Jump to content

Talk:Unified process

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
(Redirected from Talk:Unified Process)

UP verses RUP

[edit]

I think the Rational Unified Process article needs to be refactored. However, before tackling that more ambitious project I decided to create this small "Unified Process" article to try to explain (to the extent that anyone understands them) the differences between RUP and UP. My intent is not to supplant the RUP article. I think the RUP article should be the main article describing the process. More has been written about RUP than about all other "versions" of UP combined. My intent is that this UP article should explain what is common among all the different "versions" of UP and that it should provide links to the various related articles. --GFLewis 18:28, 4 July 2006 (UTC)[reply]

Comment by FredThwaites

[edit]

I have a significant problem with this article. I'm not sure that the Unified Process exists in any concrete sense.

When the three amigos were working on unifying their various OO methods they decided to separate off the process aspects from the modeling language hence UML and RUP. While they strove to keep UML non-proprietary, the RUP was from the outset a Rational product. From a chronological point of view the RUP was thus the original UP rather than an extension of it.

I can see that in developing other related processes the authors outside of Rational and now IBM, have strived to keep away from infringing the RUP copyrights by 'de-Rationalising' it, and so refer to it obliquely as UP.

I also note that within Rational and the books authored by them they tend to refer to RUP as UP themselves, further muddying the water.

While I can understand that both the RUP and subsequent derivatives share some common aspects, I think that your article in some ways tries to make this abstraction more concrete that it really is. FredThwaites 23:41, 15 July 2006 (UTC)[reply]

I agree that these waters are very muddy. I also agree that in practice it is generally impossible to distinguish RUP from UP. Since both are customizable process frameworks, and since organizations always remove the elements they consider non-essential and add other elements they feel are missing, who can ever say what the starting point was? The problem is aggravated by the fact that RUP and UP are both moving targets. RUP, at least, is well defined. With each release of the product IBM lists the changes from the prior version. But what is UP? Is it the original process described in the 1999 book by Jacobson, Booch and Rumbaugh? Or is UP the collective consensus of those who claim to follow it? Does UP, like RUP, continued to evolve as authors publish books and articles under the UP appellation?
In spite of these difficulties, I think there are two arguments which favor a separate article for UP.
  • There are several authoritative sources (not the least of which is Philippe Kruchten, himself) that define RUP as a specialization or instance of the generic Unified Process. Please see my comments under Talk:Rational Unified Process#UP verses RUP. This implies that RUP is different from UP. I have not (yet) found any authoritative source that says that RUP and UP are the same thing.
  • RUP is different from UP in the sense that the RUP software product from IBM describes and documents the process in excruciating detail. If this were not so then IBM would not be able to charge money for the RUP product (which is basically just a customizable knowledgebase). This is what Larman and other authors mean when they say that RUP is a refinement of UP. From a Wikipedia perspective this provides an opportunity to use the UP article to describe the general or common characteristics of the process, while enabling the RUP article to describe RUP specific elements (such as the division of work into 9 disciplines) which are not a characteristic of most refinements.
--GFLewis 02:37, 1 August 2006 (UTC)[reply]
[edit]

The Unified Process name is also used to avoid potential issues of copyright infringement since Rational Unified Process and RUP are trademarks of IBM.

Should this instead read "patent infringement"? —The preceding unsigned comment was added by 68.193.240.87 (talk) 00:58, 6 March 2007 (UTC).[reply]

Problem with Iterative Development diagram

[edit]

Iterative Development Diagram

The Iterative Development Diagram on this page (the so-called RUP 'hump chart') uses RUP terminology and is based on the RUP disciplines rather than the UP workflows - this is confusing for the reader because UP and RUP use different terminology and have different numbers of core activities (workflows in UP, disciplines in RUP).

As originally specified in Jacobson's book, which is the definitive reference, UP has 5 core workflows:

  • Requirements
  • Analysis
  • Design
  • Implementation
  • Test

whereas RUP has 6 disciplines, namely:

  • Business modelling
  • Requirements
  • Analysis and design
  • Implementation
  • Test
  • Deployment

There is a UP version of the RUP 'hump chart' in Jacobson's book (Fig 1.5), which is based on the UP workflows rather than the RUP disciplines, so I think that WIkipedia needs to have two versions of this diagram, one for UP and one for RUP.

Nrjs (talk) 16:22, 27 September 2010 (UTC)[reply]

Elicitation phase?

[edit]

Does elicitation phase belong to UP? It is not referenced in any other place than a single reference to a 3-bullet list with no explanation and context. Should we remove it? Rentzepopoulos (talk) 14:43, 17 May 2012 (UTC)[reply]

After reviewing the update history of the article, where there is no support for the inclusion of the elicitation phase, I remove it completely. Rentzepopoulos (talk) 14:47, 17 May 2012 (UTC)[reply]

Usecase modeling

[edit]

Explain clearly 61.2.46.42 (talk) 05:15, 16 December 2022 (UTC)[reply]