|
May 20
|
|
| Ruth Kneale |
NSO |
What's a web CMS anyway, and why would we want one? |
|
After hand-coding HTML for years with emacs, a major review
and web site overhaul brought with it fits and eye-crossing numbness. A
better way must be found! Thus began a six-month evaluation of web
content management systems (CMSs), and a further project to migrate the
existing Advanced Technology Solar Telescope (ATST) website into the
chosen CMS. I'll describe the evaluation process and then go into more
detail on the winning solution, Drupal, and how ATST will be using it,
along with plans for the future.
|
|
May 13
|
|
| Jeff Kantor |
LSST |
LSST DM Middleware Overview |
|
The Data Management System (DMS) has a vertical architecture with: an
Applications Layer containing scientific pipelines and data products; a
Middleware Layer containing control & management, data access ,
distributed processing, security, and interface services; an Infrastructure
Layer containing computing, storage, and communications hardware and system
software.
We will discuss the Middleware Layer support for: Portability to clusters,
grid, workstations; standardization of services so applications behave
consistently (e.g. recording provenance); keeping "thin" for
performance and scalability; open source, Off-the-shelf Software, and
Custom Integration development approach.
|
|
|
|
Apr 29
|
|
| Rob Seaman |
National Optical Astronomy Observatory |
A Target of Opportunity Observing System for NOAO |
|
During its half century, NOAO has accumulated extensive knowledge of block
scheduled ("classical") observing. This experience is
logistically realized in a system of systems from the observing proposal
process and run preparation, to evolving telescope and instrument
capabilities, through data acquisition and handling, data transport to
archives and VO portals, data reduction both through pipelines and toolkit
frameworks (IRAF), to data analysis and scientific publication.
Over the past twenty years or so, NOAO and partner observatories have
likewise mastered techniques of queue observing, remote observing, and
other single telescope alternate observing modes.
The advent of worldwide networks of small-to-midsize robotic telescopes,
and of the "System" of medium to very large human-mediated
telescopes, offers scientifically compelling opportunities for
commissioning new modes of transient response observing.
I will discuss the system engineering, architecture and design of a
VOEvent-driven system to support adaptive, perhaps autonomous, observing
modes at NOAO.
|
|
|
|
Apr 22
, 11:30am - 1:30pm
|
|
| Terry Bahill |
Department of Systems and Industrial Engineering, University of Arizona |
System Architecture with Tinker Toy Towers |
|
The software engineering process is a perennial topic for boisterous debate
- from Waterfall and Spiral Development through ICONIX to Extreme
Programming and Agile Methods. Software engineers may benefit from
comparing and contrasting these methods to those of the broader Systems
Engineering Community.
Dr. Terry Bahill of the UA's SIE Department will facilitate a novel
hands-on exercise that will clarify key aspects of System Engineering
shared across disciplines.
This is a special two hour Software Brown Bag beginning one half hour
early at 11:30 am and continuing until 1:30 pm.
|
|
Apr 15
|
|
| John Bolding |
National Solar Observatory |
GONG Merge/Ring Pipeline II: Using Conductor to Make Pipelines Sing or Velocities to Ring Diagrams in 1664 Easy Steps |
|
Conductor, a pipeline tool developed by the Planetary Image Research
Laboratory at the University of Arizona, is an application for managing
queues of source files to be processed by sequences of procedures.
