr311 thing, not the crap state

13:14

<jdlee>

hahaha

13:14

<jdlee>

fwiw, no one has spoken yet today :)

13:14

<hani>

I'll ask in 3 hours, when the sun people wake up!

13:15

<jdlee>

jdlee is a sun person and awake

13:15

<jdlee>

:)

13:16

<jdlee>

though I don't do anything with jersey

13:16

<hani>

what do you do then?

13:16

<jdlee>

glassfish admin console

13:16

<hani>

oh nice

13:16

<jdlee>

i like it ;)

13:16

<hani>

does it let me upload jdbc drivers yet?

13:16

<hani>

I filed an issue like a million months ago

13:16

<jdlee>

not yet, but it seems like I saw some activity on the tracker regarding that

13:16

<hani>

oh nice

13:17

<jdlee>

yeah, that would be

13:17

<jdlee>

jdlee looks at the tracker

13:19

<jdlee>

issue 3355 :)

13:19

<jdlee>

i'm going to vote on that one :)

13:20

<hani>

hehe

13:20

<jdlee>

the actual upload would be easy. i'm not sure how much work it would take to make the app server aware of the new jars w/o a restart, though

13:20

<jdlee>

should be possible, but that's not my area :)

13:20

<hani>

its fine if it pops up a msg saying you need to restart now

13:20

<jdlee>

jdlee nods

13:20

<hani>

it'd be a good first step

13:20

<jdlee>

that would be the easy way...

13:20

<jdlee>

certainly

13:20

<hani>

basically the console should never force you to drop to command line or faffing about with the fs

13:20

<jdlee>

jdlee nods

13:21

<jdlee>

i really wish we could restart the entire app server from the console (I even filed an RFE for that forever ago), but I'm getting rumblings that that won't be added any time soon if at all

13:21

<hani>

yeah, not surprised

13:22

<hani>

how would you do it? Apps generally arent good at rebooting themselves

13:22

<jdlee>

i'm not sure. there was talk about keeping the startup thread around and using that, but i think kedar or jerome nixed that

13:22

<jdlee>

fwiw, oc4j does allow that. one of the few things they got right :P

13:23

<jdlee>

this issue might be of interest to you: https://glassfish.dev.java.net/issues/show_bug.cgi?id=3897

13:23

<jdlee>

that poses versioning/upgrade issues, but is an interesting take on the issue

13:23

<hani>

ohyeah,t his was very annoying too: https://glassfish.dev.java.net/issues/show_bug.cgi?id=3354

13:23

<hani>

such an easy thing to fix!

13:24

<jdlee>

yeah...

13:25

<jdlee>

i'll either add myself to the cc list or take ownership and see if I can get that don after we get v3 prelude out

13:29

<jdlee>

anissa: mornin'

13:29

<anissa>

good morning

13:29

<jdlee>

anissa: i think issue 4315 is a dupe of 4314

13:29

<jdlee>

looks like 4315 was an accident :)

13:30

<anissa>

let me see

13:30

<anissa>

exactly, you are right

13:31

<anissa>

i will mark that dup.

13:31

<jdlee>

ok

13:31

<jdlee>

i was just looking at some issues hani brought up and saw those

13:31

<anissa>

These days i only query on DEFECT

13:32

<jdlee>

jdlee nods

13:32

<jdlee>

for HCF, that's probably best :)

13:32

<hani>

so do the jersey people hang out here?

13:32

<hani>

or do they have their own channel, or are they all slackers?

13:33

<jdlee>

i don't have an answer for that :)

13:33

<jdlee>

i don't recall seeing any jersey people here or in #glassfish-webtier (a new channel rajiv recently created)

13:34

<hani>

so they're slackers

13:34

<jdlee>

you might need to email them and ask them to join

13:34

<jdlee>

could be :)

13:34

<hani>

you can always tell the slackers by how long issues are ignored for

13:35

<jdlee>

hehe

13:36

<jdlee>

time for breakfast...biab :)

14:17

<jdlee>

jdlee is back

15:20

<[pwgr]>

hani!

15:22

<hani>

ummm.hi!

15:22

<[pwgr]>

[pwgr] is _tenchi_

15:22

<hani>

ohyeah!

15:23

<hani>

just recognised the name

15:23

<[pwgr]>

wouldnt guess i'd ever see you on freenode

15:23

<hani>

yeah I hate freenode

15:23

<hani>

but I need to yell at some people about jersey

15:23

<hani>

and as is often the case, the people that need to be yelled at are hiding

15:25

<[pwgr]>

they have their hands full with maven and all of its antics

15:25

<[pwgr]>

keeps one quite busy

15:25

<hani>

haha, yes, they're so silly

15:26

<hani>

half the time svn doesnt build

15:26

<hani>

with internal maven errors

15:26

<hani>

and you end up with useful artifacts like this: http://download.java.net/maven/2/com/sun/jersey/jersey-bundle/0.11-ea-SNAPSHOT/

15:26

<[pwgr]>

shameful

15:26

<hani>

(note the size of the sources jar)

15:27

<[pwgr]>

that must have been run through maven's fancy ultracompress plugin

15:29

<jdlee>

yeah...maven has really burned us thus far :|

15:30

<[pwgr]>

is it really benefitting in any area?

