| Time |
S |
Nick |
Message |
| 01:06 |
|
pdurbin |
there's http://about.travis-ci.org/docs/user/languages/java/ but I wonder if Java EE is supported on Travis CI |
| 02:51 |
|
|
cem_ joined ##javaee |
| 02:57 |
|
cem_ |
hi is it possible to work in freelance ? |
| 02:58 |
|
cem_ |
????! |
| 03:06 |
|
cem_ |
no one tried freelance ? |
| 03:24 |
|
|
jenue joined ##javaee |
| 03:49 |
|
|
sross07 joined ##javaee |
| 04:42 |
|
cem_ |
sfisque: ? |
| 04:59 |
|
cem_ |
anyone there ? |
| 04:59 |
|
cem_ |
All in sleep ? >?> |
| 05:00 |
|
|
cem_ left ##javaee |
| 05:01 |
|
|
sleepyfdf joined ##javaee |
| 06:40 |
|
|
dangertools joined ##javaee |
| 06:49 |
|
|
Quest joined ##javaee |
| 07:00 |
|
|
pleroma joined ##javaee |
| 07:35 |
|
|
Bombe joined ##javaee |
| 07:53 |
|
|
weyer joined ##javaee |
| 08:35 |
|
|
Stooge joined ##javaee |
| 09:02 |
|
|
Quest joined ##javaee |
| 10:43 |
|
|
haklc joined ##javaee |
| 10:43 |
|
|
drspockbr joined ##javaee |
| 10:43 |
|
|
AartBluestoke1 joined ##javaee |
| 11:09 |
|
|
AartBluestoke1 joined ##javaee |
| 12:01 |
|
|
semiosis joined ##javaee |
| 13:23 |
|
|
haklc left ##javaee |
| 13:23 |
|
pdurbin |
hearing about this for the first time: Spring Tool Suite™ (STS) - http://spring.io/tools/sts |
| 13:24 |
|
Stooge |
you aren't alone :) |
| 13:25 |
|
pdurbin |
:) |
| 13:25 |
|
pdurbin |
it's Eclipse-based. I'm happy enough with Netbeans |
| 13:26 |
|
acuzio |
STS is shite |
| 13:26 |
|
acuzio |
<---- i have tried it |
| 13:30 |
|
pdurbin |
well, I'm able to tolerate Eclipse a little more now that I'm learning the keyboard shortcuts :) |
| 13:31 |
|
acuzio |
:-) |
| 13:31 |
|
acuzio |
STS doesn't gell and is slow as fuck |
| 13:42 |
|
|
knoppix joined ##javaee |
| 13:46 |
|
knoppix |
how to learn JEE from scratch |
| 13:47 |
|
knoppix |
i have same skill like oop laravel html5 php5 |
| 13:48 |
|
pdurbin |
knoppix: I'd start by watching this: https://blogs.oracle.com/arungupta/entry/screencast_37_java_ee_6 |
| 13:52 |
|
|
oO0Oo joined ##javaee |
| 13:54 |
|
knoppix |
pdurbin thx |
| 13:54 |
|
knoppix |
pdurbin: thx |
| 13:56 |
|
pdurbin |
knoppix: sure |
| 14:44 |
|
|
kobain joined ##javaee |
| 14:53 |
|
|
Naros joined ##javaee |
| 15:24 |
|
acuzio |
arungupta has a screencast ? |
| 15:33 |
|
pdurbin |
yep. oh I like this one too: Hello JavaEE 7 With Maven 3, JAX-RS 2.0 and JSON ...In 3 Mins : Adam Bien's Weblog - http://www.adam-bien.com/roller/abien/entry/hello_javaee_7_with_maven |
| 15:34 |
|
pdurbin |
code at https://gist.github.com/pdurbin/7652999 |
| 16:30 |
|
|
Quest joined ##javaee |
| 16:42 |
|
|
semiosis joined ##javaee |
| 16:55 |
|
* pdurbin |
finally gets around to installing Java 7 |
| 17:00 |
|
pdurbin |
hmm. this might be handy: http://apple.stackexchange.com/questions/69086/how-can-i-switch-between-the-two-java-versions-in-mountain-lion-now-that-java-preferences-is-gone |
| 17:05 |
|
acuzio |
pdurbin: you use os x ? |
| 17:06 |
|
pdurbin |
at work, yes. fedora at home |
| 17:08 |
|
acuzio |
hmm interesting |
| 17:14 |
|
|
knoppix joined ##javaee |
| 17:29 |
|
|
knoppix joined ##javaee |
| 17:51 |
|
|
weyer joined ##javaee |
| 18:35 |
|
|
balazare joined ##javaee |
| 18:41 |
|
|
SoniEx2 joined ##javaee |
| 19:40 |
|
|
balo joined ##javaee |
| 19:56 |
|
pdurbin |
