uima-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Adam Lally <ala...@alum.rpi.edu>
Subject Re: small memory footprint tradeoff configuration
Date Wed, 25 Feb 2009 14:54:50 GMT
On Wed, Feb 25, 2009 at 9:07 AM, Eddie Epstein <eaepstein@gmail.com> wrote:
>> It seems like Marshall's angle (if I understood it) is not really GC
>> at all, but a model where an annotator decides to explicitly delete
>> FS.  I could be okay with that idea, too.  A GC model by definition
>> should preserve any referenced FSs, but if we say we have an explicit
>> deletion model where anybody can delete anyone else's stuff, at least
>> we won't confuse people about what's going on.  Current applications
>> that use existing annotators would not break (because the annotators
>> would not delete anything), and if a new annotator is introduced that
>> breaks the application, it's the annotator's fault for being too
>> aggressive in deleting stuff that someone else might still need.
>
> What is different here? UIMA already lets any annotator delete anything
> in the CAS, where deletion is defined by removing FS from the index
> and removing references. GC would just add the ability to reclaim FS
> heap space.
>

By "explicitly delete" I meant something akin to a
non-garbage-collecting language like C.  The user's code has to say
explicitly what FS to delete, and if those happen to be still
referenced by something, that's their problem.  I thought that might
have been what Marshall was suggesting.  I thought he was suggesting
an explicit operation that an annotator would need to call in addition
to making an FS unreferencable.  But maybe I misunderstood.  Anyway, I
prefer a GC option rather than this.

  -Adam

Mime
View raw message