Monday, 2014-08-11

*** tgnitidle has quit IRC06:29
*** tgnit has joined #npoacct07:03
*** tgnit has quit IRC09:33
*** tgnit has joined #npoacct09:46
*** tgnit has quit IRC09:57
*** tgnit has joined #npoacct10:00
*** tgnit has quit IRC10:28
*** tgnit has joined #npoacct10:59
*** tgnit has quit IRC11:04
*** tgnit has joined #npoacct11:13
*** bkuhn has joined #npoacct15:29
tgnitbkuhn:  greetings!15:32
tgniti have submitted lot of tests and will submit more, but some of my output is different from Travis - ci build.15:35
tgnittests passes on my system15:35
bkuhnare  you sure you're building against master?15:37
tgniti first rebase on upstream master15:39
tgnitinfact i redownloaded and reinstalled ledger to test..15:40
tgnitbut still the same result.15:40
bkuhnwhat are the specific differences?15:43
tgnitmainly printing in report, earlier, output was different for a reg --lot-prices --equity  test and now i dump print balance,  the commodities are in different order15:47
tgnitchecked various failed reports at
* bkuhn looks15:48
bkuhnok, I'll pull your branch too and look.15:51
bkuhntgnit: are all your tests all on your master branch now?15:53
tgnitcode_coverage  branch15:53
tgnitrebased on master15:53
bkuhnlast commit is 689112d3c0ffeb895ad4d8b9f61b50c5f537f0ed ?15:54
bkuhn(just making sure I'm looking at the right thign)15:55
bkuhntgnit: ^^^  I'm rebuilding at that commit id from scratch15:55
tgnitbefore the tests15:58
bkuhnAh, ok15:59
bkuhn272e3cb3799937dcd3c379de56b299a3cbde52ca is your last test commit?15:59
bkuhntgnit: ^15:59
bkuhnok, building from there and will look at test reports when it's done16:00
tgnitbkuhn: any luck?17:34
bkuhnI launched the tests running and was working on something else.  I'll look at output now17:58
bkuhntgnit: I get:18:21
bkuhnThe following tests FAILED:18:21
bkuhn          2 - LedgerMathTests (Failed)18:21
bkuhnErrors while running CTest18:21
bkuhnmake: *** [test] Error 818:21
bkuhnIs that what you get when you run it on your own?18:21
tgnitno,  it passes easily..18:22
bkuhnOk, so that's what I get on the command line18:22
tgniti have run clean build too with same output18:23
tgniti mean it passes on my system.18:24
bkuhnFWIW, I ran:18:24
bkuhngit clean -dfx; ./acprep --prefix=/home/bkuhn/hacks/Installs/TripunLedger3 --python  --jobs=2 debug make ; make tests18:24
tgnitshall i change the output to what Travis give  then?18:26
tgniti will try clear my system then18:27
bkuhnI think you need to figure out why you get a different result.18:27
bkuhnSomething is very wrong if you get a different result than both me and Travis.18:27
bkuhntgnit: Is there a verbose option I can give the test process that would help you?18:27
tgniti ran with that..  i know what output is different.18:28
tgniti will try once again with a cleaned system.18:29
bkuhnok, I was saying if I run it, it might give you moreinfo.18:30
bkuhn(I don't actually know how to get verbose output on the tests)18:30
bkuhnhow do you do it?18:31
tgnitctest - V18:32
bkuhnso make check just calls ctest. :)18:32
bkuhnI wasn't sure how it did it. :)18:32
tgnitbkuhn:  one thing more, shall i try adding unit test for value_t?  right now it is not there in mathtests18:36
tgnitctest -V -R Math18:37
tgnitrun only mathtests18:37
bkuhnOk, I get:18:37
bkuhn2: Test command: /home/bkuhn/hacks/ledger/MathTests18:37
bkuhn2: Test timeout computed to be: 9.99988e+0618:37
bkuhn2: Running 71 test cases...18:37
bkuhn2: /home/bkuhn/hacks/ledger/test/unit/ error in "testPrinting": check std::string("BALANCE(EUR 456, $ 123.00)") == bufstr.str() failed [BALANCE(EUR 456, $ 123.00) != BALANCE($ 123.00, EUR 456)]18:37
bkuhn2: *** 1 failure detected in test suite "math"18:37
bkuhn1/1 Test #2: LedgerMathTests ..................***Failed    0.09 sec18:37
bkuhnwhich seems to be the same error as in Travis?18:38
bkuhntgnit: as for value_t, I'm much more familiar with the regression tests than the unit test.... why are you asking if you should try that in particular?  I'm generally fine with you adding any tests you find useful.18:38
bkuhn(and cover code not yet covered)18:39
tgnitthere are some uncalled functions and unused branches in i thought may be i could cover it with unit testing18:40
bkuhnThat seems to me useful; I can't think of an obvious reason why not.  Can you>?18:48
tgnitbkuhn:  i try add as many as i can.18:49
tgniti will discuss with johnw or tbm if it is fine to add it under mathtests18:50
tgnitbkuhn:  including this printing test, once fixed  it 25 tests i think, way to go  :)18:57
bkuhntgnit: (re putting tests under math tests) Yeah, I'm not completely sure about the unit test structure.  johnw is definitely the one to ask.  If he's not around, I'd suggest submitting it that way in the PR and note that you weren't sure, and he'll tell you if it belongs somewhere else.18:58
bkuhntgnit: (re: 25 tests): good, and I see some have been accepted by johnw upstream already, I haven't counted how many though.18:58
tgnitbkuhn:  please do tell me if you feel the count is different.18:59
bkuhnI'll take a look.19:01
bkuhntgnit: I agree with your overall count of 25 in the tripun/code_coverage branch.19:05
bkuhntgnit: It seems that at least 10 have been accepted upstream, but I'd really prefer if you could get all 35 you write accepted upstream -- particularly figuring out why your equity test fails for everyone else but you.19:10
tgnitbkuhn: i will add to mathtests and create a pull request for johnw to review19:10
tgnitbkuhn: ok19:10
*** bkuhn is now known as bkuhnIdle19:55
*** bkuhnIdle is now known as bkuhn20:42

Generated by 2.12.1 by Marius Gedminas - find it at!