ctakes-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Chen, Pei" <Pei.C...@childrens.harvard.edu>
Subject Re: 3.1.2 release
Date Tue, 24 Jun 2014 11:03:31 GMT
I'm hoping to get a release candidate out this week. 
By the way- it'll be called minor 3.2.0 rather than a patch release... Due to the major changes.


Sent from my iPhone

> On Jun 23, 2014, at 11:53 PM, "Anirban Chakraborti" <chakraborti.anirban@googlemail.com>
wrote:
> 
> When is the release planned.
>> On 24 Jun 2014 02:11, "Masanz, James J." <Masanz.James@mayo.edu> wrote:
>> 
>> 
>> That would be good to know.  I think a good start would be to have a
>> comparison from one release to the next, where when release N is built,
>> some test is run using release N and using release N-1.  The CPE Gui will
>> show a breakdown of how long each annotator took in total for a set of
>> documents. I assume we could get that information programmatically fairly
>> easy.
>> 
>> -----Original Message-----
>> From: AndyMC@apache.org (Andy McMurry) [mailto:mcmurry.andy@gmail.com]
>> Sent: Monday, June 23, 2014 2:57 PM
>> To: dev@ctakes.apache.org
>> Subject: Re: LVG Thread Safety
>> 
>> I wonder how much cTAKES performance changes with even minor changes to
>> LVG.
>> 
>> In principle, thread safety shouldn't change the output, but even minor
>> updates (LVG.2014) makes me wonder.
>> 
>> Thoughts?
>> 
>> On Jun 23, 2014, at 11:58 AM, "John Green" <john.travis.green@gmail.com>
>> wrote:
>> 
>>> Nice!-
>>> Sent from Mailbox for iPhone
>>> 
>>> On Mon, Jun 23, 2014 at 2:57 PM, Chen, Pei <
>> Pei.Chen@childrens.harvard.edu>
>>> wrote:
>>> 
>>>> +dev@
>>>> Chris,
>>>> This awesome news.  Yes we'll be happy to try out the fix.
>>>> Thanks again,
>>>> Pei
>>>> https://issues.apache.org/jira/browse/CTAKES-151
>>>> From: Lu, Chris (NIH/NLM/LHC) [C] [mailto:chlu@mail.nih.gov]
>>>> Sent: Monday, June 23, 2014 2:54 PM
>>>> To: Chen, Pei
>>>> Cc: Browne, Allen (NIH/NLM/LHC) [E]; Lu, Chris (NIH/NLM/LHC) [C]
>>>> Subject: RE: LVG Thread Safety
>>>> Hi Pei,
>>>> How are you doing?
>>>> We have evaluated your request regarding to the thread safety issue on
>> the SPECIALIST Lexical Tools. Bellows are the summary. Please let us know
>> if you have any suggestions, comments, or questions. This SCR is a major
>> source code change and, if you like, we will send you a nightly built
>> version of lvg.2015 (using lvg.2014 data) in July so you can try it before
>> the official NLM internal release of lvg.2015 is released (in Oct.). Please
>> let us know, Thank you!
>>>> 1). Static variables
>>>> => Thank you (and Kim) to point out this issue. Here are how we plan to
>> modify "static variables" in the Java codes:
>>>> 1-a). Change static variables to final static variables (if applicable)
>>>> 1-b). Change static variables to local variables (if applicable)
>>>> 1-c). Keep "static String fieldSeparator_" in Lib.GlobalBehavior.java
>> and use "synchronized" for the associated static methods.
>>>> ð  There will be too many changes for Flows APIs if we decided to
>> change it to local variable.
>>>> 1-d). Keep the rest of static variables under GuiTool
>>>> ð  Assuming users do not use GuiTool (lgt) under multi-thread
>> environment.
>>>> There are 88 files need to be modified for this software change request.
>>>> 2) Standardize Java package namespace convention on lvg API:
>>>> => There were legacy reasons that all Lvg Java codes under directory of
>> Tools do not use the standard Java package convention. However, we are
>> happy to make the change for your requests.
>>>> There are 11 Java files of command line tools and 43 Java files of GUI
>> tool need to be modified.
>>>> Hope this helps. Thank you!
>>>> Best Regards
>>>> -          Chris
>> 
>> 

Mime
View raw message