2007-12-24 GnuCash IRC logs
00:01:33 <andrewsw> fun. annoying (diapers) but still fun.
00:01:39 <warlord> yeah, well.
00:03:11 <warlord> okay, I'll build 2.2 and check tests.
00:03:41 <andrewsw> my make is already running.
00:04:06 <warlord> well, we'll see who's finishes first ;)
00:04:24 <andrewsw> heh. should have run it on my laptop... my old k-7 desktop is getting clunky
00:04:41 <andrewsw> the lappy whips through it.
00:05:59 <warlord> Yeah, mine too
00:07:20 <andrewsw> I realise now that k-7 is on its third mobo... that chips was cheap tech 8 years ago. maybe its time to upgrade.
00:08:12 <warlord> yeah, probably..
00:12:09 <warlord> Huh... 2.2 passed "make check"
00:12:18 <andrewsw> dang.
00:12:39 <warlord> (it's not done yet)
00:12:43 <warlord> ahh, now it is.
00:12:47 <warlord> Hmm..
00:14:17 <andrewsw> so what does that mean?
00:14:35 <andrewsw> I saw your comment to andi5, but I don't really understand what make check does.
00:14:44 * andrewsw starts reading Makefile
00:15:21 <warlord> it runs the test suite
00:15:52 <warlord> This makes no sense.. In trunk this patch breaks the tests, but in 2.2 the test succeeds!
00:16:08 <andrewsw> is this 16619?
00:16:19 <warlord> Yeah
00:17:29 <warlord> Have you pulled down my new code yet?
00:17:36 <warlord> (it not, can you try something)
00:17:46 <andrewsw> no. I haven't.
00:17:49 <andrewsw> happy to
00:18:08 <warlord> Can you run "make check" in trunk? I presume you have 16619?
00:19:40 <andrewsw> yeah I'm up around 16678 or so. off it goes
00:19:54 <warlord> Thanks!
00:20:32 <andrewsw> oh this is so much faster
00:20:42 <warlord> heh
00:20:45 <andrewsw> (not done)
00:21:15 <warlord> well, did it fail in src/engine?
00:21:32 <andrewsw> yup
00:21:42 <andrewsw> 1 of 22 failed.
00:22:01 <warlord> Does it say something like:
00:22:06 <warlord> * 00:15:32 OTHER <gnc.lots> [xaccSplitAssignToLot()] Failed to split into two!
00:22:11 <warlord> then a long abort line.
00:22:12 <warlord> then:
00:22:16 <warlord> FAIL: test-lots
00:23:12 <andrewsw> yes, thats it. then:
00:23:20 <andrewsw> duude XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX prices
00:23:40 <warlord> that's from the next test.
00:24:04 <warlord> And I bet if you "svn up" and pull down my changes, when you rebuild it will pass "make check" again.
00:24:17 <warlord> But strangely, 2.2 passes the check even though 16619 was backported.
00:24:26 <andrewsw> wierd.
00:24:39 <warlord> yes. very weird
00:25:25 <andrewsw> so, what I can go is roll back to 16620 and make check there (I think, have to read some git man pages).
00:25:45 <andrewsw> I can't roll back before then though.
00:25:49 <warlord> okay..
00:25:59 <warlord> Yeah, good point.. I could go do that, too.
00:27:06 <andrewsw> oh, I can go to 16619. I have back to 16601
00:27:28 <warlord> I'll do 16620 ...
00:28:31 <andrewsw> heh. if 16620 does it, I'll turn in my key... :(
00:28:42 <warlord> hehe..
00:29:18 <warlord> would need to test 16619 specifically before turning in your key.
00:29:24 <warlord> (and 16618)
00:29:48 <andrewsw> okay... <phew>
00:30:21 <andrewsw> make check runs ./configure?
00:30:39 <warlord> Only if it needs to
00:31:01 <warlord> And there's no way that 16620 could break this test.
00:31:21 <andrewsw> okay. new branch (git)... So that means I'll hit gcc errors. gotta restart
00:31:31 <warlord> ok
00:32:01 <andrewsw> --disable-error-on-warning
00:32:12 <warlord> what gcc errors?
00:32:43 <andrewsw> some 4.2.2 stuff.
