[00:09:26] *** Quits: Horwitz (~mich1@p200300EC9BC1EC00022268FFFE64E7C4.dip0.t-ipconnect.de) (Ping timeout: 246 seconds) [00:22:25] *** Joins: Horwitz (~mich1@p200300EC9BC23000022268FFFE64E7C4.dip0.t-ipconnect.de) [00:22:25] *** ChanServ sets mode: +o Horwitz [03:49:31] *** Joins: xenos1984 (~xenos1984@25fe-8d78-cd62-0550-d480-87c4-07d0-2001.dyn.estpak.ee) [03:49:31] *** ChanServ sets mode: +v xenos1984 [04:07:57] *** Quits: xenos1984 (~xenos1984@25fe-8d78-cd62-0550-d480-87c4-07d0-2001.dyn.estpak.ee) (Ping timeout: 250 seconds) [04:45:04] *** Joins: xenos1984 (~xenos1984@6df4-e61d-bd92-f40c-d480-87c4-07d0-2001.dyn.estpak.ee) [04:45:04] *** ChanServ sets mode: +v xenos1984 [05:34:51] Hi guys, I've got a silly question: in many files I see #include "config.h". Given the fact that it has double quotes around I guess it's referring to a file that is in the same directory as the header or C file that includes it but I can't find a file named like that in the project [05:35:03] I see a file named config_.h [05:35:58] Do you have any pointer on where this file could actually come from please? [05:36:51] Nvm I found a config.h.in at the root of navit. I guess it's where it comes from [05:50:52] Config.h is auto generated it contains the build configuration once generated [05:51:37] @aerostitch do you have some time this week to do the rc1 with me? I would like to do it with you as guide [06:20:36] Sure [06:20:57] But I'm wondering if we should stick to the RC* format or just release more often [06:21:04] and do patch releases instead [06:22:14] Yes, already planned, 0.5.5 milestone already created. But for 0.5.4 I want to make things consistent like before, then we can release more often [06:22:42] I don't think we created any rc past 5.0 [06:22:48] but alright [06:23:17] Btw I'm currently trying out cunit for adding a unit test example in our code and it seems that the way we build currently makes it quite a nightmare to do anything simple 😦 [06:24:00] No, like last time, 0.5.2 had a bug, next should have been 0.5.2.1 because it was just a small detail but not its like it is [06:24:38] Unit testing in navit will take a long time to achieve, yes [06:25:01] Exactly for the plugin system stuff, gui, graphics and so on... [06:26:30] Some time ago I already thought that we would really need to split navit into pieces like core, each gui, each graphic, provide a stable api and so on, but who has time to do so? ^^ [06:46:41] Basically I have to do this to be able to test 1 function: https://gist.github.com/aerostitch/faa106d7534416c4dd9dc7e57dbfd209 [06:46:57] And I have to get all the rest out if I want it to be easy :\ [06:54:17] *** Quits: ilovekiruna (~ilovekiru@ip1f12a9be.dynamic.kabel-deutschland.de) (Ping timeout: 240 seconds) [06:57:57] What rest do you mean? The rest of the functions? [06:58:34] Yeah the rest of the content of those files [07:07:02] Interesting, with the unit tests I found a corner case not properly handled by the search_fix_space function: [07:07:46] When the string only contains spaces or comma or slash the resulting string will just be 2 spaces [07:08:43] not exactly 2 spaces. If it has 4 spaces it will get out as 4 spaces [07:11:44] But why are the other functions a problem? They don't interfere with each other. We are in c so it should also be easy to fake the class like implemented things by giving them a fake initialized object [07:27:24] *** Quits: xenos1984 (~xenos1984@6df4-e61d-bd92-f40c-d480-87c4-07d0-2001.dyn.estpak.ee) (Quit: Leaving.) [07:45:52] Oh no, it's more that I need to compile a shit ton of objects to be able to test a single function because we depend on a ton of headers [07:56:13] So only a time constraint? [07:59:48] well, if you look at the make file you can see that it needs 3 lines to be able to build the thing: https://gist.github.com/aerostitch/faa106d7534416c4dd9dc7e57dbfd209#file-makefile [08:00:37] but then if you start to add the dependency of search.c you need to add the dependencies for all these includes: [08:00:39] #include "debug.h" #include "projection.h" #include "item.h" #include "xmlconfig.h" #include "map.h" #include "mapset.h" #include "coord.h" #include "transform.h" #include "search.h" #include "country.h" #include "linguistics.h" #include "geom.h" #include "util.h" #include "search_houseno_interpol.h" #ifdef HAVE_API_ANDROID #include "android.h" #endif #include "layout.h" [08:01:00] and then add look inside of each of these includes to find the dependencies [08:01:23] and you end up with over 30 objects to compile just to do 1 test [08:01:41] Yes... But when integrated into cmake this problem didn't exist any more, just include search.c (not h) [08:02:26] well then that means that I would have to find the way to make it work with CMake [08:02:46] which seems to be not 100% trivial to me 🙂 [08:06:10] From this post it sounds quite easy http://nuno-sousa-corner.blogspot.com/2009/10/back-to-c-cmake-and-cunit.html?m=1 [08:06:54] Given the discussions I had in the past about unit tests I was more looking for a way to have, let's say something where you put a file, run your tests, put it somewhere outside to avoid bothering other people [08:09:13] The link looks interesting but I'm not sure it's going to be that easy to be honest. 