[00:17:42] plus it also doesnt seem to keep hanging up at exactly the same point [00:18:23] if you are a bit free tomorrow, would really like a second person to look on what i am doing [01:44:34] i might have some time tomorrow yeah [01:44:58] ilovekiruna: do you have that command and a link to the build? it'll help me [01:45:10] I can show you many builds [01:45:28] KaZeR: I just cancelled: https://circleci.com/gh/hoehnp/navit/540 [01:46:13] I tried to replicate @metalstrolch's script with .spec file in the script folder, but it still just seems to hang, cancelled the job then [01:48:21] another job was https://circleci.com/gh/hoehnp/navit/533 [01:50:12] the last successful was https://circleci.com/gh/hoehnp/navit/489 [01:50:36] my guess is that somewhere some pathes are set wrong [01:51:47] are you using make -j when running manually? [01:51:57] nope [01:52:01] just make? [01:52:04] I was also just running the script [01:53:30] and manually it works without hanging? [01:53:36] japp [01:53:41] looks like it hangs because it waits for a user input somewhere [01:53:53] but it hangs while compiling [01:54:13] cant imagine that any user input could be required then [01:55:07] just noticed one potential tiny difference. I think i usually manually call the script directly, not with "bash script.sh" [01:56:49] #! /bin/sh [01:57:02] so you might want to try that in ci too :) [01:57:05] afk 30 min [01:57:58] KaZeR: taht I even did [01:58:03] worked perfectly! [02:08:11] unlike when it starts in the container it is even actually pretty fast [02:08:44] just started the build via ssh manually via export VERSION_ID=2.2.1.18 && ./build_sailfish.sh from the script folder [02:25:12] will turn in now, quite frustrating, that the only way that i works is that nasty way of patching via sed [02:25:29] KaZeR: if you have any idea, please write me, will ehck tomorrow [02:27:47] as it really kills too much of my time, I would really suggest to at first use that ugly hack [06:57:32] *** Joins: xenos1984 (~xenos1984@193.185.150.2) [06:57:32] *** ChanServ sets mode: +v xenos1984 [07:03:51] *** Quits: xenos1984 (~xenos1984@193.185.150.2) (Quit: Leaving.) [07:24:03] ilovekiruna: worked perfectly, so the issue is fixed? [07:24:21] we'll figure out something to avoid the sed patching. We can move most of these stuff into a real docker file [07:52:32] *** Quits: aerostitch (~aerostitc@136.25.161.62) (Ping timeout: 244 seconds) [08:12:48] *** Joins: aerostitch (~aerostitc@c-24-23-228-135.hsd1.ca.comcast.net) [08:12:48] *** ChanServ sets mode: +v aerostitch [08:24:40] *** Joins: naggety (~naggety@112.red-95-125-19.dynamicip.rima-tde.net) [08:29:08] *** Joins: xenos1984 (~xenos1984@dyn231-156.eduroam.net.utu.fi) [08:29:08] *** ChanServ sets mode: +v xenos1984 [10:42:32] *** Quits: xenos1984 (~xenos1984@dyn231-156.eduroam.net.utu.fi) (Quit: Leaving.) [11:35:21] KaZeR: only the ugly sed hack works so far [13:30:04] Hi. I see that in navit.xml vehicleprofile item you can set attributes such as vehicle_width, vehicle_length, etc. How is this data used? [13:56:37] Also, how can I use my own variables in navit.xml. Specifically, I want to select the language according to a variable, which can be an environment variable or any other. Also for other cases other possible options than the language. [14:37:42] *** Quits: naggety (~naggety@112.red-95-125-19.dynamicip.rima-tde.net) (Ping timeout: 264 seconds) [14:38:20] *** Joins: naggety (~naggety@82.red-81-47-26.dynamicip.rima-tde.net) [14:41:33] Seems that logger is not working [14:51:00] KaZeR: what about later considering MaryTTS to replace espeak? http://mary.dfki.de/ I know it is java based :-(, but sounds quite nice to me [14:57:55] ilovekiruna: when selecting a speech synthesizer I would consider mainly languages support. I was testing various programs and only Festival had a quite good spanish voice support, for example. [14:58:22] Also mbrola project supported it quite well, but I couldn't use it because it had a non-comercial license [15:02:43] I still thing festival sounds quite robotic [18:10:45] > naggety | Hi. I see that in navit.xml vehicleprofile item you can set attributes such as vehicle_width, vehicle_length, etc. How is this data used? [18:11:08] mostly