00:32:55 <andrewsw> throws some warnings
00:33:19 <warlord> :(
00:33:31 <warlord> still have 4.1.2 here on FC7
00:34:59 <andrewsw> there were some easy fixes that I put up before, but the current one I hit is a swig issue and it's convoluted -- some const parms that get passed on to children without being const
00:35:15 <warlord> Oh....
00:35:17 <warlord> Hmm...
00:35:20 <warlord> What version of swig?
00:35:28 <andrewsw> hey test-lots passed on 16619
00:35:48 <warlord> Hmmmmm... yes, and it passed on 16620
00:36:00 * andrewsw dodges bullet
00:36:12 <andrewsw> swig 1.3.33
00:36:38 <warlord> I guess tomorrow I get to start a binary search for the breaking changeset.
00:37:56 <andrewsw> I got a half-hour left on my movie, you want me to go north or south of halfway?
00:39:20 <warlord> Well, I'm testing r16655
00:41:28 <warlord> I'll let you know when it finishes. if it passes the check then you can go halfway from there to the head. if it fails the test you can go halfway back to 16620
00:42:00 <andrewsw> would a distcheck fix do it? 16674
00:42:18 <andrewsw> nope, nm
00:43:22 <andrewsw> poor phil. he's doing a lot, looks like.
00:43:27 <warlord> yeah!
00:43:34 <warlord> I'm proud of him!
00:43:57 <andrewsw> it's a big job and he's all alone. I root for him when I see his commits come through.
00:45:00 <warlord> Yeah, me too.
00:46:57 <warlord> Looks like 16655 passed.. So the breakage must be somewhere after that.
00:47:35 <andrewsw> I've been reading the logs looking for things that touch engine, is that the thing to do?
00:47:39 <warlord> so, somewhere between 16655 and 16714.
00:48:03 <warlord> that's what I did, but at this point I think a binary search is necessary. It might be a libqof change, or... who knows?
00:48:18 <andrewsw> alright, I'll do 16685
00:49:33 <warlord> Thanks. I'm going to bed. Feel free to do as many bisections as you want before you go to bed, and just keep tabs going here, if you could.
00:49:33 <warlord> Thanks!
00:49:46 <andrewsw> :) nighty night warlord
00:49:46 *** warlord is now known as warlord-afk
00:49:48 <warlord-afk> night
00:57:00 <andrewsw> 16684 passes
01:03:14 <andrewsw> hey I was failing on 16678 before and now I'm passing on 16684. huh?
01:03:36 <andrewsw> but 16703 fail!
01:05:29 <andrewsw> that points to my local branch failing, (had a couple changes locally when I failed on 16678). might be one of my changes.
01:07:13 * nbinon1 's head hurts just reading about it
01:07:15 <andrewsw> 16693 fails
01:07:21 <andrewsw> heh
01:08:59 <andrewsw> 16690 fails
01:11:25 <andrewsw> its 16690. that currency change breaks `make check`. my bad. :(
01:13:42 <andrewsw> looks like it doesn't like the double specification of major and minor both called "afghani"
01:20:23 <andrewsw> hmmm... why doesn't iso-currencies-to-c get run?
01:25:54 <nbinon1> quick question...why are there so many make *checks ? (check, check-am, distcheck, distcleanckeck, distuninstallcheck, installcheck)
01:26:07 <andrewsw> no idea at all.
01:26:15 <andrewsw> I think a lot of them are auto-generated.
01:26:24 <andrewsw> do you know makefile syntax at all?
01:26:40 <nbinon1> yes, somewhat
01:26:57 <andrewsw> okay, what does a '-' do at the front of a line in a stanza?
01:27:07 <andrewsw> looks like this:
01:27:18 <andrewsw> iso-4217-currencies.c: iso-4217-currencies.scm iso-currencies-to-c
01:27:36 <andrewsw> -chmod u+x ${srcdir}/iso-currencies-to-c
01:27:45 <nbinon1> I'm not sure
01:28:23 <andrewsw> there is a script iso-currencies-to-c that converts the currency file to a c structure. But it looks like it's not getting run. Iwonder if that '-' makes the command not run...
01:28:29 <andrewsw> only one way to find out.
01:31:16 <nbinon1> commands may be prefixed by an `@' character to prevent make from echoing the command line prior to executing it.
01:31:25 <nbinon1> They may also optionally be prefixed by a `-' character to allow the rule to continue if the command returns a non-zero exit code.
