[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: hurd does NOT need /hurd

Joey Hess <joey@kitenet.net> writes:

> I appreciate that under the hood hurd servers are substantially
> different from linux kernel modules, but that's beside the point; perl
> modules are substantially different from C libraries under the hood, and
> yet they both belong in lib because of the way the files are used.

But Hurd passive translators are not only different under the hood,
but different to the user's perspective.  

> I think one problem is that settrans seems to have no search path
> capabilities, so you are stuck trying to find a directory near the root
> to put these in to prevent excessive typing all the time. 

Passive translators are, by convention, stored in /hurd.  It's not
settrans that lacks a search path capability, but rather the parent
filesystem that lacks it.  And part of that is because the user's
context is long gone by the time the translator is started.  The
settrans program just squirrels its arguments away into the
filesystem, and then when the requested node is opened, the specified
command is run as a translator.

I think the best way to understand why path search is Very Hard here,
is that it's basically the same issue as the location of command
interpreters for #! execution.


To UNSUBSCRIBE, email to debian-hurd-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

Reply to: