Book a Demo

Author Topic: RUP  (Read 9072 times)

Lolke

  • EA Novice
  • *
  • Posts: 2
  • Karma: +0/-0
    • View Profile
RUP
« on: July 28, 2004, 07:10:24 am »
Now that Rational and the RUP are widely used throughout industry, you probably are keeping a close watch.

Do you plan to follow the RUP guidelines on (as an example) supporting Business Modeling? In RUP as you know they are proposing a proprietary (different from EA) notation and method for supporting business processes: I.e. Business Use Cases, Business Actors, etc.

What is your opinion on this?

sargasso

  • EA Practitioner
  • ***
  • Posts: 1406
  • Karma: +1/-2
  • 10 COMFROM 30; 20 HALT; 30 ONSUB(50,90,10)
    • View Profile
Re: RUP
« Reply #1 on: August 04, 2004, 04:56:07 pm »
AFAIK the elements you mention are the rather poorly defined stereotypes from the original UML 1.2 Profiles spec.

IMO Erikkson-Penker business modelling and the profiles available from Sparx are an order of magnitude more powerful than the simple OMG profile.  Furthermore, the visual similarity of the OMG elements to structural design elements - used for dynamic modelling -  is confusing at best.

B

(I think that's the topic you wanted an opinion on - not my opinion on Rational, the ex-company and RUP, the comercialisation of a public domain technique/process model  ;D )
"It is not so expressed, but what of that?
'Twere good you do so much for charity."

Oh I forgot, we aren't doing him are we.

Lolke

  • EA Novice
  • *
  • Posts: 2
  • Karma: +0/-0
    • View Profile
Re: RUP
« Reply #2 on: August 05, 2004, 12:12:12 pm »
This is not the point however. As a consultant I am working for larger companies and forced to work with  RUP because they have dominance in the marketplace. I do not particularly favour XDE (rather than for instance EA), but RUP as a methodology is widely supported by a very large community. The point is I would rather have a good CASE tool that supports RUP than the current XDE/Rational tools.
Looking at it from the other side: If a tool does not support RUP it is simply excluded for use with many many companies.

mchiuminatto

  • EA User
  • **
  • Posts: 113
  • Karma: +0/-0
    • View Profile
Re: RUP
« Reply #3 on: August 06, 2004, 04:57:59 pm »
Hi

EA can support any UML extension. RUP support for Business Modeling is an UML extension, in other words, classifiers with specific stereotypes adequate to business process and an extended graphic notation. There's a couple of ways you can define an UML extension with EA. The most simple is adding your own stereotypes to the model; for instance, you can stereotype a use case as a business use case. But if you want support graphic notation besides, you can define a profile with the required UML extension.  

SparxSystems site has some profiles you can download: Eriksson-Penker, as sargasso mentioned, Business Process, derived from UML 1.4. among others.


I’m a software engineering process consultant; my company promotes RUP as the process and we have adopted EA as the UML modeling tool that we deliver with our solutions.



Regards

Marcello
Regards.

Marcello

dwscott

  • EA Novice
  • *
  • Posts: 1
  • Karma: +0/-0
    • View Profile
Re: RUP
« Reply #4 on: August 09, 2004, 06:22:37 am »
Business modeling with RUP is brutal, very confusing to end users and non-standard with respect to the real standards body for business modeling (BPMI.Org). I would much rather see EA adopt the standard in this area (BPMN - Business Process Management Notation).

SpoonsJTD

  • EA User
  • **
  • Posts: 39
  • Karma: +0/-0
    • View Profile
Re: RUP
« Reply #5 on: August 13, 2004, 12:03:52 pm »
Quote
I would much rather see EA adopt the standard in this area (BPMN - Business Process Management Notation).


Which would be (will be?) a lot easier if (when?) the BPMN group creates a UML profile for BPMN.

SamJolly

  • EA User
  • **
  • Posts: 51
  • Karma: +0/-0
    • View Profile
Re: RUP
« Reply #6 on: August 16, 2004, 09:26:51 am »
This is a really interesting point on 2 counts:

a) I have recently scanned the list of Architect jobs, and of course they all require knowledge of Rational/RUP. So should EA only be useld by those enlightened software developers building their own solutions.

b) Regarding worflow very interesting point about BPMN. I do not know much about profiles as I am a relative novice, but the provision of ready made profiles by Sparx, perhaps created in association with a standards body would be really proactive. A RUP profile would also be included in this.

Sam

sargasso

  • EA Practitioner
  • ***
  • Posts: 1406
  • Karma: +1/-2
  • 10 COMFROM 30; 20 HALT; 30 ONSUB(50,90,10)
    • View Profile
Re: RUP
« Reply #7 on: August 17, 2004, 04:53:02 pm »
Horses for courses.

When I (very repeatedly) have to explain how UML RUP EA Rose XDE etc fit together into a system design "factory" I use the following model.

There are 4 components that contribute to the success of my factory.  These are:
  • how we manage the factory
  • the process we use to convert a request into product
  • the techniques we use to design and build the product
  • the tools we use to develop the design and to construct the product


Of these I will ignore the management component for the minute.