01:31:50 <andrewsw> hmmm... okay, that's not it. thanks
01:33:18 <andrewsw> holy crap it's a swig error in iso-currencies-to-c, I bet.
01:33:41 <andrewsw> nah, that's dumb
01:37:07 <nbinon1> for the make *checks... do you know which ones are useful to run? (kinda thinking of getting my build machine to run scheduled make checks)
01:37:40 <andrewsw> THe only ones I hear about are make check and distcheck. Ask the big boys though.
01:38:21 <nbinon1> ok, thanks
01:40:50 *** Pepe66 has joined #gnucash
01:41:26 <andrewsw> yup, I was right. looks like iso-currencies-to-c is not getting run (or is failing). If I run it manually, the iso-4217-currencies.c is updated. checking to see if that fixes make check.
01:42:44 <andrewsw> dang. still fails
01:50:04 <andrewsw> @tell warlord r16690 breaks make check. or at least reverting it allows it to pass. But why does reverting 16619 then allow it to pass again?
01:50:04 <gncbot> andrewsw: The operation succeeded.
01:51:17 <andrewsw> @tell warlord in the process I discovered that src/engine/iso-currencies-to-c is not being run during make, but that doesn't seem to be the source of the problem. I ran it manually and it still failed make check.
01:51:17 <gncbot> andrewsw: The operation succeeded.
01:54:19 <nbinon1> nightime for me
01:54:29 <andrewsw> goodnight nbinon1
01:54:35 *** nbinon1 is now known as nbinont-afk
01:56:17 *** basanta has joined #gnucash
02:11:31 *** Pepe66 has quit IRC
02:15:38 *** Pepe66 has joined #gnucash
02:33:43 <andrewsw> @tell warlord ignore comments about iso-currencies-to-c
02:33:43 <gncbot> andrewsw: The operation succeeded.
02:51:36 <andrewsw> @tell warlord I don't see how, but somehow 16690 definitely breaks it. remove either of the two changed lines, and it works. leave both in, it fails. also make check fails later in sx, btw.
02:51:36 <gncbot> andrewsw: The operation succeeded.
03:12:48 *** andrewsw is now known as andrewsw-afk
04:33:07 *** litsunglin has joined #gnucash
04:33:09 *** litsunglin has left #gnucash
05:05:50 *** basanta has quit IRC
05:21:23 *** basanta has joined #gnucash
06:04:03 *** basanta has quit IRC
06:33:33 *** kling0n has joined #gnucash
08:03:41 *** tjanssen has joined #gnucash
08:11:06 *** twunder has joined #gnucash
08:26:37 *** warlord-afk is now known as warlord
08:26:46 <warlord> .
08:26:46 <gncbot> warlord: Sent 6 hours and 36 minutes ago: <andrewsw> r16690 breaks make check. or at least reverting it allows it to pass. But why does reverting 16619 then allow it to pass again?
08:26:47 <gncbot> warlord: Sent 6 hours and 35 minutes ago: <andrewsw> in the process I discovered that src/engine/iso-currencies-to-c is not being run during make, but that doesn't seem to be the source of the problem. I ran it manually and it still failed make check.
08:26:48 <gncbot> warlord: Sent 5 hours and 53 minutes ago: <andrewsw> ignore comments about iso-currencies-to-c
08:26:49 <gncbot> warlord: Sent 5 hours and 35 minutes ago: <andrewsw> I don't see how, but somehow 16690 definitely breaks it. remove either of the two changed lines, and it works. leave both in, it fails. also make check fails later in sx, btw.
08:58:52 <wharp> All that is so far over my head.
09:05:44 <warlord> that's okay.
09:05:47 <warlord> it's not over mine.
09:06:11 <wharp> Oh yeah, I realize that.
09:06:20 <warlord> Adding the currency changes the test, and then it winds up picking different currencies/values..
09:06:26 <warlord> So I think it's a bug in the test.
09:06:26 <wharp> You mentioned something about git, does gnucash use both svn and git?
09:06:30 <warlord> No.
09:06:39 <warlord> Some developers use git-svn
09:06:42 <wharp> ah, ok
09:20:12 <wharp> What's the diff b/n the branches and the trunk?
09:29:51 *** ural0 has joined #gnucash
09:31:12 <warlord> Ummmm....
