uimaj 3, UIMA AS, and UIMA C++

["Fox, David" <da...@optum.com>]

I know I’ve asked these questions before, but after a while in which my UIMA plans have been on hold, they may be becoming active again, and I thought maybe things are clearer now that uimaj-3.0.0 has reached rc4:

1.  Is there a plan or timeframe for a UIMA AS version of uimaj-3?  (As I understand it, UIMA AS 2 contains its own fork of uimaj-2, rather than being a separate framework)

2.  Is uimaj-3 compatible with the latest UIMA C++ (2.4)?   If not, is there a plan or timeframe for a new release of UIMA C++ which would be compatible?

3.  assuming there are or will be one or more versions of UIMA C++ which would be compatible with uimaj-3, would it/they be compatible with the uimaj-3 support for serializing arbitrary Java objects, or would compatibility be limited to CASes or type systems containing only types which were supported in UIMA 2?

Thanks,
David
This e-mail, including attachments, may include confidential and/or
proprietary information, and may be used only by the person or entity
to which it is addressed. If the reader of this e-mail is not the intended
recipient or his or her authorized agent, the reader is hereby notified
that any dissemination, distribution or copying of this e-mail is
prohibited. If you have received this e-mail in error, please notify the
sender by replying to this message and delete this e-mail immediately.

Re: uimaj 3, UIMA AS, and UIMA C++

[Marshall Schor <ms...@schor.com>]

Hi David,

Here's some answers

1. There's been some thought about the uima-as version that would be compatible
with uimaj-3. There's a branch for uima-as in svn already that is basically
working already with v3 (not released).  We do plan to release a version of
uima-as compatible with v3, but haven't decided on when - these things depend on
priorities, volunteer time, etc.

2. uimaj-v3 was designed to be compatible at the serialization level with v2, so
it should work with uima c++ (2.4).  I think there was an earlier email stating
someone had tried this already and found it was working.

3. Re: support for serializing arbitrary Java objects: that support depends on
the JCas, and is designed so that at serialization time, the data required to
re-create the Java object is updated into the CAS as normal CAS-serializable
data.  When serialized to uima-c++, this data would just stay as normal Cas
data, because (of course) there's no support for Java objects in C++.  But there
would be no limitations re: keeping these objects in the CAS.

Hope that answers your questions; if not, ask more...

And, if you're able/willing, the UIMA project welcomes new volunteers to help
get this work done :-)

-Marshall


On 2/21/2018 10:11 AM, Fox, David wrote:
> I know I’ve asked these questions before, but after a while in which my UIMA plans have been on hold, they may be becoming active again, and I thought maybe things are clearer now that uimaj-3.0.0 has reached rc4:
>
> 1.  Is there a plan or timeframe for a UIMA AS version of uimaj-3?  (As I understand it, UIMA AS 2 contains its own fork of uimaj-2, rather than being a separate framework)
>
> 2.  Is uimaj-3 compatible with the latest UIMA C++ (2.4)?   If not, is there a plan or timeframe for a new release of UIMA C++ which would be compatible?
>
> 3.  assuming there are or will be one or more versions of UIMA C++ which would be compatible with uimaj-3, would it/they be compatible with the uimaj-3 support for serializing arbitrary Java objects, or would compatibility be limited to CASes or type systems containing only types which were supported in UIMA 2?
>
> Thanks,
> David
> This e-mail, including attachments, may include confidential and/or
> proprietary information, and may be used only by the person or entity
> to which it is addressed. If the reader of this e-mail is not the intended
> recipient or his or her authorized agent, the reader is hereby notified
> that any dissemination, distribution or copying of this e-mail is
> prohibited. If you have received this e-mail in error, please notify the
> sender by replying to this message and delete this e-mail immediately.
>