I will discuss the use of Conductor in the second generation GONG
Merge/Ring Pipeline.
|
|
|
|
Apr 8
|
|
| Phil Warner |
National Optical Astronomy Observatory |
Spring Fling |
|
Spring is a popular, unobtrusive dependency injection framework used by
many Java applications and services developers. Spring was created, among
other reasons, to simplify the life of a Java/J2EE/JavaEE programmer by
eliminating the need to hard-code references to implementation-specific
classes and coupling between, e.g., business logic, protocol and service
implementations, and the underlying framework. I'll present a brief
overview of some of Spring's features, and discuss how the framework is
used in the NOAO VOEvent software.
|
|
|
|
Mar 25
|
|
| Steve Wampler |
National Solar Observatory |
The ATST's Software Framework |
|
The NSO Advanced Technology Solar Telescope (ATST) software is
built using an in-house developed application framework based on
a Container/Component Model. This is a highly flexible framework
that attempts to be "middleware neutral". Some of the key
features
of the framework are introduced along with some performance measures
when used with the ICE communications middleware and a PostgreSQL
database.
|
|
|
|
Mar 18
|
|
| Jeff Kantor |
LSST |
LSST DM Infrastructure Overview |
|
The Data Management System (DMS) has a vertical architecture with: an
Applications Layer containing scientific pipelines and data products; a
Middleware Layer containing control and management, data access,
distributed processing, security, and interface services; an Infrastructure
Layer containing computing, storage, and communications hardware and system
software.
We will discuss the Infrastructure Layer support for: Distributed
processing and data; specialization of environments for real-time alerting
vs peta-scale data access; Off-the-shelf, Commercial hardware &
Software, Custom Integration development approach.
|
|
|
|
Mar 11
|
|
| Jeff Kantor |
LSST |
Requirements Flowdown with LSST SysML and UML Models |
|
LSST is a survey telescope that has scientific requirements driven by 4
reference missions at the single image and full survey levels. These
requirements flow down to system level requirements for optics, cadence,
calibration, data products, and data quality. Finally, these system
requirements flow down into requirements on the main system elements of the
LSST: Telescope and Site, Camera, and Data Management. All of the above
requirements are documented in System Modeling Language (SysML), and the
Data Management requirements are then allocated to Unified Modeling
Language (UML) Use Cases to define the DM software requirements. We will
discuss the structure and content of the requirements, and how we are using
SysML, UML, and the Enterprise Architect tool for this effort.
|
|
|
|
Mar 4
|
|
| David Gasson |
National Optical Astronomy Observatory |
Ruby.learn { |stuff| } |
|
Ruby is a a dynamic, open source programming language with a focus on
simplicity and productivity. In this overview, I'll take a look at Ruby
and its ecosystem from a (relatively) high-level perspective, and explore
some of the ways that Ruby is currently being used.
|
|
|
|
Feb 26
|
|
| Evan Deaubl |
National Optical Astronomy Observatory |
Tapping into a World of Services using Mule |
|
With the expansion of widely distributed networks and the continued growth
of the ideas of SOA (Service Oriented Architecture) and SaaS (Software as a
Service), software systems increasingly need the ability to connect
services running on multiple different systems, connect to software
services run by other organizations, and make their own services visible to
users from the outside. An Enterprise Service Bus solution simplifies
these tasks by greatly reducing the amount of work needed to create and
connect services. We will discuss one particular Java ESB implementation,
Mule, and how it can be used to fulfill these needs.
|
|
|
|
Feb 12
|
|
| Brian Thomas |
University of Maryland |
Semantic Technologies and Standards |
|
Semantic technologies and standards (ex. RDF/OWL; Jena/ARQ/Joseki Java
packages for semantic data representation, query and webservice, Reasoners
such as Pellet; query languages such as SPARQL) together offer a gleaming
vision of powerful, precise, intelligent and distributed access to data.
Our group has been investigating how to leverage this technology for doing
archive science.
Some requirements are clear: adopting semantic technologies requires that
we
are able to layer the new technologies on top of old ones, that we ensure
that we construct basic scientific data models and queries which make
science possible (e.x. errors!) and all at the same time make the
(scientist) user interface to this system both precise and easy to
learn/manipulate/understand.
In this talk I will canvas semantic technologies we have found useful for
doing archive science, indicate a framework for combining these
technologies in an archive setting, and then present our effort at
developing a graphical user interface which overlies a science
knowledgebase.
|
|
|
|
