fenics team mailing list archive

["Johan Hoffman" <jhoffman@xxxxxxxxxx>]

Anders,

Sounds like a very good idea! Below is some input on the plan:

Part I: Our group could add material on mesh refinement, ALE technology
including mesh smoothing, parallel computing (including mesh refinement),
adaptivity and a posteriori error estimation.

Part II: We could contribute to the DOLFIN chapter.

Part III: Here we could contribute with a number of industry/research
examples. What is the difference between "Contributions X" and "Something
X" ?

/Johan


> Dear all,
>
> I have talked to Springer about publishing a FEniCS book and they are
> very interested. The plan is to have a draft ready in time for the
> FEniCS'09 meeting next summer.
>
> The book will summarize the current state of FEniCS, function as a
> reference for users, and showcase interesting applications built with
> FEniCS.
>
> I therefore invite anyone interested to review the suggested outline
> below and give me feedback (either online or offline). Let me know how
> and to which extent you would like to contribute to the book. I have
> suggested some chapters below, but it's entirely up to the authors to
> form their contribution!
>
> The suggested deadline for contributions (title + authors) is
> October 1 2008, with a final deadline for the finished chapters
> on May 1 2009, after which we will have a review process.
>
> The book will consist of four parts:
>
>   Part 0:
>     Introduction  (20-50 pages)
>     Tutorial      (20-30 pages)
>
>   Part I: Methodology (10 x 15 pages)
>
>     This part will contain summaries/reviews of published results,
>     but could also contain new results.
>
>   Part II: Implementation (10 x 30 pages)
>
>     This part will give a detailed presentation of our packages and
>     their design by the package authors, more in the style of
>     tutorials than journal papers. Other contributions may include
>     discussions of particular aspects not related to a specific
>     package like code generation and interface design.
>
>    Part III: Application (10 x 15 pages)
>
>     This part will demonstrate interesting applications where FEniCS
>     has been used as the computing platform. It would be very
>     interesting to receive contributions from people outside the
>     circle of package authors.
>
> Below follows a list of suggested contributions. This is just from the
> top of my head so consider the titles/authors a rough initial value
> for further iteration.
>
>   Introduction: Logg (+ others?)
>   Tutorial: Langtangen
>
>   Methodology:
>
>     Finite element code generation: Alnaes, Logg, Mardal
>     Finite element assembly: Logg + others
>     Optimal evaluation of finite element variational forms: Kirby, Scott
>     Efficient finite element mesh representation: Karpeev, Knepley, Logg?
>     Computing finite element basis functions: Kirby, Mardal
>     A gallery of finite elements: Kirby? Terrel? Rognes? Mardal? Logg?
>     Stabilized Galerkin: Hoffman, Johnson
>
>   Implementation
>
>     DOLFIN: Logg, Wells, Oelgaard
>     FErari: Kirby
>     FFC: Logg, Oelgaard, Rognes, Wells
>     FIAT: Kirby
>     Instant: Mardal
>     SyFi: Mardal, Alnaes
>     UFC: Alnaes, Logg, Mardal, Skavhaug, Langtangen
>     UFL: Alnaes, Logg, Mardal
>     Unicorn: Hoffman, Jansson
>     Viper: Skavhaug
>
>     Generic linear algebra interfaces: Alnaes, Logg, Mardal, Skavhaug,
> Wells
>     A note on build systems: Knepley, Odegard
>     Gotran: Skavhaug
>     Exterior calculus: Logg, Mardal
>     Sieve?
>
>   Application
>
>     Contributions from CBC
>     Contributions from KTH
>     Contributions from Umeå
>     Contributions from Argonne?
>
>     Something on inverse problems: Myklebust, Mardal, Nielsen
>     Something on flow problems: Valen-Sendstad, Logg, Mardal
>     Something on electro-mechanics: Alnaes, Sundnes
>     Something on FSI: Selim
>
>     Other contributors: Narayanan, Schroll, Lines, Ostien, Lindbo,
>     Hake, Brown, Nazarov, Lopes, Wilbers, Voigt, Kaasbjerg, etc
>
>   Appendix
>
>     Summary of notation (Logg)
>     UFC reference cells
>     UFC numbering of mesh entities
>     Something about Scons?
>     Something about SWIG?
>
> PS: Further discussion of this topic will take place on the fenics-dev
> mailing list. So if you are interested in following the discussion,
> please sign up at http://www.fenics.org/mailman/listinfo/fenics-dev
>
> --
> Anders
> _______________________________________________
> FEniCS-dev mailing list
> FEniCS-dev@xxxxxxxxxx
> http://www.fenics.org/mailman/listinfo/fenics-dev
>