setting up persistence is a pain... <jdbc-connection-pool datasource-classname="org.postgresql.ds.PGSimpleDataSource" (and etc.) in glassfish's domain.xml |
| 20:01 |
|
whartung |
wut? |
| 20:01 |
|
whartung |
put the keyboard down, step away from domain.xml |
| 20:02 |
|
Naros |
lol |
| 20:02 |
|
pdurbin |
in the end I used the glassfish gui to figure out the specific change I needed for domain.xml |
| 20:02 |
|
whartung |
you don't say? |
| 20:03 |
|
pdurbin |
netbeans says you can do it through netbeans: https://netbeans.org/kb/docs/web/mysql-webapp.html |
| 20:03 |
|
whartung |
you can set stuff like that in NB projects the NB will deploy when the app is run |
| 20:04 |
|
whartung |
when that happens to me it's more happenstance than by design |
| 20:04 |
|
pdurbin |
it's confusing. that's all I'm saying :) |
| 20:05 |
|
pdurbin |
thankfully someone documented what needs to be done in the glassfish gui: https://github.com/IQSS/dvn/blob/develop/doc/sphinx/source/dataverse-installer-main.rst#jdbc-resources |
| 20:05 |
|
whartung |
The GUI is a bit confusing because all of these things I'd consider required are dumped on a "additional properties" tab when you EDIT the connection. When you create one, it's more straightforward but it would be nice if some of those fields were "first class" fields (notably, URL, server, user, password) |
| 20:06 |
|
pdurbin |
whartung: EXACTLY. "additional properties" |
| 20:06 |
|
pdurbin |
I didn't test to see if I actually need JDBC30DataSource true or ConnectionAttributes ;create=true |
| 20:07 |
|
|
drspockbr joined ##javaee |
| 20:08 |
|
pdurbin |
ah, yes, hmm, it seems like netbeans may have created src/main/setup/glassfish-resources.xml for me during all my dinking around |
| 20:08 |
|
* pdurbin |
wonders if he needs it |
| 20:11 |
|
pdurbin |
apparently not |
| 20:15 |
|
|
balo joined ##javaee |
| 20:37 |
|
|
drbrule-away joined ##javaee |
| 20:38 |
|
|
drbrule-away left ##javaee |
| 20:50 |
|
|
kobain joined ##javaee |
| 21:09 |
|
|
weyer joined ##javaee |
| 21:20 |
|
whartung |
that's a NB only artifact pdurbin, it's not necessary for the application |
| 21:22 |
|
sfisque |
they are additional properties because they are driver specific and not part of the actual spec |
| 21:23 |
|
whartung |
yes, but things like host, dbname, user and password -- pretty common properties |
| 21:23 |
|
sfisque |
does not matter. some drivers do not need them |
| 21:25 |
|
pdurbin |
whartung: I removed it |
| 21:25 |
|
sfisque |
make a recomendation for them in jdbc 4.1 i guess |
| 21:27 |
|
|
knoppix joined ##javaee |
| 21:30 |
|
whartung |
No, it's not a jdbc issue, it's a GUI issue |
| 21:31 |
|
whartung |
the point being yes, not all driver need them, but MOST do. since MOST dbs are hosted over a network socket, so while it's not 100% the case, it's very likely 80-90% the case. |
| 21:34 |
|
sfisque |
right but you still have to provide the "attribute name". for one driver it might be "hostname", for another it might be "hostName", and for another it might be "Hostname". how is the GUI supposed to differentiate that when as far as the driver is concerned, it's just a key in a map |
| 21:34 |
|
whartung |
oic |
| 21:34 |
|
whartung |
no, that's valid |
| 21:34 |
|
|
kobain_ joined ##javaee |
| 21:34 |
|
whartung |
still a GUI issue -- GF defaults those based on the driver you select. |
