[Dune] Returning Geometries As Objects

[Carsten Gräser]

Hi,

Am 07.02.2012 11:14, schrieb Martin Nolte:
> Hi Christian,
> 
> first of all, your objection is noted and I will not merge the branch
> before the issue is decided upon.
> 
> However, I think this issue does not just concern the grid implementors.
> It might also have a performance impact on existing codes. Currently, it
> is usually ok to fetch the geometry multiple times, because the entity
> can ensure that it is built up only once.
> 
> Therefore, I think that a formal vote on the lifetime issue is in order
> (that's why I wanted to postpone it until the next meeting).
> 
> Since you are pressing the issue, I think we should split up the issues
> a little. For example:
> 
> (a) Should we postpone the merge?
> (b) Should we decide on the lifetime issue now or later?
> (c) May the lifetime be limited (by entity / intersection) or not?
> (Maybe capability?)
> (d) Should the lifetime issue be a showstopper for the next release
> (i.e., no transition phase)?
> (e) Should we make another release before we merge?
> 
> I think postponing the merge will increase maintenance costs and slow
> down development, so I'd like the merge to happen soon. Moreover, I also
> think there will be a lively discussion of (c), so I don't know whether
> deciding now makes sense.
> 
> What are the opionions of the other developers?
I also consider the branch tested well enough to be merged.
However, the crucial point is (b)/(c):

I'd not like to have objects that behave like references
and I do totally agree with Oliver on the potential bad
consequences.

On the other hand I'm afraid that postponing the merge
until after a discussion on the next meeting would mean
a significant amount of extra work for Martin who has
already invested much time to implement this.

@(b): I propose to decide on (c) now
@(c): I'd expect geometry objects to have the same
      life time as EntityPointers now, i.e., until
      the next grid modification (*)
If we decide to have (*) I'd furthermore like to...

@(a): ... do the branch now to avoid extra work.
@(d): ... implement (*) before the branch even if one
      would have the adapt the application code to get
      the same performance (this has to be done anyway).
      If there are serious concerns about this we
      might think of making this adjustable via
      some macro switch.
@(e): ... not do have a release before the branch in
      view of the extra work for testing things we
      already consider outdated.

Best,
Carsten

> 
> Best,
> 
> Martin
> 
> 
> 
> On 02/07/2012 10:50 AM, Christian Engwer wrote:
>> Hi Martin,
>>
>> just as a follow up of my last mail.
>>
>>> Currently, my impression is as follows: Carsten and Oli would like
>>> to enhance the lifetime of the geometry objects, Robert agrees to
>>> merge, and some users are afraid of breaking their code. Of course,
>>> the positive decision on the last developer meeting is positive
>>> towards this change.
>>>
>>> *NOTICE*: This is the final announcement. Unless a developer
>>> objects, the branch will be merged 2 PM today!
>>
>> I'm against a merge, unless we agree to disable the reference feature
>> before the next release. Yes, this will require updates in all grid
>> implementations, but I think it is necessary to make this change show
>> its full potential and to actually value all the work and time you
>> spent here.
>>
>> If we can agree on this, I have no furthe objections regarding a
>> merge. I think it is well enough tested to actually do a merge. My
>> main concern is about adding undesired features which we (might) have
>> to deal with for an unforseable time.
>>
>> And "we agree" does not necessarily require a formal vote. I think it
>> just means that those people involved in (core) grid implementations
>> have to agree. I, for my part, am willing to do add the necessary
>> changes. Now we only need an opinion from Robert and Oliver [assuming
>> that you actually want to use you new feature ;-)].
>>
>> Cheers
>> Christian

-- 
----------------------------------------------------------------------
Dr. Carsten Gräser       | phone: +49-30 / 838-75349
Freie Universität Berlin | fax  : +49-30 / 838-54977
Institut für Mathematik  | email: graeser at math.fu-berlin.de
Arnimallee 6             |
14195 Berlin, Germany    | URL  : http://page.mi.fu-berlin.de/graeser
----------------------------------------------------------------------