Enabling technologies and constraints for software sharing in large astronomy projects

About This Presentation
Title:

Enabling technologies and constraints for software sharing in large astronomy projects

Description:

Enabling technologies and constraints for software sharing in large astronomy projects G.Chiozzi ESO, Germany A.Bridger UK Astronomy Technology Centre, UK –

Number of Views:93
Avg rating:3.0/5.0
Slides: 27
Provided by: Gianl6
Learn more at: http://www.eso.org
Category:

less

Transcript and Presenter's Notes

Title: Enabling technologies and constraints for software sharing in large astronomy projects


1
Enabling technologies and constraints for
software sharing in large astronomy projects
  • G.Chiozzi ESO, Germany
  • A.Bridger UK Astronomy Technology Centre, UK
  • K.Gillies Space Telescope Science Institute
  • B.Goodrich, S.Wampler - National Solar
    Observatory, Tucson, AZ
  • J.Johnson, K.McCann W.M.Keck Observatory,
    Kamuela, HI
  • G.Schumacher National Optical Astronomy
    Observatories, La Serena, Chile

2
Aspects analyzed
  • Why are we now discussing sharing?
  • What do we mean bysoftware sharing?
  • Benefits of software sharing
  • Barriers to software sharing
  • Fostering reuse
  • Opportunities for sharing
  • Conclusion or just the beginning?

3
The right time?
1990
4
Software Sharing what is it?
Reusing software assets
Black boxorwhite box
Horizontalorvertical
Opportunisticorplanned
5
Benefits
  • SW Engineers agree that sharing can bring major
    benefits
  • Improved productivity and quality while
    decreasing cost.
  • Saving time and money while achieving greater
    reliability
  • More effective use of domain experts
  • Building an improving knowledge base
  • Improvements are propagated throughout all
    systems
  • Standards can be developed and deployed
  • Reduced maintenance costs (shared maintenance and
    less code)

6
are they measurable?
Lines of code for TCS or ESO telescopes
7
are they measurable (cont.)?
Reduction of lines of code for instruments
8
Barriers
  • Technical
  • Functionality, programming language, OS
  • Local standards and data formats
  • Sociological/project management
  • Timing between projects
  • Not invented here, project culture
  • Cost and politics of sharing.

9
Fostering reuse
  • Open source practices
  • Solid core
  • Process of contributing
  • Connecting the community
  • Understanding collaboration
  • Recognition of creative nature
  • Community interest and management sponsorship
  • Working groups
  • Getting people involved
  • Involve management, staff and customers
  • Conferences and joint papers
  • Meetings in small groups, exchanges, sabbaticals
  • Federating with related communities

10
Opportunities for sharing
11
Opportunities for sharing (2)
  • Software Engineering Standards
  • Coding, development, design and process
  • Design, knowledge and experience
  • Similar architectural designs
  • Library of domain patterns?
  • Knowledge base for astronomical control SW
    engineers?

12
Opportunities for sharing (3)
  • Functional Interface Standards
  • Very similar observatorys SW architecture
  • Should be possible to standardize functional
    interfaces
  • Why is there NO generic and portable TCS?
  • Data Interface Standards
  • Data structures are often hidden behind
    functional IF
  • Is it possible to standardize instrument
    configuration or the configuration of some
    subsystems?
  • Project Data Model? User Database?
  • The Data Handling community is ahead (FITS)!
  • But do we have the requirement?

13
Opportunities for sharing (4)
  • Middleware and infrastructure
  • Common services
  • Hide distribution and heterogeneity
  • Is a common infrastructure feasible?
  • Should we put up a working group?
  • Software packages
  • There are a number of packages in use (sla,
    cfitsio)
  • Need tools to collect, review and discuss packages

14
Conclusion
  • We do not want to develop yet another
    infrastructure
  • We want to concentrate on our domain work
  • We have analyzed different means for sharing,
    depending on the context
  • We have seen that collaborations are now more
    practical and affordable than ever
  • We aim at
  • Sharing lessons learned
  • Identifying areas for greater cooperation
  • We dream of broad scale, component based, reuse
  • How do we get there?.....

15
Questions?
We represent just a tiny subset of the projects
in astronomy. Many more colleagues in the
astronomical observatory community have given
their ideas and time as we have developed this
paper. We want to thank in particular Dave Silva
and Anders Wallander for their participation and
essential contribution to these discussions in
the last year.
  • Contact
  • gchiozzi_at_eso.org

In the paper you can find an interesting
bibliography!
16
OPEN DISCUSSION Creating an open software
community for observatorycontrol and operation
Is it possible? Is it worthwhile?
17
What is the right path?
  • Is this a good idea, in practice?
  • What can/should we do?
  • How?
  • Who would like to join in the effort?
  • The next steps?

18
Opportunities for sharing
  • Software Engineering Standards
  • Design, knowledge and experience
  • Functional Interface Standards
  • Data Interface Standards
  • Middleware and infrastructure
  • Software packages

19
OPEN DISCUSSION Creating an open software
community for observatorycontrol and operation
Is it possible? Is it worthwhile? Summary!
20
The discussion
120 people
  • Is this a good idea?
  • Yes it is!

21
Highlight quotes
  • We DO LIKE to reuse what we do NOT LIKE to do
    ourselves
  • Analyze first what succeeded
  • To not get entangled in technology discussions
  • Focus on domain expertise
  • I reuse well documented things
  • I do not want to find The Solution but I want
    to see multiple options

22
What can/should we do for a start?
  • Mailing list / forum
  • Wiki
  • Who is going to host them?
  • Create an initial list of topics
  • Rate them and select the 5 top by
    priority/interest
  • Identify responsible person for each area
  • Look after 1 year and make a checkpoint

23
Periodical Meetings
  • ADASS (once/year)
  • BoF session
  • Already some more concrete ideas
  • SPIE (every two years)
  • Some phone meetings to coordinate the activities
  • Sub-groups would decide if/how/when to meet

24
Some issues
  • Look at similar initiatives for procedures/tools
  • NASA Earth Science Data Systems Software Reuse
    Working Group (http//www.esdswg.org/softwarereuse
    )?
  • Opticon?
  • Can we directly piggy back on one?
  • Keep it very domain specific?
  • People should offer their body and soul for
  • Enhancing the original core team
  • Coordinating and keeping alive the topics
  • Look for resources
  • Man power approval inside our organizations
  • Funding from agencies (NSF, EU FPx programs?)

25
Tomorrow?
  • Think think think what can YOU do to make it
    really happen?
  • Get in touch with me (gchiozzi_at_eso.org) or
    anybody of the authors of the paper, to offer
    your collaboration.
  • We will quickly setup an initial mailing
    list//forum and wiki
  • Follow up
  • Spread the word
  • Convince your boss to give you some resources

26
Thanks!!!!!
Contact gchiozzi_at_eso.org I will put links and
this PPT in my home page
http//www.eso.org/gchiozzi Will move things as
soon as we will have the wiki and redirect from
there.
Write a Comment
User Comments (0)
About PowerShow.com