Time |
S |
Nick |
Message |
00:26 |
|
|
SLovenberg joined ##javaee |
00:37 |
|
SoniEx2 |
https://github.com/SoniEx2/TheModernClassicGame/commit/4680f77b709351455a804231367822330ad34334 |
00:37 |
|
SoniEx2 |
I'm stuck D: |
00:46 |
|
|
shoky left ##javaee |
00:47 |
|
|
SLovenberg joined ##javaee |
01:07 |
|
|
kobain joined ##javaee |
01:47 |
|
|
Quest joined ##javaee |
07:46 |
|
|
weyer joined ##javaee |
08:42 |
|
|
cem_ joined ##javaee |
08:43 |
|
cem_ |
how to store 4GB of data in database ? |
08:55 |
|
sess |
cem_: probably depends on DB vendor |
08:55 |
|
sess |
oracle has max blob size 4gb - 1 |
08:56 |
|
sess |
wait it was (4 GB - 1) * DB_BLOCK_SIZE |
08:56 |
|
sess |
whatever block size is |
09:05 |
|
|
callMeBaby joined ##javaee |
09:50 |
|
|
dangertools joined ##javaee |
09:50 |
|
|
dangertools joined ##javaee |
11:11 |
|
|
drspockbr joined ##javaee |
11:52 |
|
|
MegaMatt joined ##javaee |
13:32 |
|
SoniEx2 |
I'm not stuck anymore: https://github.com/SoniEx2/TheModernClassicGame/commit/afb41388f0c5340e30756efc56dba0b734eb9523 |
14:20 |
|
|
Naros joined ##javaee |
14:44 |
|
|
Quest joined ##javaee |
14:45 |
|
Quest |
password for my nick got compromised |
14:45 |
|
Quest |
strange that the guy gave it back... |
14:46 |
|
Quest |
and this was the 3rd time |
14:49 |
|
SoniEx2 |
O_o |
14:51 |
|
Quest |
Gues that was a guy, unless some hacker girl wants to impress me |
14:52 |
|
SoniEx2 |
or a hacker guy likes you? |
14:52 |
|
Quest |
it would be useless as iam not gay |
14:53 |
|
SoniEx2 |
you don't have to be gay to like a gay guy... |
14:54 |
|
Quest |
ya, but they are very dangerous to like |
14:54 |
|
Quest |
lolz |
14:55 |
|
SoniEx2 |
unless you're a hacker yourself...? |
14:58 |
|
Quest |
ahm. |
14:58 |
|
Quest |
ethical hacker. sometimes. |
14:58 |
|
Quest |
not a pro |
15:07 |
|
Quest |
SoniEx2, hows your project going |
15:07 |
|
Quest |
what technologies are you working on? |
15:08 |
|
SoniEx2 |
it's doing good |
15:08 |
|
SoniEx2 |
I'm not stuck anymore |
15:08 |
|
SoniEx2 |
https://github.com/SoniEx2/TheModernClassicGame/tree/master/src/com/github/soniex2/tmcg/ym2612 |
15:10 |
|
Quest |
hm |
15:56 |
|
|
Quest joined ##javaee |
16:09 |
|
|
jieryn_ joined ##javaee |
16:13 |
|
|
jieryn joined ##javaee |
16:17 |
|
|
cem_ joined ##javaee |
16:21 |
|
jieryn |
anyone know how to write an arquillian test that sets the javax.security.Principal when operating against @ArquillianResource URL |
16:21 |
|
jieryn |
? |
16:21 |
|
jieryn |
sorry, java.security.Principal |
16:29 |
|
cem_ |
i just cant predict future |
16:30 |
|
sfisque |
cem_ but you can have it tell you things… Future<?> |
16:32 |
|
cem_ |
sfisque: i thought can start servlet by now but working out things on postgres throw PassiveException(); |
16:56 |
|
|
sajjadg joined ##javaee |
17:08 |
|
|
SLovenberg joined ##javaee |
17:26 |
|
|
jypie joined ##javaee |
17:48 |
|
|
SLovenberg joined ##javaee |
18:01 |
|
whartung |
We had 108 threads trapped in a shared HashMap get today -- whee! |
18:01 |
|
whartung |
it was fun, 8 cores pegged! |
18:05 |
|
acuzio |
Tahts when you know Queues are needed |
18:05 |
|
whartung |
nah, it was a crummy cache. wrong part for the job |
18:52 |
|
Guest20878 |
http://stackoverflow.com/questions/20050447/cannot-load-params-in-struts-2-action any views ? |
19:33 |
|
|
weyer joined ##javaee |
19:50 |
|
|
WileTheCoyot joined ##javaee |
19:50 |
|
WileTheCoyot |
hello everyone |
19:50 |
|
whartung |
Wile T Coyot -- SUUPRA Speller! |
19:51 |
|
WileTheCoyot |
