[developers] mrs.system dead?

Francis Bond fcbond at gmail.com
Mon Oct 31 03:54:28 CET 2005


> hmm - having looked at the files, it appears to me that mrs.system is now a
> misnomer, since it does not contain all the MRS code - the function has
> changed a lot compared to the old mrs.system.  It might be cleaner to rename
> it?
> oe at csli.Stanford.EDU said:
> > although it took me some time to remember: actually no.  when building PET,
> > the `mrs' system gets used to compile an MRS library in ECL.  so, to the
> > extent that its current out-of-sync-ness will be problematic in the PET
> > context, whoever re-builds PET next would have to fix things.

Does this mean that the MRS(s) created by cheap with -mrs will be
different from those created by the lkb for the same string?

Francis Bond  <www.kecl.ntt.co.jp/icl/mtg/members/bond/>
NTT Communication Science Laboratories | Machine Translation Research Group

More information about the developers mailing list