/srv/irclogs.ubuntu.com/2015/12/31/#ubuntu-devel.txt

LocutusOfBorg1I wish you a nice new year!18:29
tewardstupid question, but is there a way to refer to the version string used in the Debian changelog during building in a rules file?  (such that a variable could extract the value of whatever the version string was on the changelog entry)19:14
tumbleweedteward: dpkg-parsechangelog19:15
tewardtumbleweed: thank you!19:16
tewardtumbleweed: is that able to be used during the actual building of the binaries in the rules file to extract the version for use in a later flag passed to 'go' for building?  From what testing I'm doing, I'm seeing nothing in that variable (I have the rules file echoing the output to the log during build, and it's showing nothign...(20:10
tumbleweedteward: it's a command, not a variable20:12
tewardtumbleweed: i know this20:13
tewardtumbleweed: trying to put it INTO a variable20:14
tewardgetting no output20:14
tumbleweedif you're in a source package, it should have output20:14
tewardI.E. in makefile: export PACKAGE_VER_STRING=$(dpkg-parsechangelog --show-field Version)20:14
tewardthat to then later be referred to in arguments when the actual building of the binaries happens in sbuild20:15
tumbleweed--show-field is fairly recent. Is it available in the relase that you're building on?20:15
tumbleweedwe used to have to pipe dpkg-parsechangelog to sed20:15
tewardtumbleweed: building in both Debian stable (jessie) and Wily, both in sbuild schroots20:15
tewardso it *should* be available in there?20:15
tumbleweedanything since dpkg 1.17.020:16
tumbleweedaccording to the manpage20:16
tewardyeah that's what I read too, for some reason it just does *not* want to do what i want it to :/20:17
* teward tests something20:18
tewardtumbleweed: only thing I can think of is that when we get to the dh_auto_install step which is overridden in the rules, it's no longer able to use dpkg-parsechangelog20:23
tewardbut I can't for the love of me figure out why20:23
tewardgoing into the schroot, it *LOOKS* like it would work, but it's just not getting output how it should and providing it later on (i.e. empty string)20:27
tewardwhich breaks things :/20:27
tumbleweedteward: show me your rules file20:29
tewardtumbleweed: http://paste.ubuntu.com/14340616/20:34
tewardoops20:34
tewardwait20:34
tewardthat's the older one with git lol20:34
tumbleweedI can tell you what your problem there is :P20:34
tumbleweedeach line of a rule gets executed in a seprate subshell20:34
tumbleweedif you want to export a variable for every subshell, you mustn't do it from a rule20:35
tewardtumbleweed: so it should go out where GOPATH is defined, then?20:35
teward(in your mind replace that `git describe` line with the dpkg-parsechangelog stuff, i'm too lazy to repaste xD)20:35
tumbleweedyep20:37
tumbleweedand you probably don't need to export it - you can use a make variable, rather than a shell variable, for this20:37
tewardwill that get picked up later with variable substitution when it passes it to go in the flags?20:38
tumbleweedif you use the make syntax $(VARIABLE_NAME)20:38
tewardtumbleweed: no dice :/20:44
tewardno idea why it's still failing :?20:44
tewardand I have an Echo line to echo what's there too, which is how I know it's failing20:44
tewardabout fed up with it right now and am just gonna hardcode the damn thing for now into place >.<20:45
tewardTHEN tackle variable sub20:45
cjwatsonif you're setting it as a make variable, remember to use $(shell ...) not $(...)20:48
tewardtumbleweed: cjwatson: I owe you both a beer/drink!21:00
tewardtumbleweed: thank you for the guidance on where to get the string set21:00
tewardcjwatson: thanks for reminding me about needing 'shell' there21:00
tewardit's now behaving as it needs to :)21:00
tumbleweedexcellent :)21:00

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!