| 21:35 |
|
whartung |
just saying they could have a better experience, especially for the top common dbs (oracle, mysql, pg, sql server) |
| 21:36 |
|
sfisque |
maybe a good recommendation for jdbc 4.1 have the driver's expose a dictionary of keywords, enum pairs where the enums are standardized keys, .. like "HostName", Enum.JDBC_HOST_NAME_KEY |
| 21:37 |
|
sfisque |
thent he container admin consoles could introspect what keys to bind to, and which things to solicit for those fields |
| 21:38 |
|
whartung |
yup |
| 21:38 |
|
whartung |
start your new gig yet? |
| 21:38 |
|
sfisque |
dont have one lined up yet. they extended me to the 20th though, people started freaking out |
| 21:39 |
|
pdurbin |
new gig? I missed this |
| 21:39 |
|
whartung |
ah -- good, I think :) |
| 21:39 |
|
whartung |
here's something for musing though. |
| 21:40 |
|
whartung |
How would you go about creating a transactional mail SENDER service. i.e. I want to mail a message, but only on commit. |
| 21:40 |
|
sfisque |
XA land |
| 21:40 |
|
whartung |
yea |
| 21:40 |
|
whartung |
so it seems to me |
| 21:40 |
|
whartung |
that basically |
| 21:41 |
|
sess |
just close the transaction before sending the mail? |
| 21:41 |
|
whartung |
that's not quite the same thing sess |
| 21:41 |
|
sess |
if the first succeeds that is |
| 21:41 |
|
|
kobain joined ##javaee |
| 21:41 |
|
sfisque |
right but how do you "roll back" and email send |
| 21:41 |
|
sfisque |
that's the question |
| 21:42 |
|
sess |
wait, you want to mail on successful commit or rolled back commit? |
| 21:42 |
|
whartung |
but I was thinking, that you'd simply snapshot and fsync the mail message to disk, identified by the transaction ID. That would be the "ready for commit phase". |
| 21:42 |
|
|
kobain joined ##javaee |
| 21:42 |
|
sfisque |
yah, you need a way to queue and rollback if you need to abort the email send |
| 21:42 |
|
whartung |
then when it does commit, I'd have a background task that's supposed to flush the messages to SMTP listener. |
| 21:42 |
|
whartung |
you also need a way to restart the transaction after it's "been committed" |
| 21:42 |
|
sfisque |
OR you can queue them in a MQ, with some type of "ready to consume" flag |
| 21:43 |
|
whartung |
the problem with MQ |
| 21:43 |
|
whartung |
is if the mail server is down |
| 21:43 |
|
sfisque |
right. asynch == no real tx semantics |
| 21:43 |
|
whartung |
the retry logic goes awry, and the message (ideally) ends up in the DL |
| 21:45 |
|
whartung |
the key think about XA, that most folks don't think about, is that after you've committed, the transaction is not supposed to fail. |
| 21:45 |
|
whartung |
so you have to manage that aspect. |
| 21:45 |
|
whartung |
there's a hidden "final commit" in XA that's not obvious to the developer |
| 21:46 |
|
sfisque |
aye. somethign to consider, front load all validation so that you're pretty much guranteed a successful commit after you "go remote" |
| 21:46 |
|
whartung |
and that's the what the persistent transaction store is for in GF |
| 21:46 |
|
whartung |
with the automated transaction semantics of the container, you're never quite in full control. |
| 21:47 |
|
sfisque |
BMT |
| 21:47 |
|
whartung |
but it's still a race condition |
| 21:47 |
|
sfisque |
aye |
| 21:47 |
