On Mon, May 20, 2002 at 08:25:19PM +0200, Marcus Brinkmann wrote: > On Mon, May 20, 2002 at 10:37:07AM -0500, Steve Langasek wrote: > > Freeswan has gotten by just fine without the creation of a new toplevel > > directory for these commands; the supplemental binaries all exist in > > /usr/lib/ipsec/, > So the program ipsec is looking in that place for them. This does not > happen with settrans. And it is not ipsec <anything the user likes>, but > ipsec <some specific command from a fixed set selected by the system > administrator>, right? > This looks quite different to me. It seems to me as that those programs are > called indirectly by the ipsec program, which has some intimate knowledge > about where to find helper programs to execute the desired operation. > Knowledge the user doesn't has and doesn't need. > No such knowledge about Hurd translators exists in settrans, and it does not > look anywhere for running them. It doesn't even care what you try to > invoke. The user has to do all the work here. Ah, ok -- because the argument to settrans can be an arbitrary command, such as the sh example you gave for preloading debugging libraries; and therefore, settrans cannot hope to successfully interpret and 'autocomplete' the path to the translator. That means the push for a /hurd directory boils down to user convenience (i.e., shorter path name). I can understand that. Steve Langasek postmodern programmer
Attachment:
pgplCnsmxo0E2.pgp
Description: PGP signature