After a Linux build has completed and installed, the manpage /usr/X11R6/man/man1/XDarwin.1x.gz is installed on the system. Not very useful on Linux.
Any idea how this happens? It should be possible to track this down from an install log. On Darwin, the XDarwin man page is installed by the Imakefile in Xserver/hw/darwin, which Linux should not be using. The manpage lives in this directory. On the other hand, a reason someone might want the XDarwin man page installed is that it is mentioned in the the Xorg manpage along with XWin, the Cygwin X server. Is the XWin manpage installed?
No, the XWin manpage isn't installed. Documenting software that doesn't ship with the packages seems to be a broken idea IMHO. Ultimately the Xorg manpage being dynamic and only referencing other pages that actually exist on a given platform would be nice, but I'm not sure how feasible it would be to implement that. The simple solution here, is to just not install XDarwin.1x on Linux.
A quick "grep -r XDarwin.man *" reveals that the XDarwin man page is included by xc/doc/man/misc/ Imakefile. The comment at the top of the Imakefile reveals it is doing exactly what it is designed to do and that it installs a number of other manpages without binaries besides XDarwin. /* * This Imakefile is intended to make it possible to install man pages * for architectures other than the one the build is done on. So far, * it is only set up for builds on XFree86/ix86 platforms, which means that * it only includes those man pages not normally installed there: * * XFree86 Linux/mips: newport * XFree86 Linux/sparc: sunbw2, suncg14, suncg3, suncg6, sunffb, sunleo, * suntcx * XFree86 Linux/ix86: v4l, glide * XFree86 Linux: ur98 * XFree86 Darwin: XDarwin, dumpkeymap */ I don't have a strong opinion on whether or not this is desirable behavior. (According to the Imakefile comment it is a bug that it does not include the XWin man page.) There is precedent to install man pages for files that don't ship with packages. The XDarwin man page also clearly indicates that it is for Darwin/Mac OS X only and would be taken by a Linux reader as purely informational. Perhaps Roland can give more details on this since he was the last person to touch this file.
It has been this way for quite a while -- going back at least until 2001. I don't know whether I consider this a bug or not. Should this really be a release blocker?
In a standard Red Hat X build, the XDarwin manpage has never appeared, which is how I discovered this. When updating to a new X release, any files that are present after installation into the buildroot, but are not listed in the file lists, get flagged by rpm, which warns of the files existing in the buildroot. The manpage was there for a long time perhaps, but was never installed on Linux before, at least on Red Hat systems. I marked the issue as "trivial" as it isn't a major issue either way, however while I assessed it as a bug myself, if others consider it not to be a bug, I've no strong feelings either way, and I've patched our rpms to remove the file post %install section. Not particularly a release blocker I guess.
After discussing this with the release wranglers, it was decided that if anyone has time to do this before the release it would be considered for inclusion, but it is not a release blocker. Taking off blocker list.
Is this still valid?
Yes, this is still valid.
doesn't happen in 7.1.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.