On 2021/02/22 14:05, Alex Raschi wrote: > I attached the new versions of skalibs (2.10.0.2) and execline > (2.8.0.0), these fixes a few bugs and change backtick(1) options > slightly. > > I also attached a new port of the newly created mdoc(7) ports of the > execline HTML documentation. With this one i get: > > Warning: execline-man-pages-2.8.0.0.1 conflicts with etsh-5.4.0v0 > (shells/etsh):/usr/local/man/man1/if.1 > > However shells/etsh does not seem to provide a real if(1) command, i > checked the PLIST of etsh and the if command seems to be an internal > shell command (execline provides an if command but does not conflict > with etsh). Any suggestion to fix this?
either register the conflict with @conflict markers (in both ports), or install docs to a different dir. conflict markers seems ok to me. it might be better to just include the manuals in the main package. you can use multiple DISTFILES. > As said in the previous emails i get these with execline too: > > in default FLAVOR: the following libraries in WANTLIB look like masked by > RUN_DEPENDS: skarnet > in FLAVOR "static": the following libraries in WANTLIB look like masked by > RUN_DEPENDS: skarnet > > I have also checked that these ports work with -fno-common. > > Any comments and/or OKs? SHARED_LIBS = execline 2.8 SHARED_LIBS = skarnet 2.10 start with 0.0. if the build system doesn't produce a library with the right name to match this, patch or pass in via make(1) variables until it does. https://www.openbsd.org/faq/ports/specialtopics.html#SharedLibs @so lib/libexecline.so @lib lib/libexecline.so.${LIBexecline_VERSION} @bin lib/libexecline.so.2.8.0.0 @so lib/libskarnet.so @lib lib/libskarnet.so.${LIBskarnet_VERSION} @bin lib/libskarnet.so.2.10.0.2 there should just be "@lib lib/libxyz.so.${LIBxyz}" lines, no symlinks etc. these are probably what's responsible for portcheck's "masked by" warning. pkg/DESCR says "has no security issues" a bold claim! I don't think it's a good idea to include that bit. I would drop the static flavours unless there's a really good reason for it (usually "it's helpful to run in a chroot for a webserver").