for a truck specific routing profile. the height is used to ensure the vehicle fits under bridges and tunnels [18:11:57] naggety | Also, how can I use my own variables in navit.xml [18:12:19] i guess that you want a property on the navit object for example? [18:17:37] KaZeR: i start to wonder if i am stupid or circleci is not one hundred percent reproducable [18:17:56] haha [18:18:09] why do you laugh? [18:19:13] feels pretty mean to me, after fighting almost 24 hours getting my stupid script running in a relyable way [18:21:34] sorry [18:21:50] i laughed because you're obviously not stupid :) [18:22:04] let me point naggety to some code and we can look into it [18:22:16] naggety: first you need to define a config entry, like this one : https://github.com/navit-gps/navit/blob/daee89c2cd2ddfd0965fc597b1d15b3eb45af9b7/navit/attr_def.h#L397 [18:23:24] then you can access the config item like this : https://github.com/navit-gps/navit/blob/daee89c2cd2ddfd0965fc597b1d15b3eb45af9b7/navit/audio/player-spotify/spotify.c#L425-L429 [18:23:56] we usualy do it in a given plugin's _new() method as this is where the struct attr **attrs is available [18:24:11] you can then copy the value into a private struct like it's done here or just save a pointer to it [18:24:25] ilovekiruna: i have some time. point me to your latest error [18:25:05] KaZeR: if you see this job: https://circleci.com/gh/hoehnp/navit/649 [18:25:16] apparently the build script stops in between [18:25:20] I have seen this many times [18:25:55] for easier analysis I would suggest to go back to a run that was successful [18:26:42] this one here worked as it should: https://circleci.com/gh/hoehnp/navit/594 [18:27:24] just the next change broke it where i was changing the name of the build script: https://circleci.com/gh/hoehnp/navit/604 [18:27:57] I tried to reproduce now all the changes from that time again, but no success [18:28:42] what also bugs me is that sometimes build_android fails despite i didnt change anything about it [18:30:57] for example android failed in https://circleci.com/gh/hoehnp/navit/646, now when just changing the shebang in the build script it runs successful again [18:31:03] so totally clueless [18:38:04] KaZeR: any ideas? [19:07:34] naggety | Seems that logger is not working [19:07:55] yeah and Number6 is on a plane today. Or recovering from jetlag. He should probably be able to fix it soon [19:08:38] ilovekiruna: can you give me write permissions on your repo? Otherwise i can't interact much with your ci jobs [19:14:07] just a second, will grant u now [19:15:46] sne you an invite [19:21:08] ok [19:21:10] thx [19:28:10] unrelated but stumbled on it while looking at your issue : xslt transformations for android : saxonb-xslt -s:navit_shipped.xml -xsl:xslt/android.xslt -o:navitldpi.xml OSD_SIZE=1 ICON_SMALL=24 ICON_MEDIUM=32 ICON_BIG=48 [19:28:39] what is up with this xslt? [19:35:53] the cool thing about this example is that it shows you how to use parameters [19:36:22] <- define a default value [19:36:33] and then in [19:36:40] [19:36:50] set the value of this attribute to the one that was given as parameter [19:38:33] u talk there again about the changes for win32 speech, right? [19:39:44] yes. i'm looking into your sailfish issue, but stumbled upon this example while reading your config.yml [19:39:53] i think that albeit a bit complex it's a good example [19:39:53] ah, I c [19:39:59] yeah [19:40:19] I made now one modification to my build script, removed the actual building in the last commit [19:40:25] wanted to check it manually again [19:40:41] right now i'm testing locally. need to install sb2 it seems [19:40:57] damn used the wrong base image [19:40:58] if you pull my container it should be there [19:41:01] yeah [19:45:55] mmm i'm missing something [19:45:56] sb2: Error: Invalid target specified, aborting. [19:46:06] i've been following the steps from your config.yml [19:46:19] I also sometimes had this [19:46:27] have you defined VERSION_ID? [19:46:46] yep, 2.2.1.18 [19:47:42] am testing right now on the circleci, seems like this time it stops after 8 percent [19:47:49] "sb2 -t SailfishOS-2.2.1.18-armv7hl [...]" so yeah the version id is defined correctly [19:48:04] one question [19:48:16] are you