| Time |
S |
Nick |
Message |
| 01:53 |
|
|
tommmied joined ##javaee |
| 03:23 |
|
|
acuzio joined ##javaee |
| 03:31 |
|
|
SoniEx2 joined ##javaee |
| 03:32 |
|
|
tommmied joined ##javaee |
| 03:32 |
|
|
firebird1 joined ##javaee |
| 04:07 |
|
firebird1 |
damn jasper reports suck |
| 04:07 |
|
firebird1 |
cant understand |
| 04:32 |
|
|
tommmied joined ##javaee |
| 04:47 |
|
firebird1 |
i totally came with 53 tables still dont know this is correct :( |
| 05:01 |
|
|
firebird1 joined ##javaee |
| 05:33 |
|
|
tommmied joined ##javaee |
| 05:33 |
|
|
Bombe joined ##javaee |
| 05:45 |
|
|
firebird1 joined ##javaee |
| 05:45 |
|
firebird1 |
hi |
| 06:32 |
|
|
tommmied joined ##javaee |
| 06:49 |
|
firebird1 |
difference between cookbook and bok ? |
| 07:22 |
|
firebird1 |
hell need to report client on saturday |
| 07:22 |
|
firebird1 |
:'( |
| 07:30 |
|
firebird1 |
i wont let sfisque,pdurbin moveout today :D |
| 07:31 |
|
|
neuro_sys joined ##javaee |
| 07:32 |
|
|
tommmied joined ##javaee |
| 07:35 |
|
* firebird1 |
going to hell will be back with my smart phone |
| 08:02 |
|
|
kobain_ joined ##javaee |
| 08:32 |
|
|
tommmied joined ##javaee |
| 08:35 |
|
|
firebird1 joined ##javaee |
| 08:49 |
|
|
TomyWork joined ##javaee |
| 08:49 |
|
TomyWork |
hi |
| 08:51 |
|
TomyWork |
this might be off-topic, but i have a jboss running in a VM on my machine and I would like to create a link that automatically logs me in with a certain user/pass |
| 08:51 |
|
TomyWork |
err, into the management console |
| 08:52 |
|
TomyWork |
i tried http://user:pass server:port, but that just makes firefox spew warnings about attempted subterfuge and then the login prompt is back |
| 09:00 |
|
firebird1 |
what is the warning ? |
| 09:06 |
|
TomyWork |
does it matter? |
| 09:09 |
|
TomyWork |
it's one of these: http://lh5.ggpht.com/_0G45OzGsIro/TD93wlhLZlI/AAAAAAAADg0/XUJhc0JMalI/image_thumb%255B6%255D.png%3Fimgmax%3D800 |
| 09:15 |
|
|
knoppix joined ##javaee |
| 09:32 |
|
|
tommmied joined ##javaee |
| 09:33 |
|
firebird1 |
i see |
| 10:30 |
|
|
knoppix__ joined ##javaee |
| 10:32 |
|
|
tommmied joined ##javaee |
| 11:32 |
|
|
tommmied joined ##javaee |
| 11:40 |
|
|
acuzio left ##javaee |
| 12:19 |
|
|
sajjadg joined ##javaee |
| 12:22 |
|
|
Ziberius joined ##javaee |
| 12:22 |
|
|
SoniEx2 joined ##javaee |
| 12:32 |
|
|
tommmied joined ##javaee |
| 13:49 |
|
|
firebird1 joined ##javaee |
| 13:56 |
|
firebird1 |
hello world |
| 13:59 |
|
|
tommmied joined ##javaee |
| 13:59 |
|
|
jieryn joined ##javaee |
| 13:59 |
|
|
jieryn joined ##javaee |
| 14:14 |
|
|
kobain joined ##javaee |
| 14:30 |
|
|
Naros joined ##javaee |
| 14:46 |
|
sajjadg |
firebird1: you forgot the system.out.println(); |
| 15:09 |
|
firebird1 |
:D |
| 15:09 |
|
firebird1 |
where is sfisque ? |
| 15:09 |
|
firebird1 |
ill be back with my first report |
| 16:19 |
|
|
trollolol joined ##javaee |
| 17:26 |
|
pdurbin |
if (cause instanceof ConstraintViolationException) |
| 17:26 |
|
pdurbin |
fun with bean validation :) |
| 17:42 |
|
|
trollolol joined ##javaee |
| 17:45 |
|
|
sajjadg joined ##javaee |
| 18:35 |
|
|
sheenobu joined ##javaee |
| 18:41 |
|
|
sfisque joined ##javaee |
| 18:49 |
|
|
sajjadg joined ##javaee |
| 18:51 |
|
|
Naros left ##javaee |
| 19:06 |
|
whartung |
anyone looked at the recent JEE 8 survey? |
| 19:16 |
|
sfisque |
link? |
| 19:18 |
|
whartung |
https://blogs.oracle.com/theaquarium/entry/java_ee_8_community_survey |
| 19:21 |
|
whartung |
what do you think about Guice vs CDI sfisque |
| 19:24 |
|
sfisque |
depends on container. if you're in an EE container, why add unnecessary libs. if you're in a servlet container, you kind of have to go with either guice or spring |
