[15:11] <attente> https://www.irccloud.com/pastebin/XrJE3Xpc/
[15:11] <attente> alan_g: hey, are you getting those errors when building miral-shell? ^
[15:13] <alan_g> attente: no. Compiler? Platform?
[15:15] <attente> alan_g: gcc via ccache
[15:15] <attente> on zesty
[15:16] <attente> gcc 4:6.2.1-1ubuntu1, ccache 3.3.3-1
[15:21]  * greyback also got that today
[15:27]  * alan_g hasn't got zesty to hand
[15:31] <attente> greyback, alan_g: adding an extra comma there fixes it
[15:32] <greyback> attente: alan_g: ack I'll look after it
[15:32] <attente> greyback: thanks
[15:32] <alan_g> thanks
[15:48] <alan_g> greyback: that's down to a gtest change?
[16:03] <greyback> attente: where did you add a comma? To designate an extra macro argument?
[16:07] <attente> greyback: yeah: )); -> ),);
[16:07] <attente> greyback: in all three places the error occurred
[16:08] <greyback> attente: bah, I thought I had tried that and compiler rejected it
[16:08] <attente> https://www.irccloud.com/pastebin/LyZvkGRM/
[16:08] <attente> oh. really?
[16:09]  * attente shakes fist at googletest
[16:16] <greyback> alan_g: well, this is the patch. I'm unclear what's really changed: http://bazaar.launchpad.net/~gerboland/miral/gtest-bug-workaround/revision/457
[16:16] <ubot5`> Ubuntu bug 457 in k3b (Ubuntu) "k3b doesn't have any dependencies on cdrdao" [Medium,Invalid]
[16:17] <greyback> ooh ubot got that wrong
[16:18] <alan_g> greyback: for earlier versions of googletest (vivid, xenial, yakkety) that will break. "error: macro "INSTANTIATE_TEST_CASE_P" passed 4 arguments, but takes just 3),);
[16:18] <greyback> alan_g: yeah I guessed that
[16:19] <greyback> but I'm still unclear what really changed
[16:19] <greyback> qtmir uses that macro in one test, and it compiles.
[16:23] <alan_g> greyback: could you paste the definition of INSTANTIATE_TEST_CASE_P?
[16:25] <alan_g> I see: # define INSTANTIATE_TEST_CASE_P(prefix, test_case_name, generator) \ etc
[16:27] <greyback> alan_g: yep it has changed since: http://pastebin.ubuntu.com/23549170/
[16:28] <alan_g> /sigh
[16:33] <alan_g> Why do they do these things? AFAICS qtmir doesn't use "-pedantic" which would be the difference
[16:34] <alan_g> greyback: you still working on fix? or shall I take over?
[16:35] <greyback> alan_g: I'm still investigating
[16:35] <greyback> just comparing qtmir testcase with miral: http://pastebin.ubuntu.com/23549194/
[16:35] <greyback> difference is not obvious to me yet
[16:36] <greyback> no obviously different compiler flag
[16:36] <alan_g> I didn't see -pedantic in qtmir
[16:36] <greyback> ah
[16:37] <greyback> I didn't think pedantic was reported with -Werror
[16:37] <greyback> ok, that must be it
[16:54] <greyback> alan_g: wdyt? http://pastebin.ubuntu.com/23549271/
[16:54] <greyback> I can't see any way to get the googletest version, to gate it on 1.8
[16:56] <alan_g> I'd prefer not to rely on a gcc extension: https://gcc.gnu.org/onlinedocs/gcc-4.9.2/gcc/Variadic-Macros.html
[16:57] <alan_g> And I remember the  GTest .cmake doesn't set the standard properties
[17:02] <alan_g> greyback: I'll sleep on it.
[17:02] <greyback> alan_g: ack
[17:02] <greyback> have a nice evening
[17:02]  * greyback not sure what is least worst fix