00:36 <drez> i couldn't find this in the faq, so :P ... why not musl?!

00:37 <drez> well, i guess it would be devastating at this point considering crux's age

00:37 <drez> but besides that, is there any philosophical factor involved?

00:41 <drez> and I suppose pragmatism also

00:44 <frinnst> =======> ERROR: Signature mismatch found:

00:44 <frinnst> MISMATCHglslc.1

00:44 <frinnst> =======> ERROR: Building '/usr/ports/pkg/shaderc#2021.2-1.pkg.tar.gz' failed.

00:47 <cruxbot> [contrib.git/3.6]: [notify] crypto++: 8.5.0 -> 8.6.0 security update CVE-2021-40530 plaintext recovery attack on OpenPGP ciphertexts encrypted with ElGamal. Only certain combinations of sender and receiver software are exposed. We found that GPG (via Libgcrypt) and Crypto++ are affected when acting as sender, while Go is not. https://ibm.github.io/system-security-research-updates/2021/07/20/insecurity-elgamal-pt1#faq

02:21 <cruxbot> [contrib.git/3.6]: shaderc: add regeneratd glslc man page

09:30 <cruxbot> [opt.git/3.6]: libgd: readd macros used by php

09:30 <cruxbot> [opt.git/3.6]: scite: update to 5.1.2

09:33 <cruxbot> [opt.git/3.6]: [notify] php: update to 8.0.11, includes fix for CVE-2021-21706

10:20 <cruxbot> [contrib.git/3.6]: p5-datetime-timezone: 2.47 -> 2.48

11:08 <cruxbot> [core.git/3.6]: sysvinit: update to 3.00

11:08 <cruxbot> [core.git/3.6]: tzdata: update to 2021b

14:00 <cruxbot> [opt.git/3.6]: rust: 1.54.0 -> 1.55.0

14:00 <cruxbot> [opt.git/3.6]: rust-bin: 1.54.0 -> 1.55.0

14:02 <cruxbot> [opt.git/3.6]: vulkan-validation-layers: build as shared not static

14:05 <cruxbot> [compat-32.git/3.6]: vulkan-validation-layers-32: build as shared not static

14:06 <stenur> drez: i have no idea, but ressources are limited, and not too long ago quite some software packages required patches for musl. And musl did/does not cover it all, it goes for size not feature-richessness.

14:07 <stenur> Also not for speed; though some problems are solved brilliantly as far as i can tell. (And some of those they may even have developed themselves, that is to say.)

14:08 <stenur> Also i dunno CRUX has lots of scandinavian roots, and maybe UTF-8 was an issue for some.

14:13 <stenur> And .. then. A GNU Libc based _can_ indeed be not that much larger than a musl one. But for static compilation (i would say). sshd, postfix qmgr are not that much larger, and for the rest i had to look what libraries are effectively all linked in, compared to my AlpineLinux VM.

14:16 <stenur> And finally i do not even know whether musl supports all the symbol versioning and other compatibility shims that GNU libc has, at all.

14:17 <stenur> Here for example i still run binaries from installation time, Mar 24 2019. Two glibc updates and one kernel version jump are crossed by those, but it just works.

14:18 <stenur> (In defiance of dynamic linking)

15:50 <cruxbot> [opt.git/3.6]: rust: fix source sum somehow got corupted on second download

16:07 <cruxbot> [contrib.git/3.6]: python3-appdirs: dup opt/python3-appdirs

16:07 <cruxbot> [contrib.git/3.6]: python3-pip: dup opt/python3-pip

16:07 <cruxbot> [contrib.git/3.6]: iso-codes: dup opt/iso-codes

16:09 <cruxbot> [contrib.git/3.6]: nghttp2-libs: dump to v1.45.0

16:09 <cruxbot> [contrib.git/3.6]: acpid(2): update to 2.0.33

16:11 <stenur> As a normal user with only cwm and st i wonder how Wayland gives you the little copy+paste i use.

16:22 <cruxbot> [contrib.git/3.6]: zathura-pdf-mupdf: initial commit, version 0.3.7

16:22 <cruxbot> [contrib.git/3.6]: zathura: initial commit, version 0.4.8

16:22 <cruxbot> [contrib.git/3.6]: mujs: initial commit, version 1.1.3

16:22 <cruxbot> [contrib.git/3.6]: gumbo-parser: initial commit, version 0.10.1

16:22 <cruxbot> [contrib.git/3.6]: girara: initial commit, version 0.3.6

16:23 <ppetrov^^> guys, just to let you know that I think I managed to solve my problem with the static binary compilation

16:28 <ppetrov^^> apparently, it should be like this:

16:28 <ppetrov^^> g++ -o outputFile objFile1.o objFile2 myLibraries.a

16:28 <ppetrov^^> while in the makefile it was:

16:29 <ppetrov^^> g++ -lmyLibraries -o outputFile objFile1.o objFile2

16:29 <ppetrov^^> so, the libs should be in the end, i guess

16:35 <tilman> ppetrov^^: for static linkage the order of arguments matters, yes. static libraries are *not* included in the final link when there are no consumers of the symbols provided by those libraries. mentioning foo.o and bar.o first means the linker will remember which symbols it still needs to resolve

16:36 <tilman> that's the default behavior anyway

16:36 <ppetrov^^> the thing is, this fixed compilation of dynamically linked executable on debian, as well

16:36 <cruxbot> [contrib.git/3.6]: gtk-layer-shell: initial commit, version 0.6.0

16:36 <cruxbot> [contrib.git/3.6]: lfs: initial commit, version 1.0.0

16:37 <ppetrov^^> compiling dynamic binary with the default order in the Makefile worked fine on CRUX, fine on Slackware but not on Debian

16:37 <ppetrov^^> no idea why

17:49 <tilman> ppetrov^^: if you put "libfoo.a" in the linker call, then libfoo will *always* be linked statically. it may still link other libraries (especially libc, or libstdc++) dynamically though

17:49 <tilman> ppetrov^^: use "readelf -a myexecutable | grep NEEDED" or "ldd myexecutable" to see which libraries it will load dynamically

17:51 <tilman> ppetrov^^: if you use "-lfoo", then libfoo may be linked dynamically or statically. depends on the use of the linker switch -Wl,-static , on the presence of libfoo.so and libfoo.a and on the use of linker flags -L (i.e. where to look for libraries)

17:51 <ppetrov^^> thank you

17:54 <ppetrov^^> do I need -Wl if I build static binary?

17:54 <ppetrov^^> sorry for the stupid question..

18:07 <tilman> -Wl just means "pass the next option string to the linker"

18:08 <tilman> i think you want -static -static-libgcc -static-libstdc++ -Wl,-static *before* any of the -lfoo options

18:10 <ppetrov^^> thanks a lot, much appreciated

18:39 <cruxbot> [contrib.git/3.6]: python3-requests: added missing dependency: python3-idna

18:40 <cruxbot> [contrib.git/3.6]: p5-datetime-calendar-julian: 0.105 -> 0.106

19:42 <cruxbot> [core.git/3.6]: ca-certificates: moved comment in a new line