| 19:25 |
|
whartung |
or weld, which is CDI |
| 19:25 |
|
whartung |
The one thing that attracts me to guice |
| 19:25 |
|
sfisque |
aye but weld is the reference implementation |
| 19:25 |
|
whartung |
is that it's explicit |
| 19:25 |
|
whartung |
@Inject MyBean bean |
| 19:25 |
|
whartung |
and you need to configure that, in Java actually, so it's type safe. |
| 19:26 |
|
whartung |
vs in CDi |
| 19:26 |
|
whartung |
you don't, it'll "hunt it down" |
| 19:26 |
|
whartung |
"hunt it down" == black magic |
| 19:26 |
|
sfisque |
unsure what you mean? |
| 19:26 |
|
whartung |
by default, in CDI, you can inject pretty much anything |
| 19:26 |
|
whartung |
all you need is an empty beans.xml |
| 19:26 |
|
whartung |
so it's pretty liberal "convention over configuration" |
| 19:27 |
|
whartung |
@Inject MyBean bean is about this ][ close to just be "new MyBean()" out of the box |
| 19:27 |
|
sfisque |
well according to the spec, there are very explicit rules on the convention. it's not black magic. it's pretty clear |
| 19:27 |
|
sfisque |
well, except the impl is free to pool/proxy/delegate as it sees fit |
| 19:27 |
|
whartung |
yes. |
| 19:27 |
|
whartung |
but that also means |
| 19:27 |
|
sfisque |
so you don't ahve to care about wether you new or obtain |
| 19:29 |
|
whartung |
that if a new class lib is added to your project, then the configuration implicitly is changed. Simple example, if you're injecting an interface, and now, suddenly, two implementation of that interface are now on the class path, a default injection that worked before with one, now no longer works, as there's a choice to be made. |
| 19:30 |
|
whartung |
so now, "suddenly", you code is not working as expected. |
| 19:30 |
|
sfisque |
well if you're adding new libs that provide parallel impl, one would assume you have vetted the code, which means you'd detect the ambiguity and cleared it up with a specifier annotation |
| 19:31 |
|
sfisque |
qualifier annotation |
| 19:31 |
|
whartung |
maybe, maybe not :) |
| 19:31 |
|
sfisque |
which is about 4 min of code |
| 19:31 |
|
sfisque |
well, you're confusing a process breakdown with a coding standard |
| 19:31 |
|
sfisque |
if the team isnt vetting code, it's not the platforms job to save them |
| 19:31 |
|
whartung |
my point is that it makes the system "more" environmentally sensitive. |
| 19:32 |
|
sfisque |
EE is ALL ABOUT ENVIRONMENT |
| 19:32 |
|
whartung |
how do you mean |
| 19:33 |
|
sfisque |
we not talkinga bout a simple swing toy. this is enterprise. stitching together systems and services, funneling data/messages over conduits. everything is environment in EE. to not accept that is to go in naively |
| 19:34 |
|
sfisque |
IMO that's one of the big dangers of ejb3. it simplified the coding, and people lost sight of the fact that it's still enterprise with all the issues. they just removed alot of the boiler plate and config |
| 19:34 |
|
whartung |
that's kind of the thing right there. |
| 19:34 |
|
whartung |
when the "simple stuff", "just works", then as you use the system, you don't spend much time on the edge cases. |
| 19:35 |
|
whartung |
so they're not foremost in your consciousness |
| 19:35 |
|
sfisque |
people still need to read the fracken spec. i had people asking me questions at the last gig that were straight up answered in the java ee spec pdfs |
| 19:35 |
|
sfisque |
it was saddening |
| 19:35 |
|
whartung |
so when things change, you start getting in to the "darker corners" of the specification. |
| 19:36 |
|
sfisque |
that's another failing of enterprise implementation. too many shops have phased out the role of arch and that has lead to crummy cohesion between systems |
| 19:36 |