i want to read a properties file which is located (under eclipse) in <Project Name>/web/WEB-INF/prop.properties |
19:52 |
|
WileTheCoyot |
how could i refer to this file in a sinple java class? |
19:52 |
|
WileTheCoyot |
for now i use private static final String FILE = "/web/WEB-INF/properties/c3p0.properties"; |
19:52 |
|
WileTheCoyot |
InputStream inputStream = this.getClass().getClassLoader().getResourceAsStream(FILE); |
19:53 |
|
WileTheCoyot |
this give me null pointer exception |
19:53 |
|
WileTheCoyot |
whartung: what do you mean? |
19:56 |
|
whartung |
Out here, where the coyotes eat several cats per day, we spell it with an extra E |
19:57 |
|
WileTheCoyot |
whartung: i know but my nick doesn't end with E :) |
19:57 |
|
whartung |
I noticed |
20:00 |
|
WileTheCoyot |
how could i refer to /web/WEB-INF/ in java classes? |
20:00 |
|
whartung |
refer to it how? |
20:00 |
|
WileTheCoyot |
i want to load a properties file under that dir |
20:02 |
|
whartung |
ServletCotext has getRealPath(), so MIGHT be able to use ctx.getRealPath("/WEB-INF/yourfile.properties") |
20:03 |
|
whartung |
that would get the actual file on the filesystem |
20:03 |
|
whartung |
not sure if WEB-INF is hidden from getRealPath tho |
20:03 |
|
whartung |
if it is, you can getRealPath("/") and then build the WEB-INF/propfile parts |
20:03 |
|
WileTheCoyot |
whartung: unfortunately i'm not in a servlet |
20:04 |
|
whartung |
everything is in a servlet |
20:04 |
|
whartung |
if it's in a web ap |
20:04 |
|
whartung |
unless it's a thread you started |
20:04 |
|
whartung |
filters have access to that as well |
20:04 |
|
whartung |
what do you have? |
20:04 |
|
WileTheCoyot |
even if i don't extens HttpServlet? |
20:04 |
|
whartung |
what do you hae? |
20:05 |
|
whartung |
gotta step out... |
20:05 |
|
whartung |
biab |
20:05 |
|
|
SLovenberg joined ##javaee |
20:06 |
|
WileTheCoyot |
i have a storage class (which doesn't extends HttpServlet) that loads the db pool and take the parameter from that file |
20:08 |
|
WileTheCoyot |
http://pastebin.com/9H1BPKpf |
20:18 |
|
WileTheCoyot |
the path of the file is /web/WEB-INF/pool.properties |
20:19 |
|
|
javier joined ##javaee |
20:19 |
|
WileTheCoyot |
but tomcat give me nullPointerException |
20:26 |
|
|
weyer joined ##javaee |
20:40 |
|
|
weyer joined ##javaee |
21:06 |
|
|
cem_ joined ##javaee |
21:12 |
|
jieryn |
i have a @ManyToMany and a DAO with JPA criteria, i want to select just a subset of the mapping... |
21:12 |
|
jieryn |
the collection is Fetch.LAZY, but i can't seem to construct a criteria query to get it |
21:13 |
|
sfisque |
does the child have a back reference to the parent? |
21:13 |
|
sfisque |
you can do it two ways. link from child to parent, OR parent_.join( child.class ); |
21:16 |
|
jieryn |
there's only one reference |
21:17 |
|
jieryn |
criteriaQuery.join ? |
21:19 |
|
jieryn |
nope |
21:20 |
|
whartung |
"select just a subset"? |
21:21 |
|
whartung |
you'll have to specifically query for that, you can't filter on a property join |
21:21 |
|
sfisque |
can use .fetch() if need to specify limits on the joined entity |
21:22 |
|
jieryn |
i do |
21:22 |
|
sfisque |
or, like i said earlier, you can link from child to parent and put criteria on the child |
21:23 |
|
jieryn |
i have @Entity which has @ManyToMany(LAZY) Collection<Blah> blahs; |
21:23 |
|
sfisque |
oh MtoM, yeah, that's a bit trickier |
21:23 |
|
jieryn |
i have a DAO where i want to findAllBlahs(Entity, Range) |
21:23 |
|
jieryn |
