home > plop > 2006 > Papers > Notes > 49-notes.php

PLoP 2006 Shepherded Papers

Notes on Paper 49: ACCEPTED

here's my review of Mirko Raner's submission "The Mutator Pattern" (plop2006-mraner0). I recommend the paper for workshopping at PLoP. My detailed reasoning follows below.

On 7/29/06, Joseph W. Yoder <joe@joeyoder.com> wrote:
> The criteria for paper acceptance are:
> The quality and wholeness of the pattern(s) or pattern
> language.

Mirko proposes a pattern called the "Mutator". Mutators, the central entity of this pattern, unify sequential access to a series of elements, that are only implicitly given e.g. by a generating function. Iterator and Mutator complement each other. Iterator has the context of an explicitly given sequence, most often a Composite that hides its structure, whereas the context of Mutator is a sequence that is only implicitly given. Mutators have a similar interface as Iterators, but offer only a weaker contract to the client than Iterators do. In this sense Mutators are generalizations of Iterators. The Mutator pattern is useful by itself, but contrasting Mutator and Iterator also helps understanding Iterator better.

The pattern description mimics GoF style. It comes up with extensive and two detailed examples in Java. The paper gives both pros and cons of the application of the pattern and even shows in one of its examples its limitations. Evidence of the pattern is backed up with both the domain of testing and genetic algorithms.

I think "Mutator" qualifies as a pattern on its own.

> Degree of improvement during the shepherding process, and

We managed to go through three iterations of shepherding.

Mirko improved the pattern description during shepherding. Unimportant detail was removed, scaffolding was reduced. The resulting paper is better tailored to the expectied audience than the first version.

> openness of the authors to
> revisions; and

Mirko accepted nearly all of my suggestions. He always came up with a new draft, not only with announcements of potential changes to his submission.

> Relevance of the pattern or pattern language, based on the
> opinions of the program committee

It's up to you and your collegues to judge this issue...

You'll find Mirko's latest version attached, which is also the version my judgement is based upon.
 

To keep up on the latest
PLoP information, subscribe to:
plop-announce-subscribe@hillside.net.