12:26 <stenur> nghttp2 is "long" on .47 btw

12:26 <stenur> aka libnghttp2 as it is now called

12:30 <stenur> maybe not that important though, in the last months (before i stopped tracking) they mostly cared for their webserver and HTTP/3; from the linked things i see a bug that could affect us

13:10 <stenur> Btw having a libxml2 .MAJOR.MINOR symlink would be tremendous, i had to rebuild it all (and clang still missing) as we only have .MAJOR.MINOR.UPDATE and the normal .so is never kept as such by linking (do not ask me no questions.

14:03 <jue> stenur: I agree, we should have a "libxml2.so.2" symlink, looks like we lost that while swiching to the cmake build

14:18 <farkuhar> beerman: the upower build error seems to arise whenever libxslt is installed, even with the meson option -D gtk_doc=false. Your lxc containers never had libxslt (especially after running pkg-clean), so you never encountered the errors that SiFuh and jaeger reported.

14:20 <beerman> farkuhar: I can't check right now but I am pretty certain I have libxslt installed on my machines

14:21 <farkuhar> I managed to build upower successfully, after removing libxslt (despite the 29 packages that depend on it).

14:21 <beerman> i can run a quick test in my container

14:25 <farkuhar> it's a peculiar combination that creates this error. I think it needs libxslt installed, but gtk-doc NOT installed, and then the meson setup fails.

14:26 <beerman> http://sprunge.us/6OIjW6

14:27 <beerman> rip me..

14:27 <beerman> http://sprunge.us/7YQIXP

14:28 <beerman> i can reproduce the build error that way. alternative a) depend on gtk-doc (nobody likes that) b) hack around it (no maintainer likes that)

14:29 <farkuhar> alternative a) is what I proposed to Romster, after first reproducing SiFuh's error last weekend. I can see why it's not a popular choice.

14:30 <beerman> no matter what you do you can't please everybody (and for me, I am done with that anyway). Alternative c) disable manpages as well btw ;)

14:30 <groovy2shoes> alternative d) screw meson ¬_¬

14:31 <beerman> alternative e) if you are not contributing stop posting in this channel?

14:31 <beerman> all you ever do groovy2shoes is hate around against things.

14:32 <farkuhar> Among the stale files in the old source array (before beerman's cleanup) was a manpage. Maybe that was included in the ports tree because alternative c) was considered too drastic.

14:32 <beerman> The only thing too drastic in it that I can see is that people can't get over the fact that the world keeps spinning, with or without them

21:49 <stenur> btw on FreeBSD i saw the commit "zstd: build with C11 [.] This enables thread-aware timers which in turn fixes benchmark result reports."

21:49 <stenur> (aka -std=c11)

21:50 <stenur> (dunno (yet) whether it matters for Linux, a short git grep in the code made me sceptical though)