i could just entity.getBlahs().subset(x, y) or something |
21:24 |
|
jieryn |
but there may be 20-30k sub rows |
21:24 |
|
jieryn |
seems stupid to fetch them all just to winnow them down |
21:24 |
|
whartung |
what are you trying to do? |
21:24 |
|
jieryn |
get a subset of a @ManyToMany collection |
21:24 |
|
whartung |
what subset? How do you define it? |
21:24 |
|
jieryn |
@Entity class MyEntity { @ManyToMany Collection<MyBlah> myBlahs; } |
21:25 |
|
jieryn |
i don't want to fetch all myBlahs from the database, i only need a few at a time |
21:25 |
|
whartung |
no, what are the requirements of your subset? how do determine which ones to get |
21:25 |
|
jieryn |
there are no restrictions on it, i want to add sortby and first/max results only |
21:26 |
|
whartung |
well obviously there are restrictions. First 100, 1000 random ones, all the green ones, middle 250…what kind of criteria are you using to define your subset? |
21:26 |
|
whartung |
what's the use case? paging? sampling? reporting? |
21:26 |
|
jieryn |
i seriously and truely do not want to restrict any of them, i just want ot page them |
21:26 |
|
whartung |
ok, paging |
21:27 |
|
whartung |
so, paging sucks -- period. On any thing with volume. |
21:27 |
|
whartung |
it sucks in everything, but particularly in RDBMSs |
21:28 |
|
jieryn |
ok, that may be trtue, but i still need to craft a query which will even reference them |
21:28 |
|
whartung |
so, you need to do conscious work to manage it properly |
21:28 |
|
whartung |
the way to do it |
21:28 |
|
jieryn |
i can not even get a jpa criteria query working to reference them |
21:28 |
|
whartung |
is via a temporary table |
21:28 |
|
jieryn |
there is already a join table |
21:28 |
|
jieryn |
jpa automatically created one |
21:29 |
|
whartung |
nope, you need a temp table to index the elements in the result set that you are paging through |
21:29 |
|
jieryn |
i don't want to keep a reference to the actual pager |
21:29 |
|
whartung |
you HAVE to |
21:29 |
|
jieryn |
i just want to limit the results by first ordering and then setting the first/max |
21:29 |
|
jieryn |
i don't htink so |
21:29 |
|
whartung |
ok, don't…it's your database. |
21:29 |
|
sfisque |
lolz |
21:29 |
|
jieryn |
thanks. |
21:30 |
|
jieryn |
anyone else ? |
21:30 |
|
whartung |
don't worry, I'll welcome you back when you come to your senses. |
21:30 |
|
jieryn |
i wish to your god that did it for me |
21:30 |
|
sfisque |
another way you can do it is to model the join table as an entity, but that "can" cause performance issues if you build bad queries |
21:31 |
|
whartung |
"Can you help me find my keys?" "Where did you lose them." "Over there." "Why are you looking here?" "The lights better." |
21:32 |
|
jieryn |
i am happy to admit i may not be phrasing my question in a manner that you understand |
21:32 |
|
jieryn |
i don't know why you have to be such an asshole about it though |
21:32 |
|
whartung |
You want to page through a large result set, and order it. Yes? |
21:33 |
|
jieryn |
why is it that all database people are like my half retarded cousin that likes moonshine and picking bar fights? |
21:33 |
|
semiosis |
who doesn't like moonshine & barfights? |
21:33 |
|
jieryn |
:) |
21:34 |
|
whartung |
don't have to be half retarded to like moonshine and barfights. |
21:34 |
|
whartung |
So, you'r NOT paging a large result set? |
21:34 |
|
whartung |
that's not what this is about? |
21:34 |
|
jieryn |
i'm not a database person, but i'm fairly sure that 'paging' is a special kind of operation |