|
whartung |
and when that happens, folks are "cutting and pasting" there way in to "fixes they found on the internet" |
| 19:36 |
|
sfisque |
aye. but someone on the team should be lurking there. every EE team needs "the dude in the shadows" |
| 19:36 |
|
sfisque |
aye |
| 19:37 |
|
whartung |
it gets tiring having to be responsible for 1000 pages of specification when in day to day, you use only 50 of them. |
| 19:38 |
|
whartung |
I was thinking of looking at something like this: http://www.jooq.org/ |
| 19:38 |
|
|
kobain joined ##javaee |
| 19:39 |
|
whartung |
type safe SQL building in java, vs JPA. Something more explicit than JPA, but not as painful as raw SQL |
| 19:39 |
|
sfisque |
consider the tradeoff. write your own TX framework, connection pooling framework, directory services framework, and all the other stuff that EE provides. so the big question is… does the team learn the F'in platform, or do you reinvent the wheel |
| 19:39 |
|
whartung |
but there's also simplifying the stack to rely on only a few fundamental services. |
| 19:39 |
|
whartung |
(not the JEE stack, the "development stack" so to speak) |
| 19:40 |
|
sfisque |
EE-lite. you have that starting in EE6. the Lite spec is for "minimal containers". the downside is, there does not seem to be much motion in delivering EElite containers atm. |
| 19:41 |
|
sfisque |
what we need is a good EELite impl published by a big name, like apache, jboss, ibm, or oracle |
| 19:41 |
|
sfisque |
so we can finally have TC+spring jettisoned into orbit like it should be |
| 19:42 |
|
whartung |
just seems that something like Guice, with its explicit config, a TX wrapper of some kind, some JOOQ like thing (maybe even mybatis), and a RESTy service framework, just seems to hit a lot of primary needs to me |
| 19:42 |
|
whartung |
doesn't TomEE do that? |
| 19:43 |
|
whartung |
(I really don't know the diff between TomEE and Geronimo -- why apache needs two containers ) |
| 19:43 |
|
whartung |
write now, I'm sorta fighting a web service that "goes away" after a deploy "sometimes" |
| 19:43 |
|
sfisque |
i'm guessing having it "parted out" like that makes it easier to sell to the suits. everyone at the CxO level has a boner for cheap spring developers |
| 19:44 |
|
sfisque |
"look! it's Tomcat…. just pre-vetted with extra techs that we'd end up vetting ourselves anyway!" |
| 19:44 |
|
whartung |
EJB Lite is really close |
| 19:44 |
|
sfisque |
CxO pops a boner, calls india/russia/china |
| 19:44 |
|
whartung |
:) |
| 19:44 |
|
whartung |
"Send in the Indian!" |
| 19:44 |
|
whartung |
*Indians |
| 19:44 |
|
sfisque |
aye. only issue is, no one's pushing a viable production EELite container :-( |
| 19:45 |
|
sfisque |
we have a spec, and no good contenders |
| 19:45 |
|
whartung |
CDI + Jersey + Something-Not-JPA |
| 19:46 |
|
whartung |
not enough experience with jersey yet to hate it :) |
| 19:46 |
|
whartung |
so, i'll roll with it |
| 19:46 |
|
sfisque |
myBatis seems to be the next "flavor of the weak" |
| 19:46 |
|
whartung |
it's been around forever |
| 19:46 |
|
sfisque |
not ibatis. mybatis which replaces all the OO with xml config |
| 19:46 |
|
sfisque |
because we all know EE needs MORE xml |
| 19:46 |
|
whartung |
this Jooq thing is compelling because it's type safe, even though it relies on a code generator from the schema. |
| 19:47 |
|
whartung |
there's something warm and fuzzy about updating the schema and having the compiler a better chance at breaking things |
| 19:47 |
|
sfisque |
jpa does that too. so… i have to ask… what's the compelling part? |
| 19:47 |
|
whartung |
no object management |
| 19:47 |
|
whartung |
just query building |
| 19:48 |
|
whartung |
