At 12:41 Uhr -0700 20.01.2002, Jeff Whitaker wrote: >On Sun, 20 Jan 2002, Max Horn wrote: > >> May I ask, why does the man package a) use the man version from linux >> > >It's not "from" linux - the kernel.org site was just a convenient download >location.
OK, it's not a linux version, sorry. But it's definitly a different man implementation than that used in Darwin. E.g. it uses a different (incompatible) man.conf format, and react in different way to e.g. "man --" . I am not saying this is necessarily bad, of course, it has added functionality which might be a boon. > >> b) why does it compile its own man version at all - why not a >> thin script that calls /usr/bin/man but with a custom config file? > >Because that's the way I did it - if someone has a better solution that's >fine with me. > > > >> I don't particulary like the changed behaviour of this man, and in >> addition, I am basically force to use it, unless I decide to remove >> tcltk and python, too (which I did for now) :/ > >What behavior don't you like? For example the fact that it uses less instead of more, and doesn't call it with -e or -E. > As far as I can tell it behaves exactly the >same way as the built-in man, except it looks in the "n", "l" and "o" >sections. Well, and some other places where it now looks etc. > >> >> A better solution should be found. In particular, why does tcltk have >> to depend on the new man at all? It's working without it, too. Just >> change the depends to a Suggests field entry, I'd say. > >The most useful man pages in the tcltk package are in the "n" section - I >think it should be a dependency since it is practically essential to >anyone programming in tcl/tk. The same is true for the gmt package, all >the man pages get installed to "l". I still think that is a bug in those packages, oh well. >If anyone wants to write a "thin" script go ahead, I'd be happy to remove >the man package if it is not needed. How about this (not tested, straight from my head): #!/usr/bin /usr/bin/man -C /sw/etc/man.conf $* Better of course would be to check if the user passed another -C option. This shouldn't be hard to implement though. Cheers, Max -- ----------------------------------------------- Max Horn Software Developer email: <mailto:[EMAIL PROTECTED]> phone: (+49) 6151-494890 _______________________________________________ Fink-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/fink-devel
