Interesting task: makewhatis repeatability

[Poul-Henning Kamp]

In message <9140595F-38F1-4589-A0FB-4C0BC4AF022C at lassitu.de>, Stefan Bethke writes:
>> I just found out that makewhatis does not produce the same output
>> after two different installworld runs.
>>
># for i in 1 2; do rm -rf /data/00/build/x; mkdir /data/00/build/x;  
>make installworld DESTDIR=/data/00/build/x; md5 /data/00/build/x/usr/ 
>share/man/whatis >>/tmp/md5; done
>...
># cat /tmp/md5
>MD5 (/data/00/build/x/usr/share/man/whatis) =  
>393e1046893b6504b0d51e57a377c529
>MD5 (/data/00/build/x/usr/share/man/whatis) =  
>393e1046893b6504b0d51e57a377c529

Hmm, I may be seeing some other weirdness then.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.