sure you dont use the base image? [19:49:05] i am using hoehnp/sailfishos-platform-sdk:2.2.1.18 [19:49:16] ok, lete me try once myself [19:50:01] yes, this time building seemd to have stopped at 8 percent [19:50:42] how does your /srv/mer/targets/ folder look like? [19:52:29] # ls /srv/mer/targets/ [19:52:31] SailfishOS-2.2.1.18-armv7hl SailfishOS-2.2.1.18-i486 [19:52:42] ok, how it is supposed to be [19:53:16] first or second sb2 line? [19:57:52] both [19:58:26] very strange [19:58:38] for me it runs totally smoothly [19:58:43] https://pastebin.com/bSi9GQkq [19:59:24] are you running mannually or by calling the script? [20:00:23] calling the script yeah [20:00:45] ok, will start a new container and try again [20:00:54] u call it with bash -x ,right? [20:02:03] yep [20:02:43] btw : when this works we might want to break it down into 2 jobs, one for arm and one for intel. should build a bit faster. but no big deal for now [20:03:15] honestly, I am still not yet convinced if we really should build for intel [20:03:26] ha, that's another debate then :) [20:03:29] dont know which devices actually use it [20:03:48] pretty much like android-x86 i guess [20:04:23] ok, am now in contrib/sailfish [20:06:21] now was running the same command as you bash -x build_sailfish_ci.sh [20:09:32] # sb2-config -l [20:09:34] /root/.scratchbox2 missing, create some targets with sb2-init! [20:09:36] could this be it? [20:09:53] I will send you my commands in a pastebin [20:09:55] am just on it [20:10:40] pm is fine [20:47:53] *** Joins: xenos1984 (~xenos1984@193.185.150.2) [20:47:53] *** ChanServ sets mode: +v xenos1984 [21:19:45] *** Joins: mvglasow (~mvglasow@dslb-188-104-126-007.188.104.pools.vodafone-ip.de) [21:23:43] hi folks [21:24:47] KaZeR or anyone else who has access to NavitLogger: seems the logger died around 01:30-01:40 on October 3, see https://irclogs.navit.ie/%23navit-2018-10-03.log [21:25:32] and logs for the following two days are completely empty [21:25:50] and none of what I've written here has shown up in the logs so far... [21:57:13] *** Quits: xenos1984 (~xenos1984@193.185.150.2) (Ping timeout: 246 seconds) [22:01:59] thanks mvglasow [22:02:21] i think that Number6 is the only one currently, we might have to wait for him to be back from his trip, probably this weekend [22:03:45] all right [22:04:15] in the meantime, there's news on the traffic side [22:04:27] I've prepared a first PR [22:04:57] with routing now completely moved to LPA*, paving the way for other improvements [22:06:09] nice! [22:07:24] traffic is still opt-in (i.e. you have to configure it manually to use it), hence odds of traffic breaking anything are low [22:08:15] though CodeFactor is still somewhat naggy, is passing all of its checks a requirement for merging? [22:09:05] I tried my best, but some things are quite non-trivial to get rid of, and some rules may still need tweaking imho [22:14:57] i think that it's totally open for discussion [22:16:47] phew :-) place for discussion: https://github.com/navit-gps/navit/issues/673 [22:20:04] and since you asked about how you could help some time ago: feel free to add a traffic plugin for your preferred platform :-) [22:20:53] since I rely exclusively on Android for production use, the Android plugin scratches my itch quite well [22:21:23] the same principle should be portable to Linux, using Dbus to broadcast traffic feeds [22:22:45] for simplicity, I would recommend sticking with TraFF as a message exchange format and doing all interfacing with the sources in external apps [22:23:10] the TraFF spec is at http://traffxml.org [22:24:22] also, I'm in touch with Gabriel, who's working on an OpenTraffic succesor [22:25:17] still trying to figure out if one of their various projects produces the data we need, and how far the project has advanced [22:26:48] mvglasow: sounds great [22:40:01] ilovekiruna thanks, and contributions are always welcome :-) [22:42:57] mvglasow: me and KaZeR are right now on my current contribution in building for sailifhs [22:44:35] all right, I know that feeling when you could assign four clones of yourself on four different projects ;-) [23:47:28] *** Quits: mvglasow (~mvglasow@dslb-188-104-126-007.188.104.pools.vodafone-ip.de) (Quit: Leaving)