[Developers] [ADMB Users] Save the date - June 20 - 23, 2011

[John Sibert]

Regarding the public accessibility of the ADMB source cod. Our plan was 
to release a new version of the source code last month and keep the 
code.google site up to date with frequent updates. Unfortunately we hit 
a few snags in the buildbot to build some of the windows binary 
distributions. I'm hoping we can fix it soon. Meanwhile if Ian (or 
anyone else) wants to get at the source code, just sent an email to 
Johnoel (who is taking a few days off).

Yes we should discuss public accessibility of the source code at the 
workshop.  Checking out a working copy, modification, testing and 
committing will certainly be on the agenda. Did I mention testing? Also 
we need to discuss whether we need 2 repositories. I'm leaning to yes. 
One repository for development, testing and building binary 
distribution. And one for general access. Maintaining code for building 
binaries for multiple windoze platforms is a pain, and it would be a 
mistake to make code that does not build on all platforms publicly 
accessible. I think we are headed in the direction Ian suggests.

Cheers,
John

On 01/04/2011 08:39 AM, Ian Taylor wrote:
>  Hi Developers,
> Related to the code, I would like to revisit the long-discussed topic 
> of the public accessibility of the ADMB source code.
>
> It seems to me that a higher priority than adding to the code would be 
> to open up the working source code repositories. I think it looks 
> really bad for the ADMB Project to have publicly available code that 
> hasn't changed since 2009: 
> http://code.google.com/p/admb-project/source/browse/. And now that 
> source code for the beta version 10 is available for download, why not 
> have the Google Code page mirror the working SVN repository which 
> contain the latest stuff, or in some other way make the SVN repository 
> accessible (as read-only) to the general public? I think the 
> contributions that Mark wisely calls for will come more readily and be 
> debugged more quickly if they can be more easily viewed.
> -Ian
>
> On 1/4/2011 9:30 AM, John Sibert wrote:
>> HI Mark,
>> Good suggestions, thanks. We need to be careful to not spend too much 
>> time on C++ fundamentals. I would hope that participants would come 
>> prepared.
>>
>> Speaking of being prepared, I would suggest that participants have a 
>> look at "The Elements of C++ Style" by Misfeldt, Bumgardner, and 
>> Gray. (Available in paperback from Amazon 
>> http://www.amazon.com/Elements-Style-Sigs-Reference-Library/dp/0521893089). 
>> It helps is developers use a consistent style.
>>
>> I'll circulate a draft agenda in March.
>> Cheers,
>> John
>>
>> On 01/03/2011 08:16 AM, Mark Maunder wrote:
>>> John,
>>>
>>> One thing we need to make sure is that we go through all the steps 
>>> to add some code to the codebase. Including getting the source code, 
>>> changing the source code, commenting the source code, documenting 
>>> the source code, testing the source code, submitting the source code.
>>>
>>> Related to this is how to create a function (passing and receiving 
>>> arguments), overloading functions so they work with multiple 
>>> variable types, ....
>>>
>>> I also think we should go over creating, using, and distributing 
>>> your own libraries, which might be the initial testing bed and 
>>> distribution of code before it gets accepted to the base code
>>>
>>> Finally, we need to discuss the process of how to decide on what 
>>> code gets accepted into the base code.
>>>
>>> An ambitions, but perhaps useful task might be to add a 
>>> PRIORS_SECTION to ADMB so that priors can be added to any estimated 
>>> or derived parameter
>>> (I know that this does not add much, but it would provide more 
>>> structure to the code and could clarify things for new users).
>>>
>>> e.g.
>>>
>>> PARAMETER_SECTION
>>>    init_vector ln_a
>>>    init_number ln_b
>>>    vector a
>>>    number b
>>>
>>> PRIORS_SECTION
>>>    ln_a~normal(0,1)
>>>    b~prior_t(10,3)
>>>
>>> Mark
>>>
>>>
>>>
>>> Mark Maunder
>>> Head of the Stock Assessment Program
>>> Inter-American  Tropical Tuna Commission
>>> 8604 La Jolla Shores Drive
>>> La Jolla, CA, 92037-1508, USA
>>>    Tel: (858) 546-7027
>>> Fax: (858) 546-7133
>>> mmaunder at iattc.org
>>> http://www.fisheriesstockassessment.com/TikiWiki/tiki-index.php?page=Mark+Maunder 
>>>
>>>
>>> Visit the AD Model Builder project at
>>>   http://admb-project.org/
>>>   See the following website for information on fisheries stock 
>>> assessment
>>> http://www.fisheriesstockassessment.com/
>>>
>>>
>>> -----Original Message-----
>>> From: users-bounces at admb-project.org 
>>> [mailto:users-bounces at admb-project.org] On Behalf Of John Sibert
>>> Sent: Monday, January 03, 2011 9:55 AM
>>> To: developers at admb-project.org; users at admb-project.org
>>> Subject: [ADMB Users] Save the date - June 20 - 23, 2011
>>>
>>> Colleagues,
>>>
>>> Apologies for duplicate posting.
>>>
>>> I'm pleased to announced that we have finally been able to organize our
>>> first ADMB Developers' Workshop. It will be held at the National Center
>>> for Ecological Analysis and Synthesis in Santa Barbara, California, 
>>> June
>>> 20 - 23, 2011. We also have limited funding available to assist with
>>> travel and participation.
>>>
>>> At this point, the workshop agenda and format are completely open. Some
>>> of the topics that have been suggested include writing adjoint code,
>>> parallelization, R interfaces, installation packages, ... . Others have
>>> suggested breaking into small groups to implement specific features 
>>> (but
>>> which ones).
>>>
>>> So my purpose in writing to invite you to participate, to solicit
>>> suggestions for workshop topics, and to find the names of others who
>>> might make a substantive contribution to the proceedings.
>>>
>>> Cheers,
>>> John
>>>
>>
>

-- 
John Sibert
Emeritus Researcher, SOEST
University of Hawaii at Manoa

Visit the ADMB project http://admb-project.org/