no cascades no lazy fetches, etc. |
| 19:48 |
|
sfisque |
you can turn OM off. just query into detached objects |
| 19:48 |
|
sfisque |
they just become plain VOs |
| 19:48 |
|
sfisque |
i'm smelling solution looking for a problem |
| 19:49 |
|
whartung |
my overall goal is to be able to look at the code and know what it's doing. there's a lot of stuff today happening behind the scenes, and code doesn't mean what it says any more. |
| 19:49 |
|
sfisque |
gotcha |
| 19:49 |
|
whartung |
and I'm tired of debugging frameworks and not my code. |
| 19:50 |
|
whartung |
tired of walking on eggs and upsetting someone else's 50K lines of code I didn't write, but I'm responsible for. |
| 19:50 |
|
whartung |
"nothing changed and the deploy is failing" |
| 19:50 |
|
sfisque |
i do not think that ever goes away. switching frameworks only "moves the problem" |
| 19:50 |
|
whartung |
oh, I know…"Yea, the grass may be greener -- but there's more of it to mow." |
| 19:51 |
|
sfisque |
the other option is to fragment and make everything tiny toy apps that are portaled |
| 19:51 |
|
whartung |
that brings on other complexities -- yay -- integration! |
| 19:51 |
|
sfisque |
but then the problem becomes scale |
| 19:51 |
|
sfisque |
aye |
| 19:51 |
|
whartung |
that's funny tho |
| 19:51 |
|
whartung |
someone on tomcat room |
| 19:51 |
|
sfisque |
"moves the problem" |
| 19:51 |
|
whartung |
installing some app |
| 19:51 |
|
whartung |
that's deployed as HUNDREDS of WARs |
| 19:51 |
|
whartung |
"wut?" |
| 19:52 |
|
sfisque |
W I N!!!! |
| 19:52 |
|
whartung |
yea that was kind of scary |
| 19:52 |
|
whartung |
"war != cgi-bin script, k?" |
| 19:52 |
|
sfisque |
but but but but!!!! |
| 20:03 |
|
whartung |
how do you turn off managed objects outside of a transaction? |
| 20:03 |
|
whartung |
or I should say inside |
| 20:08 |
|
sfisque |
http://docs.oracle.com/javaee/6/api/javax/persistence/criteria/CriteriaBuilder.html#construct(java.lang.Class,%20javax.persistence.criteria.Selection...) |
| 20:08 |
|
sfisque |
javaeebot lucky CriteriaBuilder construct example |
| 20:08 |
|
javaeebot |
sfisque: https://blogs.oracle.com/ldemichiel/entry/the_typing_of_criteria_queries |
| 20:09 |
|
whartung |
where does Customer_.id come from? |
| 20:09 |
|
sfisque |
the result of a query that uses a constructed selection is auto-detached on delivery |
| 20:12 |
|
sfisque |
the paradigm is that you use an @Entity, but instead of fetching the full object, you fetch only the columns you need, into a constructor that matches the selected columns, and you get a detached object that is only populated with the data you need rather than the full tree with all the management |
| 20:12 |
|
sfisque |
perfect for times when you need to fetch the elements of a pop up list or render leaves in a tree in report generation and you don't want to pay the full price of a full fetch |
| 20:13 |
|
whartung |
that example doesn't jive with the javadoc |
| 20:15 |
|
|
neuro_sys joined ##javaee |
| 20:15 |
|
sfisque |
javaeebot google CriteriaBuilder construct example |
| 20:15 |
|
javaeebot |
sfisque: The Typing of Criteria Queries (Linda DeMichiel's Blog) - Blog: Oracle: <https://blogs.oracle.com/ldemichiel/entry/the_typing_of_criteria_queries>; Chapter 9. Criteria Queries: <https://docs.jboss.org/hibernate/entitymanager/3.6/reference/en/html/querycriteria.html>; JPA Criteria API Queries (CriteriaBuilder, CriteriaQuery) - ObjectDB: <http://www.objectdb.com/java/jpa/query/criteria>; Java (2 more messages) |
| 20:15 |
|
|
neuro_sys joined ##javaee |
| 20:17 |
|
whartung |
yea that blog returns a CriteriaQuery object, and then they call cq.from(..) on it, returning a Root<..>, but javadoc doesn't have a "from"method on criteria query |
