incubator-ctakes-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Masanz, James J." <Masanz.Ja...@mayo.edu>
Subject RE: release plans?
Date Fri, 07 Sep 2012 14:50:49 GMT
I think  we need 2 issues opened regarding the type system. I will do
that today. One for things that are additions and won't impact current
users, and another for things that require refactoring and could affect
current users, such as the fix to be creating an (instance of a subclass
of) Event for a disease/disorder etc instead of an (instance of a
subclass of)  Entity.

To mark the release that we plan to fix an issue by, perhaps we could
use the "Fix Version/s" field. Or does that normally include the version
that a patch is released for (if we ever create patches). 

I propose the following:

2.6-incubating
 CTAKES-7  Pre-Migration- Update/Migrate current cTAKES documentation  
 CTAKES-13 Pre-Migration- Create cTAKES 2.6 release (what would have
been 2.6 in SF) 
 CTAKES-14 TimeMention is missing feature for what is being manually
annotated as "class"  
 CTAKES-19 ctakes constituency parser does not generate function tags  
 CTAKES-24 extraneous jars in cTAKES 2.5 
 CTAKES-33 Add Relation Extractor  
 CTAKES-36 remove generated types from source tree 
 CTAKES-42 user guide - Note that test1.xml uses CDA not plaintext AE  

3.0-incubating
 CTAKES-12 Upgrade to cTAKES components to latest Lucene version.  
 CTAKES-16 use uimaFIT's selectCovered() instead of UIMA's subiterator 
 CTAKES-18 add XxxxEntity|EventMention types for other NEs similar to
MedicationEventMention 
 CTAKES-25 retrain rest of models using OpenNLP 1.5  
 CTAKES-27 Upgrade cTAKES dependencies/libraries 
 CTAKES-28 Externalize dictionarylookup/UMLS password to password file
or variable 
 CTAKES-30 Maven integration of cTAKES components  
 CTAKES-31 constrainToWindow assumes tokens are not sorted 
 CTAKES-32 dictionary lookup creates new LookupTokens many times 
 CTAKES-35 SyntaxAttributeCalculator.java hardcodes file names 
 CTAKES-38 Orange book is not the most recent  
 CTAKES-39 Upgrade LVG to a later version  
 CTAKES-44 Tokenizer does not handle Windows-style newlines  
 CTAKES-47 codes and OUI in CreateLuceneIndexForSnomedLikeSample.java  
 CTAKES-48 negation annotator - Sentence type used in code not in descr

 CTAKES-50 Remove UIMA Deprecated references in Annotators 
 CTAKES-52 Context sensitive tokenizer misses certain kinds of time
expressions  
 CTAKES-53 move UMLS out of Apache distribution  
 CTAKES-54 rename packages to org.apache.ctakes  
 CTAKES-55 rename modules consistently 


Regards, 
James Masanz

> -----Original Message-----
> From: ctakes-dev-return-323-
> Masanz.James=mayo.edu@incubator.apache.org [mailto:ctakes-dev-return-
> 323-Masanz.James=mayo.edu@incubator.apache.org] On Behalf Of Steven
> Bethard
> Sent: Friday, September 07, 2012 6:07 AM
> To: ctakes-dev@incubator.apache.org
> Subject: release plans?
> 
> So I know we're still waiting on Apache to upload the code to SVN
> (https://issues.apache.org/jira/browse/INFRA-5079) but perhaps we
could
> already be talking about what our upcoming release plans look like?
Here are
> several things that I know are in the pipeline, but I don't know which
release
> they're planned for:
> 
> * Move UMLS out of Apache distribution
> (https://issues.apache.org/jira/browse/CTAKES-53)
> * Rename packages to org.apache.ctakes
> (https://issues.apache.org/jira/browse/CTAKES-54)
> * Rename modules consistently
> (https://issues.apache.org/jira/browse/CTAKES-55)
> * Change build infrastructure to Maven
> (https://issues.apache.org/jira/browse/CTAKES-30)
> * Update UIMA type system so that it is possible to load SHARP
annotations
> (James, could you and Stephen Wu file an issue for this with
appropriate
> details?)
> 
> My current guess is that the release plan looks something like:
> 
> 2.6-incubating: CTAKES-53
> 3.0-incubating: everything else
> 
> Is that right? Are there other open issues that have to be addressed
before
> 2.6-incubating?
> 
> Also, does Jira let us mark the release that we plan to fix an issue
by? I didn't
> see any way of doing that. If we could do that, it would be an easy
query to
> find out what needs to be done for what release.
> 
> Steve


Mime
View raw message