21:34 |
|
jieryn |
i am not holding a reference to a paged result |
21:35 |
|
whartung |
nope, it's not a special kind of operation. |
21:35 |
|
whartung |
especially not in a web environment. |
21:35 |
|
whartung |
client/server, different problem -- but that's not what folks do today. |
21:35 |
|
whartung |
as a rule |
21:37 |
|
whartung |
so, which part am I misunderstanding? |
21:38 |
|
semiosis |
jieryn: heres the strategy i used for a rest api.... one endpoint to give a count of items (# items available, # per page (constant), # of pages) then a list endpoint which requires a page number param |
21:38 |
|
whartung |
how do you fetch the data for a page semiosis |
21:39 |
|
semiosis |
GET /items?count=true; GET /items?page=1; GET /items?page=2; ..... |
21:39 |
|
semiosis |
basically |
21:39 |
|
whartung |
How do yu get the data from yur back end store |
21:39 |
|
semiosis |
oh right |
21:39 |
|
whartung |
yea, that part |
21:40 |
|
jieryn |
figuring out how to calculate offsets isn't really all that difficult |
21:40 |
|
whartung |
offsets is easy, fetching the data, not so much |
21:40 |
|
semiosis |
a named query, which selects the items & orders by date created, setMaxRestuls to MAX_RESULTS and setFirstResultPage to MAX_RESULTS * page |
21:40 |
|
jieryn |
referencing via criteria api a Collection<Blah> blahs where CriteriaQuery<Blah> and Root<MyEntity> |
21:40 |
|
whartung |
so you don't have a large table |
21:40 |
|
semiosis |
s/setFirstResultPage/setFirstResult/ |
21:41 |
|
semiosis |
whartung: how large is large? |
21:41 |
|
whartung |
large enough that your server hates you |
21:42 |
|
whartung |
SELECT … FROM … ORDER BY … LIMIT 100 OFFSET 500 |
21:42 |
|
jieryn |
ok, thanks |
21:42 |
|
jieryn |
that really isn't even close to what i'm asking about sorry |
21:42 |
|
jieryn |
i think i am not communicating what i am experiencing, thank you, i apologize |
21:43 |
|
semiosis |
whartung: servers love me |
21:43 |
|
whartung |
That is one of the worst SQL statements you can write :), specifically with large tables. |
21:43 |
|
whartung |
but it does depend on the WHERE and ORDER BY clauses |
21:43 |
|
|
kobain_ joined ##javaee |
21:44 |
|
whartung |
so I should change from "table size" to "result set size", since that's more germane |
21:46 |
|
semiosis |
whartung: what alternative do you suggest? |
21:46 |
|
whartung |
a work/temporary table |
21:47 |
|
whartung |
the problem isn't with the query per se, it's the fact that it's called more than once |
21:47 |
|
whartung |
so lets take the extreme case |
21:47 |
|
whartung |
of a million row table |
21:47 |
|
whartung |
SELECT * FROM table ORDER BY firstName LIMIT 100 OFFSET 1000 |
21:48 |
|
whartung |
assume firstName is not indexed |
21:48 |
|
whartung |
that query just sorted 1M rows |
21:48 |
|
whartung |
now, lets' get the next page |
21:48 |
|
whartung |
SELECT * FROM table ORDER BY firstName LIMIT 100 OFFSET 1100 |
21:48 |
|
whartung |
that query, also, just sorted 1M rows |
21:48 |
|
semiosis |
clearly |
21:48 |
|
whartung |
the original query is long gone |
21:48 |
|
semiosis |
so your solution is to sort the whole thing into another table then query that multiple times |
21:49 |
|
whartung |
so instead, you have a work table, ideally with a serial key |
21:49 |
|
whartung |
yea |
21:49 |
|
semiosis |
isn't that what having an index on firstname is? |
21:49 |
|
whartung |
INSERT INTO worktable SELECT 0,primaryKey FROM Table |
21:49 |
|
whartung |
yes. |