09:31:27 <warlord> Different "lines" of code.
09:31:41 <wharp> Is there a branch for each release?
09:31:49 <warlord> Yes
09:31:50 <wharp> I just need to google and figure it all out I guess
09:31:54 <warlord> Yeah..
09:32:00 <warlord> This is like SCM 101
09:32:39 <wharp> Let's just say what I did take in school was less than educational.
09:32:46 <wharp> I don't have a programming background though.
09:33:02 *** twunder has quit IRC
09:33:02 <warlord> technically, neither do i ;)
09:33:09 <warlord> (I was EE, not CS)
09:33:27 <warlord> ...but none of my friends remember (or even thought) that
09:34:29 <warlord> OOOH! I made it up to 19!
09:35:38 *** ural0 has left #gnucash
09:36:53 <wharp> lol
09:37:17 <wharp> I was CIS. I had a javascript class where the prof talked about web economics the whole quarter and a visual basic class that was decne.t
09:37:25 <wharp> Oh, and I think I had COBOL too.
09:37:47 <warlord> COBOL?
09:38:01 <wharp> yep
09:38:08 <warlord> I took cobol during a summer back in the early 80s.
09:38:15 <warlord> I didn't think anyone still taught it.
09:38:20 <warlord> Anyways, I need to leave..
09:38:23 <wharp> Well I went to college in 98-'01
09:38:26 <wharp> K
09:38:29 <wharp> Merry Christmas
09:38:34 <warlord> Yeah, you're a young-un ;)
09:38:38 <wharp> yeah
09:38:47 <warlord> Andyou got a COBOL class?
09:38:49 <warlord> WHERE did you go?
09:39:10 <wharp> my young-un sure makes me feel old
09:39:13 <wharp> Louisiana Tech
09:39:23 <warlord> Anyways, I gotta go.
09:39:28 <wharp> take care
09:39:44 <warlord> @tell andrewsw I added to bug #492137
09:39:44 <gncbot> warlord: The operation succeeded.
09:39:48 *** warlord is now known as warlord-afk
10:15:10 *** wizkid238 has joined #gnucash
10:26:57 <wharp> lol
10:28:15 *** kling0n has quit IRC
10:28:52 *** kling0n has joined #gnucash
10:45:28 *** kielein has joined #gnucash
10:58:27 *** kling0n has quit IRC
11:07:37 *** twunder has joined #gnucash
11:17:53 *** twunder has quit IRC
11:47:16 *** sjc has joined #gnucash
12:39:42 *** pdc__ has quit IRC
12:42:23 *** pdc has joined #gnucash
12:58:19 *** nbinont-afk has quit IRC
13:02:21 *** sjc_ has joined #gnucash
13:11:11 *** sjc has quit IRC
13:18:52 *** sjc__ has joined #gnucash
13:19:22 *** warlord-afk is now known as warlord
13:27:30 *** sjc_ has quit IRC
13:31:24 *** sjc_ has joined #gnucash
13:40:31 *** sjc__ has quit IRC
13:52:32 *** sjc_ has quit IRC
14:02:26 <warlord> andrewsw-afk: I have verified that r16690 introduces the problem.. Which leads me to believe it's actually a bug in the tests.
14:07:04 *** Pepe66 has quit IRC
14:16:45 *** twunder has joined #gnucash
14:31:31 *** warlord is now known as warlord-afk
14:40:35 *** twunder has quit IRC
15:09:10 *** warlord-afk is now known as warlord
15:47:53 *** tjanssen has quit IRC
16:18:26 <warlord> andrewsw-afk: fixed in r16716.
16:18:40 <warlord> @tell andrewsw fixed in r16716. It was a bug in the tests.
16:18:40 <gncbot> warlord: The operation succeeded.
16:20:14 <warlord> Now lets see if people play with the book closing code I wrote.
16:43:11 *** conrad has quit IRC
17:01:35 <warlord> time to head out.
