dolfin team mailing list archive

[Johan Hake <hake.dev@xxxxxxxxx>]

On 09/10/2012 10:55 AM, Garth N. Wells wrote:
> On Mon, Sep 10, 2012 at 9:45 AM, Anders Logg <logg@xxxxxxxxx> wrote:
>> On Sat, Sep 08, 2012 at 11:16:49PM +0200, Johan Hake wrote:
>>>    On Sep 8, 2012 12:04 PM, <[1]noreply@xxxxxxxxxxxxx> wrote:
>>>    >
>>>    > ------------------------------------------------------------
>>>    > revno: 6896
>>>    > committer: Garth N. Wells <[2]gnw20@xxxxxxxxx>
>>>    > branch nick: assembler
>>>    > timestamp: Sat 2012-09-08 10:49:23 +0100
>>>    > message:
>>>    >   Start cleaning up assemblers.
>>>    >
>>>    >   The assembler classes are no longer full of static member functions
>>>    (this was pointless because we have free function for easy access) and
>>>    the host of optional boolean arguments have been removed from the
>>>    member function interfaces and made part of a common base class.
>>>    >
>>>    >   Simple usuage remains unchanged. For more advanced usage,
>>>    FooAssembler object should be created and the boolean options set via
>>>    >
>>>    >     assmebler.reset_tensor = false;
>>>    >
>>>    >   etc. This should be much more intelligible and less error prone.
>>>    > renamed:
>>>
>>>    Nice!
>>
>> Yes nice, but the parameter system should be used as for other classes:
>>
>>   assembler.parameters["reset_tensor"] = false;

++

> I don't think so. There is no advantage to using parameters in this
> case. It just adds complexity.

What complexity more than having to deal with an extra parameters type
instead of the bools attributes?

The whole thing with parameters attached to objects is that these can
then be nested into other parameters easily.

Johan

> Garth
> 
>> --
>> Anders
>>
>>
>>>    Johan
>>>
>>>    >   dolfin/fem/AssemblerTools.cpp => dolfin/fem/AssemblerBase.cpp
>>>    >   dolfin/fem/AssemblerTools.h => dolfin/fem/AssemblerBase.h
>>>    > modified:
>>>    >   demo/undocumented/periodic/cpp/main.cpp
>>>    >   demo/undocumented/smoothing/python/demo_smoothing.py
>>>    >   dolfin/ale/HarmonicSmoothing.cpp
>>>    >   dolfin/fem/Assembler.cpp
>>>    >   dolfin/fem/Assembler.h
>>>    >   dolfin/fem/LinearVariationalSolver.cpp
>>>    >   dolfin/fem/OpenMpAssembler.cpp
>>>    >   dolfin/fem/OpenMpAssembler.h
>>>    >   dolfin/fem/SymmetricAssembler.cpp
>>>    >   dolfin/fem/SymmetricAssembler.h
>>>    >   dolfin/fem/SystemAssembler.cpp
>>>    >   dolfin/fem/SystemAssembler.h
>>>    >   dolfin/fem/assemble.cpp
>>>    >   dolfin/fem/dolfin_fem.h
>>>    >   dolfin/swig/modules/fem/dependencies.txt
>>>    >   dolfin/swig/modules/fem/module.i
>>>    >   site-packages/dolfin/compilemodules/swigimportinfo.py
>>>    >   dolfin/fem/AssemblerBase.cpp
>>>    >   dolfin/fem/AssemblerBase.h
>>>    > The size of the diff (1283 lines) is larger than your specified limit
>>>    of 500 lines
>>>    >
>>>    >
>>>    > Your team DOLFIN Core Team is subscribed to branch lp:dolfin.
>>>    > To unsubscribe from this branch go to
>>>    [4]https://code.launchpad.net/~dolfin-core/dolfin/trunk/+edit-subscript
>>>    ion
>>>
>>> Referenser
>>>
>>>    1. mailto:noreply@xxxxxxxxxxxxx
>>>    2. mailto:gnw20@xxxxxxxxx
>>>    3. https://code.launchpad.net/~dolfin-core/dolfin/trunk
>>>    4. https://code.launchpad.net/~dolfin-core/dolfin/trunk/+edit-subscription
>>
>>> _______________________________________________
>>> Mailing list: https://launchpad.net/~dolfin
>>> Post to     : dolfin@xxxxxxxxxxxxxxxxxxx
>>> Unsubscribe : https://launchpad.net/~dolfin
>>> More help   : https://help.launchpad.net/ListHelp
>>
>>
>> _______________________________________________
>> Mailing list: https://launchpad.net/~dolfin
>> Post to     : dolfin@xxxxxxxxxxxxxxxxxxx
>> Unsubscribe : https://launchpad.net/~dolfin
>> More help   : https://help.launchpad.net/ListHelp