🙂 Based on the tests I was doing there's at least a bunch of flags missing here for glib for example [08:09:29] You can already see in your one test [08:09:30] Don't put it outside, then it will become out of sync [08:09:32] Unit testing is annoying, but when done properly it will help [08:10:33] Maybe we should revive the old github issue about it and see, if we want to start unit testing, what convention we would like to take [08:10:41] like file naming [08:11:43] or location (should it be in a subfolder called "tests" at the root or in a subfolder called "tests" in each directory or directly along with the actual code file?) [08:13:29] I would vote for one tests folder with files named test_plugintype_component.c so for example the full path could be navit/tests/test_gui_gtk.c [08:15:45] sounds good! 🙂 [08:16:16] I'll revive the old thread. I see that lains proposed 2 other frameworks that I've seen and that I need to test [08:17:18] I looked into it also, and found cunit to be the best fitted for our needs. At least as far I remember [08:19:57] Will you still be online in about 6-7 hours? That would be the moment we could start the rc1 because then I will be at pc again (currently only smartphone) [08:21:18] Well it's 1:21AM right now so I need to go to bed [08:21:31] and this week I'm traveling so I might be out a bunch [08:22:07] But just ping me when you're available and I'll see if I need to ask for another time [08:22:39] Around 7PM Pacific time would probably be the best tomorrow I think if you're up for it [08:22:46] But I'll see what I can do [08:25:05] Don't be to stressed because of navit, you know we had time where releases took way longer ;) [08:26:50] looooool [08:27:12] I'm not stressed out, I'm passionate! rofl [08:28:17] I haven't done a lot of technical stuffs over the last year so when I do I'm kind of "in the zone" 😄 [08:28:34] and not letting go until I have something working [08:28:47] Perfect ^^ [08:29:38] Alright! Talk to you in a few hours! 😉 [08:49:33] *** Joins: xenos1984 (~xenos1984@2001:bb8:2002:200:6651:6ff:fe53:a120) [08:49:33] *** ChanServ sets mode: +v xenos1984 [10:38:39] *** Joins: ilovekiruna (~ilovekiru@46.183.103.17) [10:38:39] *** ChanServ sets mode: +v ilovekiruna [12:41:14] *** Quits: ilovekiruna (~ilovekiru@46.183.103.17) (Ping timeout: 240 seconds) [13:55:21] *** Joins: ilovekiruna (~ilovekiru@46.183.103.8) [13:55:21] *** ChanServ sets mode: +v ilovekiruna [14:36:08] hi again [14:51:24] *** Quits: xenos1984 (~xenos1984@2001:bb8:2002:200:6651:6ff:fe53:a120) (Quit: Leaving.) [14:55:18] *** Quits: ilovekiruna (~ilovekiru@46.183.103.8) (Ping timeout: 245 seconds) [15:01:02] *** Joins: jandegr (d5db9cc5@213.219.156.197.adsl.dyn.edpnet.net) [15:01:02] *** ChanServ sets mode: +v jandegr [15:08:58] *** Joins: xenos1984 (~xenos1984@bb63-b135-13e9-4094-d480-87c4-07d0-2001.dyn.estpak.ee) [15:08:58] *** ChanServ sets mode: +v xenos1984 [16:07:01] hi [16:07:19] hi kazer [17:10:24] *** Quits: jandegr (d5db9cc5@213.219.156.197.adsl.dyn.edpnet.net) (Remote host closed the connection) [17:27:16] *** Joins: jandegr (d5db9cc5@213.219.156.197.adsl.dyn.edpnet.net) [17:27:16] *** ChanServ sets mode: +v jandegr [18:17:53] *** Quits: jandegr (d5db9cc5@213.219.156.197.adsl.dyn.edpnet.net) (Remote host closed the connection) [20:00:41] *** Joins: ilovekiruna (~ilovekiru@217-195-233-59.dsl.easynet.nl) [20:00:41] *** ChanServ sets mode: +v ilovekiruna [20:02:29] jkoan: I just checked lovely launchpad [20:02:45] I saw in German there are now 12 expressions needing review [20:02:54] is this something i should look into? [20:03:20] All of them, should be easy ;) [20:03:39] i guess so [20:04:17] actually sometimes i wished to really hear an example sentence [20:05:22] honestly with the country names I am not sure if i agree with you [20:05:42] am wondering if we shouldn't use there the real official names of the countries [20:06:03] Those are the official names [20:06:45] Taiwan is the official name? [20:07:04] isnt it republic of China [20:08:06] As I looked up, I think both are right [20:08:36] same goes for China, I believe the official name is Peoples Republic of China [20:10:58] Possible, we could recheck, but I haven't time today any more [20:11:48] probably i dont manage today, but I would recheck them and then confirm them, ok? [20:12:06] OK [21:03:13] *** Quits: xenos1984 (~xenos1984@bb63-b135-13e9-4094-d480-87c4-07d0-2001.dyn.estpak.ee) (Ping timeout: 245 seconds)