|
whartung |
"INSERT IN TO DB" "SEND EMAIL" "COMMIT" -- not particularly atomic :) |
| 21:48 |
|
sfisque |
that's why there is a whole consortium around XA. no real solution, just many "ways" to solve the problem in context |
| 21:49 |
|
whartung |
yea |
| 21:49 |
|
whartung |
well, XA has in built retry semantics that the container manages for you as well |
| 21:50 |
|
sfisque |
the problem is, it's unclear how to build up an XA Tx if you have to roll your own (basically wrap a remote in-house service with your "local" db tx.) |
| 21:50 |
|
sfisque |
i couldnt discern how to build an XA TX from scratch in an EJB. the docs are unclear |
| 21:51 |
|
whartung |
what do you mean? |
| 21:54 |
|
sfisque |
i have a remote service, that talks JSON over web (not rest, but essentially like rest). i have an oracle db connection that i am funneling things to. the remote json service has no boundary demarcation, so i wanted to build a TX on my side, and have it interact with the oracle TX as well. from what i read, that's a basic XA context, but it's not clear to me how to create that XA tx and then bundle the two services into it. |
| 21:54 |
|
sfisque |
i could find no way to make/configure a BMT to be XA |
| 21:55 |
|
whartung |
oh, yea, I think you need a JCA adaptor to do XA |
| 21:55 |
|
whartung |
that's where the XA workflow is exposed |
| 21:56 |
|
sfisque |
that's what i was afraid of. i've written jca's before, but the dev cycle on this project is too short to build out something like that and also implement the other stuff. i was hoping i could "get away" with somethign the container would provide |
| 21:56 |
|
sfisque |
/cry |
| 21:56 |
|
sfisque |
QQ |
| 21:56 |
|
whartung |
yea |
| 21:57 |
|
whartung |
it would be nice to have a generic JCA, I wonder if that's practical, or if the JCA is already as "abstract as it can be". Something that exposed just the few methods you needs |
| 22:03 |
|
sfisque |
someone has to o have solved this problem… time to dive into google… hopefully some has built a generic harness i can leverage |
| 22:04 |
|
sfisque |
hazelcast looks interesting |
| 22:05 |
|
whartung |
what's that got to do with this use case? |
| 22:06 |
|
sfisque |
it's open source, so i can peek at their code to see how they build up the XA boundaries and mimic it |
| 22:06 |
|
sfisque |
they have XA support |
| 22:06 |
|
whartung |
oh |
| 22:06 |
|
whartung |
there's this too |
| 22:06 |
|
whartung |
https://github.com/dblevins/mdb-improvements |
| 22:07 |
|
sfisque |
hrm, possibly leveragable |
| 22:20 |
|
|
sfisque joined ##javaee |
| 22:32 |
|
|
sfisque joined ##javaee |
| 22:37 |
|
sfisque |
this might be what i need: http://txconnect.sourceforge.net |
| 22:41 |
|
whartung |
that looks pretty interesting |
| 22:44 |
|
sfisque |
now if can only get the guys who wrote the remote service to support rollback. currently we can only insert and update, there is no delete or abort method in the service profile. blech |
| 22:44 |
|
sfisque |
i had a cow when i discovered that |
| 22:45 |
|
sfisque |
so… you wrote a remote centralized service for an ESB, and you provided no rollback, abort, or delete option. hrm...... |
| 22:45 |
|
whartung |
yea, you get to do all that your self \o/ |
| 22:46 |
|
whartung |
his service isn't transactional, why would it rollback? |