21:50 |
|
whartung |
but you can either use an index to SORT the result set, or NARROW the result set, rarely both (unless they're the same thing, right?) |
21:51 |
|
whartung |
so obviously, this works great "SELET * FROM table WHERE firstName >= "A" and firstName < "B" ORDER BY firstName |
21:51 |
|
semiosis |
but not when ordering by lastname |
21:51 |
|
whartung |
current, i fyou want last name your SOL :) |
21:51 |
|
|
jaip joined ##javaee |
21:52 |
|
whartung |
and, again, if you have a narrow raw result set, the sort is "cheap" |
21:52 |
|
whartung |
obviously once you capture the worktable, your result set is stale, but such is life. |
21:53 |
|
semiosis |
http://dev.mysql.com/doc/refman/5.5/en/multiple-column-indexes.html |
21:53 |
|
semiosis |
looks helpful |
21:53 |
|
whartung |
but now you can query by the index first column, and use that instead of LIMIT and OFFSET |
21:54 |
|
whartung |
if you're narrowing and ordering by the same indexed columns, then you're fine. but most of the time, folks do't do that |
21:55 |
|
whartung |
most pages page by number, not data (rows 1-50, 51-100 vs all the A's and all the B's) |
21:56 |
|
whartung |
and its completely different in the Client Server world |
21:56 |
|
whartung |
you can do "SELET * FROM TABLE ORDER BY firstName" and just hold the cursor open over the DB connection |
21:56 |
|
whartung |
you typically can not do that in a stateless web app |
21:56 |
|
whartung |
holding connections is verboten |
21:57 |
|
whartung |
(you can do it with a stateful EJB, however ;) ) |
21:57 |
|
whartung |
but DBA's don't like random users holding there DBs hostage for arbitrary amounts of time |
21:59 |
|
whartung |
alas, it wasn't this way back in the golden days of ISAM…then it was easy :) |
22:11 |
|
cem_ |
when to use syslog in postgres |
22:13 |
|
cem_ |
pdurbin: when to use syslog in postgres |
22:22 |
|
whartung |
when you want to route your postgres logs to a syslog server? |
22:23 |
|
sfisque |
in production, so that IT doesnt have to monitor multiple log sources per server |
22:37 |
|
cem_ |
there are so many log option in postgres i guesss other then default log we can choose |
22:38 |
|
cem_ |
one of the feature i liked is inheritance |
23:07 |
|
whartung |
log inheritance?? |
23:14 |
|
cem_ |
no ...colum inheritance |
23:28 |
|
cem_ |
org.postgresql.util.PSQLException: FATAL: password authentication failed for user "postgres" |
23:29 |
|
cem_ |
even after entering correct password it says wrong |
23:34 |
|
whartung |
I guess it's not the correct password then |
23:35 |
|
cem_ |
whartung: but i entered the same password to get into CLI and it worked psq; -U postgres and i entered the same password and i logged in |
23:35 |
|
cem_ |
psql* |
23:36 |
|
whartung |
then there's something else wrong with your request to PG |
23:41 |
|
cem_ |
how to check bit mask ? |
23:42 |
|
cem_ |
host postgres postgres 127.0.0.1/32 password part of my pg_hba,config |
23:53 |
|
cem_ |
apparently i cant connect as superuser |
23:53 |
|
whartung |
not as root, no |
23:53 |
|
cem_ |
A common setup is that you cannot connect as the database user postgres unless you are running as the OS user postgres. |
23:53 |
|
whartung |
I haven't seen that |
23:54 |
|
whartung |
sure it could be done |
23:56 |
|
cem_ |
but how |
23:56 |
|
sfisque |
i believe the "default" postgresql setup works like that. it assumes your os user is the same as the db user |
23:56 |
|
sfisque |
but you can reconfigure postgresql to not do that |