UML is a technique - it is a notation that let's skilled designers describe accurately the design of the product.  It's just like the symbols used in an engineering design or an electronic design or a house design.  UML can be done on a piece of paper with a crayon, on a whiteboard with a coloured marker or on a PC screen using a.....

tool such as EA, Rose, XDE, poseidon, etc etc etc etc etc.  The tools let us construct these designs quickly and efficiently - and to some extent accurately within the "rules" for use of the notation.  In addition, most of them allow for the production of the design artifacts in media suitable for communicating the design....

and arriving at an agreement with the stakeholders on its suitability through a process such as RUP, Iconix, OOSE etc etc etc.  Such processes define where specific information is to be found as well as the "steps" we go through to ensure the design.

So, forgive me, but I find this argument tiresome.  There is no EA is better/worse than RUP because its like asking whether a hammer is better than a recipe for cannelloni.  Is BPMN better/worse than EA?  Or is impressionist painting better than a food processor?

What do we (here) preach?  
1) flexible and pragmatic use of UML as the notation for all system design - from single developer, single user systems through to enterprise solutions.
2) island solution, exploratory (paper prototype) design and component solution projects use EA as the tool, projects and programs over $2M use (guess which) heavyweight tools (only) in order to take advantage of existing infrastructure support.
3) the process is totally and entirely dependent on the situation. In general we follow an Iconix variant because I personally like the use of collaboration models as a formal step in moving from the business/logical view into the design views.  If the situation is big enough to warrant the use of a formal development process (our yardstick is quite complex but lets say "over $1.2M") then the RUP (actually the UP) framework is used as the basis for developing a project specific process.

Managing the game is quite another topic - just as the "rules" for managing a footbal team are quite different to the rules of the game itself.

So "should EA only be useld by those enlightened software developers building their own solutions"? Piffle!


.. and that's all I've got to say about that  ;)

Bruce
"It is not so expressed, but what of that?
'Twere good you do so much for charity."

Oh I forgot, we aren't doing him are we.

TrtnJohn

  • EA User
  • **
  • Posts: 176
  • Karma: +0/-0
    • View Profile
Re: RUP
« Reply #8 on: August 18, 2004, 01:17:25 pm »
Piffle to you too, Bruce  ;D

I am also new to these design process/frameworks and am looking to implement some of their best pieces into a process we have already working.  (Or more accurately limping along).   Which do you recommend to look at first?  Any good books to recommend?

Edit:

I should have added we are a group of software developers that work on projects that are fairly short in duration. (6 - 18 months of effort).  We usually work in small teams of 3 - 15 developers/designers.  But, the team can be spread out across a couple of different locations.  (Even countries in some cases).
« Last Edit: August 18, 2004, 01:20:54 pm by TrtnJohn »

sargasso

  • EA Practitioner
  • ***
  • Posts: 1406
  • Karma: +1/-2
  • 10 COMFROM 30; 20 HALT; 30 ONSUB(50,90,10)
    • View Profile
Re: RUP
« Reply #9 on: August 18, 2004, 04:33:39 pm »
If you are looking at des/dev process engineering then the lead article by Sinan Si Alhir in http://www.methodsandtools.com/PDF/dmt0102.pdf has some great foundational material - especially diagram 6 and page 14 on "Focus" and "Balance".

hth
Bruce


As a matter of fact the second article (Sassenberg) might be worth a look too.


As far as books are concerned  - I favour Rosenberg & Scott, Ambler, and Fowler (with caution).
« Last Edit: August 18, 2004, 04:45:16 pm by sargasso »
"It is not so expressed, but what of that?
'Twere good you do so much for charity."

Oh I forgot, we aren't doing him are we.

Gary W.

  • EA User
  • **
  • Posts: 139
  • Karma: +0/-0
    • View Profile
Re: RUP
« Reply #10 on: September 14, 2004, 11:04:28 am »
Hi,

>  As far as books are concerned  - I favour Rosenberg &
> Scott, Ambler, and Fowler (with caution).

Okay, having read these authors, and agreeing with your recommendation, I have to ask why the 'with caution' caveat on Martin Fowler?

Cheers,
Gary

sargasso

  • EA Practitioner
  • ***
  • Posts: 1406
  • Karma: +1/-2
  • 10 COMFROM 30; 20 HALT; 30 ONSUB(50,90,10)
    • View Profile
Re: RUP
« Reply #11 on: September 14, 2004, 04:20:19 pm »
'cause I'm not a big fan of the agile so-called method.  

Although some of the ideas on pragmatism in modelling and the idea of focussing on the delivery of systems rather than procedure checklist ticking are appealing, there are IMO fundamental flaws in the overall philosophy.

One does not embark on a single handed around the world yacht race by throwing a bathtub into the sea, jumping in and then when it sinks "refactoring" it.

N.B. The topic has been adequately explored elsewhere (principally by Rosenberg etal).

I'm just saying I'm on the not-so-agile-it-sinks side of the fence.

Bruce
"It is not so expressed, but what of that?
'Twere good you do so much for charity."

Oh I forgot, we aren't doing him are we.

Gary W.

  • EA User
  • **
  • Posts: 139
  • Karma: +0/-0
    • View Profile
Re: RUP
« Reply #12 on: September 16, 2004, 01:42:59 pm »
Bruce,

Thanks.. I was just curious.  Of course, Ambler is just as, if not more, focused on Agile Methods than Fowler is.

Gary