| 22:46 |
|
sfisque |
if ONLY |
| 22:46 |
|
sfisque |
lolz.. i know… crazy talk |
| 22:46 |
|
* sfisque |
wanders off, muttering to himself |
| 22:46 |
|
whartung |
egads..what..WHY!?? http://d.pr/i/VNrw |
| 22:47 |
|
whartung |
hibernate??? |
| 22:48 |
|
whartung |
but that does looks pretty cool |
| 22:48 |
|
sfisque |
yah. i'll probably have to strip down the code if i end up using it |
| 22:48 |
|
whartung |
it doesn't expose the transaction semantics the way I'd like |
| 22:48 |
|
|
kobain_ joined ##javaee |
| 22:53 |
|
whartung |
I wonder if this needs to be updated for JEE 5+ |
| 22:53 |
|
whartung |
it was written in 2006 |
| 22:53 |
|
whartung |
probably not, an under appreciated fact about EJB -- it's backwards compatability. |
| 22:53 |
|
* whartung |
glares at maven 3 |
| 22:54 |
|
sfisque |
aye but with some selective annotations, i'm sure it could be streamlined |
| 22:54 |
|
sfisque |
gives the finger to LocalHome |
| 22:54 |
|
whartung |
heh |
| 22:54 |
|
whartung |
but it STILL works…we have a project with ENTITY BEANS /cough running in GF3 |
| 22:57 |
|
sfisque |
of course. i'd expect nothing less from a Sun product |
| 22:57 |
|
whartung |
just kudos to a standard |
| 22:57 |
|
whartung |
oh, and that was a web logic project that we ported to GF3 |
| 22:58 |
|
whartung |
it certainly wasn't drag and drop, but it wasn't "DO OVER IN NODE, YO!" either |
| 22:58 |
|
sfisque |
they were quite masterful in smoothing transitions. a stellar enterprise shop. too bad they hit the skids and needed oracle to buy them |
| 22:58 |
|
sfisque |
lolz |
| 22:59 |
|
sfisque |
my favorite cross port. we had a product that was going to deploy on WS6.x and i couldnt get the Friggen thing running properly so i dev'ed on GF. product dropped in with only a few config tweaks in the admin console. :-P |
| 22:59 |
|
sfisque |
made my life much more tolerable |
| 22:59 |
|
sfisque |
WS is a BEAST to manage/maintain |
| 23:00 |
|
sfisque |
typical ibm product. you need a phd just to install the damn thing and patch it appropriately |
| 23:00 |
|
whartung |
I have not, EVER, heard a good word for WS :) |
| 23:00 |
|
whartung |
never seen anyone sing praise of it. |
| 23:00 |
|
sfisque |
yah. ibm never figured out that software is no longer the purview of lanky dudes with pocket protectors hiding in the machine room |
| 23:00 |
|
whartung |
I mean, it must work…somebody must use it |
| 23:01 |
|
sfisque |
lolz, yeah… at gun point, maybe |
| 23:01 |
|
sfisque |
:P |
| 23:01 |
|
sfisque |
points the big gun at the dev…. "i said….. USE IT....." |
| 23:01 |
|
whartung |
"Oh goody -- 2 weeks of off site training to get a web app to run!" |
| 23:03 |
|
|
knoppix joined ##javaee |
| 23:03 |
|
whartung |
Frankly, EJB really turned a corner with the release of Sun app server 8, when they made it free to use (before GF, before open sourcing it) |
| 23:03 |
|
whartung |
getting EJB in to the hands of departmental app developers, that was a good thing |
| 23:08 |
|
sfisque |
aye. before then you really only had geronimo and that was clunky at best from my experience |
| 23:09 |
|
whartung |
you also had JBoss, but talk about clunky! |
| 23:09 |
|
sfisque |
yeah. before 4.2 jboss was |
| 23:09 |
|
sfisque |
… ok…. |
| 23:39 |
|
|
SoniEx2 joined ##javaee |