15:31

<jdlee>

i think so

15:31

<jdlee>

if we can solve some of the dep issues, I think we gain a lot

15:31

<jdlee>

pom "inheritance", project structure, etc.

15:31

<[pwgr]>

maven solves those issues ?

15:32

<[pwgr]>

or your saying those issues are with maven?

15:32

<jdlee>

the build has been a lot more stable recenently, so maybe The Powers That Be are getting a god handle on things

15:32

<jdlee>

solves those issuse

15:32

<jdlee>

not that Ant can't do that, but we get those free with maven

15:32

<jdlee>

"free"

15:32

<jdlee>

you pay with blood ;)

15:34

<[pwgr]>

I dont

15:34

<[pwgr]>

i had to learn the maven lession 2 years ago

15:37

<jdlee>

hehe

15:38

<jdlee>

i thought I'd drop maven some time ago, but as I learn it more, i've found it to be mostly usable

15:38

<jdlee>

there's still too much gotta-hold-your-mouth-right issues, but I've found them easier to stomach as I learn more

17:14

<robilad>

robilad waves at hani

17:15

<hani>

hey

17:15

<hani>

so, still no jersey people around?

17:15

<[pwgr]>

jersey?

17:15

<[pwgr]>

i dont think they hang here

17:15

<hani>

where do they hang out?

17:16

<[pwgr]>

no idea... this channel seems to have shrunk

17:16

<robilad>

jdlee: you get the issues for free? ;)

17:16

<[pwgr]>

some days i feel as though ive driven anyone off with my bad habit of spreading my misery and complaining

17:17

<jdlee>

over a year ago, I asked on the dev list for more Sun engineers to come on to IRC, and there was a brief flurry, but most either didn't stay on for more than a few weeks, or log on and just idle :(

17:17

<jdlee>

there are a few of us ;) now, but not as many as I'd like

17:17

<jdlee>

robilad: haha

17:17

<robilad>

fwiw, a recent build of trunk gave me 500+ jars in the m2 directory

17:18

<jdlee>

most of which probalby aren't used

17:18

<jdlee>

probably the biggest problem with transitive dep management

17:18

<robilad>

and for each one of those m2 had to go out and do half a dozen separate http requests

17:18

<robilad>

so it took about an hour.

17:19

<[pwgr]>

dispicable

17:19

<robilad>

in better systems, the repository access tool just goes out and fetches the contents once.

17:19

<jdlee>

there's some investigative work down with nexus, which would be awesome. if we could point to ONE repo, and let that one proxy/cache artifacts, I think a clean build would go a LOT faster, as there would be *1* network connection per artifact, not 12

17:19

<robilad>

and then figures out what needs to be done, locally, and does it.

17:21

<robilad>

well, the idea of having poms as little xml snippets that you're supposed to traverse remotely is just so enterprisey.

17:22

<jdlee>

well, it IS GlassFish Enterprise Server :P

17:22

<robilad>

probably works great if your repo is running on your laptop and you're hitting localhost like a pro

17:22

<jdlee>

what would you suggest as an alternative system?

17:22

<robilad>

is kind of silly if your repo is half the world away. ;)

17:23

<robilad>

dpkg, et. al. have a single file the package manager downloads (or just a partial update since the last update) that has all the information on packages in the repo.

17:23

<robilad>

so it connects once, then you have all the data you need localy to figure out what you need to update/download/etc.

17:25

<robilad>

so it's one network connection for all artifacts to get the information + N connections for N artifacts you actually download.

17:27

<jdlee>

robilad: ah! i see what you mean

17:27

<robilad>

but all java-specific solutions in that repo/module space suck pretty badly, so ... it's not really a big deal.

17:27

<jdlee>

you're caching on the client all the packages on the server, their deps, etc, so you know what to download and where to get it

17:27

<jdlee>

that might be a nice maven plugin

17:27

<robilad>

you kind of expect to see enterprise java repo management systems stuff to emulate linux from 1992 ;)

17:27

<jdlee>

haha

17:28

<jdlee>

rajivm: good morning :)

17:28

<jdlee>

rajivm: hani is looking for some jersey people. that's not your group is it?

17:29

<hani>

jdlee: yeah, async comms arent very useful in debugging

17:29

<hani>

jdlee: eg, right now I have an NPE deep in jersey, and I bet posting a stacktrace will be useless for it

17:29

<jdlee>

jdlee nods

17:30

<jdlee>

IRC+pastebin has been veru useful for me

21:03

<ankylose1>

http://pastebin.us/?show=d7725802e

21:04

<ankylose1>

i have created a SessionCustomizer but only postLogin is perfomed

21:05

<ankylose1>

why postBeginTransaction doesn't invoked when i do a entityManager.begin()

21:05

<ankylose1>

this code is on a standalone application

21:07

<ankylose1>

session.getEventManager().addListener(this);

21:09

<ankylose1>

:(

Drone v1.4 © 2002-2005 Uwyn RIFE powered