devfs is a translator sitting on /dev and providing what is to be provided in there in a dynamic fashion -- as compared to static passive translator settings as they're used now.

devfs has not yet been written.

If applicable, it has to be taken care that all code concerning the page-in path is resident at all times.

IRC, freenode, #hurd, 2012-01-29

<pinotree> what would be an hurdish way to achieve something like the
  various system (udev, devfs, devd, etc) for populating devices files
  automatically according to the found system devices?
<pinotree> (not that i plan anything about that, just curious)
<youpi> it's not really a stupid question at all :)
<youpi> I guess translators in /dev
<youpi> such as a blockfs on /dev/block
<antrik> pinotree: in an ideal world (userspace drivers and all), the
  device nodes will be exported by the drivers themselfs; and the drivers
  will be launched by the bus respective bus driver
<antrik> an interesting aspect is what to do if we want a traditional flat
  /dev directory with unique device names... probably need some
  unionfs-like translator that collects the individual driver nodes in an
  intelligent manner

IRC, freenode, #hurd, 2012-04-22

<antrik> braunr: I don't think it's a problem that translators are invoked
  when listing /dev
<antrik> the problem is that they linger around although they are very
  unlikely to be needed again any time soon
<youpi> for now it's not too much a problem because there aren't too many
<youpi> but that can become problematic
<pinotree> a devfs on /dev could also fill it with new devices
<youpi> but only with the ones that actually exist
<pinotree> yeah
<braunr> antrik: i mean, the hurd may lack a feature allowing the same
  translator to be used for several nodes not hierarically related
<braunr> antrik: or rather, it's a special case that we should implement
  differently
<braunr> (with e.g. a devfs that can route requests for different nodes to
  a same translator
<braunr> )
<antrik> I agree BTW that some intermediary for /dev would be helpful --
  but I don't think it should actually take over any RPC handling; rather,
  only redirect the requests as appropriate (with the actual device nodes
  in a hierarchical bus-centric layout)
<braunr> right
<antrik> braunr: actually, the Hurd *does* have a feature allowing the same
  translator to be attached to several unrelated nodes
<braunr> i keep getting surprised :)
<antrik> though it's only used in very few places right now
<youpi> pfinet and ptys at least ?
<antrik> yeah, and console client IIRC