17:03:21 *** warlord has quit IRC
17:25:16 *** pdc_ has joined #gnucash
17:27:53 *** pdc has quit IRC
17:34:07 *** warlord has joined #gnucash
17:34:07 *** gncbot sets mode: +o warlord
17:35:22 *** dan_ has joined #gnucash
17:36:14 <dan_> Marry Christmas, everyone
17:37:17 <dan_> My today's woe though is: "There is no documentation available for usr/share/gnucash/gnome/help/gnucash/C/gnucash-guide.xml"
17:37:54 <dan_> This msg shows up upon calling Gnucash online help
17:38:30 <dan_> Funny thing is that I did install documentation along with the gnucash itself
17:38:32 <warlord> dan_: it means you dont have the gnucahs-docs package instaled
17:38:40 <warlord> Or you're on SuSE
17:39:04 <dan_> Yep, I am on OpenSuSe 10.3 indeed. Is it a prob?
17:39:09 <warlord> Yep.
17:39:16 <warlord> Search the list archives.
17:39:20 <warlord> (or maybe even the wiki)
17:39:43 <dan_> k
17:40:02 <dan_> Could u please point me to the right direction?
17:40:33 <dan_> How can i get rid of this issue cause I do need gnucash before the New Year
17:41:12 <dan_> YaST shows gnucash-docs as installed
17:41:22 <warlord> Did you search the list archives?
17:41:55 <dan_> R u talking about gnucash email list archives?
17:42:08 <warlord> Say what?
17:42:22 <warlord> Russian Talking about gnucash.. what?
17:43:46 <dan_> Warlord, are you talking about checking gnucash email list archives or some other ones, please?
17:43:53 <warlord> Yes
17:43:55 <warlord> Of course.
17:44:02 <warlord> What other list archives would I be talking about?
17:44:08 <dan_> k
17:45:57 <warlord> Try https://lists.gnucash.org/pipermail/gnucash-user/2007-March/019914.html
17:47:20 <warlord> Gotta run.
17:47:22 *** warlord is now known as warlord-afk
17:48:00 <dan_> thank you, warlord
17:48:07 <dan_> do appreciate it, man
18:08:44 *** dan_ has quit IRC
18:17:57 *** nbinont has joined #gnucash
18:18:11 *** nbinont is now known as nbinont-afk
18:32:35 *** nbinont-afk has quit IRC
19:07:57 *** kielein has quit IRC
22:01:34 *** nbinont has joined #gnucash
22:01:42 *** nbinont is now known as nbinont-afk
22:09:31 *** warlord-afk is now known as warlord
22:13:25 *** nbinont-afk is now known as nbinont
22:13:34 <nbinont> hi warlord
22:13:53 <warlord> Hi
22:14:51 <nbinont> quick question...I'm wondering which of the make *checks we really use. (there appears to be quite a number autogenerated)
22:14:56 <nbinont> would you know?
22:15:51 <warlord> many of them are related.
22:16:03 <warlord> There's "make check" which run all the tests.
22:16:36 <warlord> then there's "make distcheck" which is a combination of "make dist" and "make check" where it builds a disk tarball, then untar's, builds it, and runs "make check" on the resulting build..
22:20:38 <nbinont> ok, thanks. I guess I can just run "make check" and be happy :)
22:20:56 <warlord> Yep.
22:21:03 <warlord> That's really the only thing you should care about.
22:21:32 <nbinont> cool
22:23:38 *** nbinont is now known as nbinont-afk
22:32:52 <warlord> nbinont-afk: when you get the chance you should try my book closing feature ;)
22:52:54 *** Zoolooc__ has joined #gnucash
22:58:58 *** twunder has joined #gnucash
23:01:13 *** tjanssen has joined #gnucash
23:02:12 *** Zoolooc_ has quit IRC
23:14:45 *** andrewsw-afk is now known as andrewsw
23:15:03 <andrewsw> hi. just a quick hello and have a happy holiday!
23:15:03 <gncbot> andrewsw: Sent 13 hours and 35 minutes ago: <warlord> I added to bug #492137
23:15:04 <gncbot> andrewsw: Sent 6 hours and 56 minutes ago: <warlord> fixed in r16716. It was a bug in the tests.
23:15:19 <andrewsw> :) cool.
23:15:21 <warlord> :)
23:15:41 <andrewsw> I'm likely not around tomorrow at all, so have a really nice day!
23:15:53 <warlord> thanks. happy holidays to you and yours
23:16:15 <andrewsw> :) goodnight and listen for the bells on the roof...
23:16:18 *** andrewsw is now known as andrewsw-afk
23:17:02 <warlord> good night
23:20:01 *** tjanssen has quit IRC
23:21:05 *** warlord is now known as warlord-afk