| 20:50 |
|
semiosis |
feline.speak() |
| 21:01 |
|
sfisque |
inherited from AbstractQuery |
| 21:24 |
|
sfisque |
what is a good tool to redirect port 80 to my ee container other than using apache as a proxy. i could use stunnel but i dont want all of the traffic encrypted so i'd rather just funnel through straight without having jboss run as root. |
| 21:30 |
|
whartung |
linux can't run privileged ports on unprivileged processes yet? |
| 21:37 |
|
|
mbc joined ##javaee |
| 21:37 |
|
sfisque |
afaik no. still need root to bind to under 1024. |
| 21:37 |
|
sfisque |
but i don't let root run java because it's a huge security hole |
| 21:37 |
|
sfisque |
currently i have apache do it because it does suid well |
| 21:37 |
|
whartung |
yea |
| 21:38 |
|
sfisque |
but i'd rather run something lighter weight than apache just to proxy because apache isn't doing anything else |
| 21:38 |
|
whartung |
you consider a server running java as a security threat? |
| 21:38 |
|
whartung |
does it consume any real memory doing this? |
| 21:38 |
|
sfisque |
no, but running java as root is never a good idea |
| 21:38 |
|
whartung |
I mean, really, what's the difference between a proxy and an http server. |
| 21:38 |
|
sfisque |
not sure, but apache has all sorts of facilities that i don't need |
| 21:39 |
|
sfisque |
all i want is port 80 redirect and that's it. don't need the process to do any other magic |
| 21:39 |
|
whartung |
ah, the good old days of inetd :) |
| 21:40 |
|
sfisque |
guess i'll google. i figured ask here first to see if there was a recommended solution |
| 21:40 |
|
whartung |
yea I don't know of anything |
| 21:41 |
|
sfisque |
lots of hits for iptables. but i never liked the configuration of iptables. not very clear or obvious |
| 21:41 |
|
whartung |
yea |
| 21:41 |
|
whartung |
dangerous :) |
| 21:41 |
|
whartung |
"Wait, what happened to ssh?" |
| 21:42 |
|
whartung |
$ iptables save<ENTER> … NO CARRIER |
| 21:42 |
|
sfisque |
lolz |
| 21:42 |
|
sfisque |
i guess i could just build a ssh tunnel in a bash script. not elegant, nor fault tolerant |
| 21:43 |
|
whartung |
I routed LP over port 25 using inetd once, that was fun - thankfully nobody mailed use anything haha |
| 21:43 |
|
sfisque |
oh man |
| 21:43 |
|
whartung |
that was pretty cool back in the day |
| 21:44 |
|
whartung |
we had our main distribution system, ran on an HPUX machine, green screen stuff. |
| 21:44 |
|
whartung |
and once a year |
| 21:44 |
|
whartung |
we had an off site location, so we wanted to connect them. |
| 21:44 |
|
whartung |
what we did was we built up a simple linux box, and that gave us two terminals -- the console, and one of the serial ports. |
| 21:44 |
|
whartung |
the other serial port was for the modem. |
| 21:45 |
|
sfisque |
so basically a relay box |
| 21:45 |
|
whartung |
basically a simple terminal server. |
| 21:45 |
|
whartung |
originally we had them call our own modem and connect directly to our network, but they wanted to cut down the long distance charges |
| 21:46 |
|
whartung |
so they signed up with a local internet provider, and we routed the telnet (TELNET I TELL YOU!!) traffic from the linux box to our machine. |
| 21:46 |
|
whartung |
but we need to print locally. |
| 21:46 |
|
whartung |
before we just set the linux box up as a remote printer off of the main server -- ez mode. |
| 21:46 |
|
whartung |
but LP wouldn't route over the new provider |
| 21:46 |
|
sfisque |
looks like iptables is "the" way to do it on linux. guess i need to read up and refamiliarize myself |
| 21:47 |
|
whartung |
so I picked 25, since I knew if ANYTHING had to route, it was port 25 :) |
| 21:47 |
