[developers] Re: pending/674: generator broken on everything except the ERG?

Ann Copestake Ann.Copestake at cl.cam.ac.uk
Tue Oct 25 19:24:51 CEST 2005


> did you happen to still have revision 1.2 of
> `serialize.lisp' (i.e. one obtained between 3-oct and 5-oct)?

not unless something was wrong with CVS since it was a fresh update from 
yesterday

> while this functionality remains undocumented, supressing the warning
> message is still a good thing, though, in my view.

well, it printed as an error, not a warning.  I appreciate now that the error 
was inside its own handler-case so would not have affected anything other than 
the saving of indices, but this is highly misleading.  I came to the 
conclusion that it was breaking the generator, since something was ...  There 
is no need to have this functionality by default for small grammars, since 
indexing is very fast for them, and there is some overhead in doing so, since 
it'll mean changing the grammars and having to worry about some more files. So 
I suggest we leave it as I patched it.  Note that there is no cache of the 
lexicon for most small grammars since they use read-tdl-lex-file-aux rather 
than the cached-if-available function.

The fresh start gave me a working generator.  Before that I was getting no 
generated strings though I was getting edges.   I had increased 
*maximum-number-of-edges* because it ran out but even after that it still 
wasn't generating, though I didn't get the warning about going over the number 
of edges again. So perhaps there is some complexity from packing?

This is a grammar with qeqs etc

> regarding this second issue, i think packing should be off by default,

packing should be off by default.  it is potentially useful for teaching 
purposes if packing can be switched on interactively

> my apologies if
> the failure comparing handle-free MRSs caused you difficulties!
 
I hadn't checked that yet, but it would have caused problems ...

Ann





More information about the developers mailing list