|
sfisque |
lolz |
| 21:47 |
|
sfisque |
a safe assumption |
| 21:47 |
|
whartung |
so, the spooler on our side fired off a script that opened up port 25 on the remote machine. |
| 21:48 |
|
whartung |
and on that I had the simplest script you could imagine tied to inetd: "cat | lp" |
| 21:48 |
|
whartung |
worked like a champ :) |
| 21:48 |
|
whartung |
it basically echoed anything coming in over port 25 to the printer |
| 21:49 |
|
sfisque |
the days of duct tape and linux voodoo |
| 21:49 |
|
whartung |
unix glory days |
| 21:50 |
|
whartung |
being able to write simple network server in shell scripts…them were the days! |
| 21:50 |
|
sfisque |
it's funny. i'm not a fan of software patents, but there are some that really deserve the patents they got. pipes was just an astounding leap |
| 21:50 |
|
whartung |
heh |
| 21:50 |
|
whartung |
I used the same technique to turn a 4GL program in to a proxy db server |
| 21:51 |
|
sfisque |
:-D |
| 21:51 |
|
whartung |
takes a sql command over stdin, feeds it to the DB, dumps it back in CSV |
| 21:51 |
|
whartung |
"secure" |
| 21:51 |
|
whartung |
*snort* |
| 21:51 |
|
sfisque |
ROFL |
| 21:52 |
|
whartung |
we didn't have ubiquitous connectivity in them days :) |
| 21:52 |
|
sfisque |
yah. when rutgers uninstalled talk, a bunch of us figured out we could use pipes and write to mimic what talk did. the sysops didnt care much for that |
| 21:52 |
|
whartung |
yea nice |
| 21:52 |
|
whartung |
pipes are the bomp |
| 21:52 |
|
sfisque |
aye that |
| 21:52 |
|
whartung |
SOA 101 :) |
| 21:52 |
|
sfisque |
day 1: learn bash and pipes |
| 21:52 |
|
sfisque |
day 2: h4X0rz! |
| 21:53 |
|
sfisque |
day 3: ways to hide when sysops knocks on your door |
| 21:53 |
|
whartung |
BitD some friends of mine wrote a COMPASS simulator. COMPASS was the machine code for the CDC Cyber mainframes. |
| 21:53 |
|
whartung |
they could then execute the system binaries on their emulator…boy howdy,did they find back doors lol |
| 21:54 |
|
sfisque |
heh heh |
| 21:54 |
|
whartung |
simpler times…. |
| 21:59 |
|
neuro_sys |
is it always unsafe to use threads in a ee container? |
| 22:00 |
|
sfisque |
not "always", but you can deadlock the container. besides, how would you syncrhonize the threads, you're prohibited from using syncrhonization primitives in the container (for possibly obvious reasons) |
| 22:01 |
|
sfisque |
if you NEED threads, use @Asynch, an MDB, or build the threads in a JCA adapter |
| 22:01 |
|
sfisque |
the real question is… what are you trying to do that requries threads in the EE container? |
| 22:01 |
|
sfisque |
the facility might already be provided |
| 22:12 |
|
sfisque |
basically. there is nothing to stop the creation of threads IN the ejb container, but the spec very explicitly says, DO NOT DO THIS |
| 22:13 |
|
sfisque |
effectively there are three "no no's" in EJB land. explicit creation of threads, explicit file i/o, synchronization other than transactional demarcation |
| 22:42 |
|
whartung |
heh |
| 22:42 |
|
whartung |
that said, I do all those things lol |
| 23:12 |
|
|
semiosis_ joined ##javaee |
| 23:15 |
|
|
javaeebot` joined ##javaee |
| 23:15 |
|
|
SEx2 joined ##javaee |
| 23:16 |
|
|
semiosis joined ##javaee |
| 23:16 |
|
|
semiosis joined ##javaee |
| 23:19 |
|
|
kobain joined ##javaee |
| 23:20 |
|
sfisque |
i'm curious how. if you're in an SLSB you have no instance vars to synchronize on. if you're in a SFSB you run the risk of locking up the transaction because it needs to flush a var that is being syncrhonized on. |
| 23:21 |
|
whartung |
synchronize on external singleton stuff, etc. |
| 23:25 |
|
|
semiosis_ joined ##javaee |
| 23:27 |
|
|
semiosis joined ##javaee |
