Sparx Systems Forum

Enterprise Architect => Suggestions and Requests => Topic started by: skiwi on May 05, 2020, 01:20:36 pm

Title: Requirement priority
Post by: skiwi on May 05, 2020, 01:20:36 pm
We use RFC RFC 2119 keywords to set requirement levels.
In EA the priority field has some predefined values.
I know this have been canvassed way back in the day from a search I did,
but is it possible to specify a project or model default that is different from the predefined values?

tia
Title: Re: Requirement priority
Post by: Eve on May 05, 2020, 01:48:30 pm
Configure > Reference Data > Model Types > General Types > Priority.
Title: Re: Requirement priority
Post by: skiwi on May 05, 2020, 02:35:38 pm
Oh spendid, thankyou very much
Title: Re: Requirement priority
Post by: qwerty on May 05, 2020, 03:45:51 pm
Did you think about creating your own requirement element? Adding useful stuff to tags. Along with MOSCOW (or whatever you prefer).

q.
Title: Re: Requirement priority
Post by: Paolo F Cantoni on May 05, 2020, 04:53:11 pm
Did you think about creating your own requirement element? Adding useful stuff to tags. Along with MOSCOW (or whatever you prefer).

q.
That's what we did...

Paolo
Title: Re: Requirement priority
Post by: Sunshine on May 07, 2020, 03:30:19 pm
Didn't even know that RFC 2119 existed. Published back in March 1997.
https://tools.ietf.org/html/rfc2119 (https://tools.ietf.org/html/rfc2119)
Is there nothing more recent?
Title: Re: Requirement priority
Post by: Paolo F Cantoni on May 07, 2020, 06:16:36 pm
Didn't even know that RFC 2119 existed. Published back in March 1997.
https://tools.ietf.org/html/rfc2119 (https://tools.ietf.org/html/rfc2119)
Is there nothing more recent?
Well, we "old-timers" were around then...  ;)

Paolo
Title: Re: Requirement priority
Post by: Uffe on May 07, 2020, 06:18:13 pm
Didn't even know that RFC 2119 existed. Published back in March 1997.
https://tools.ietf.org/html/rfc2119 (https://tools.ietf.org/html/rfc2119)
Is there nothing more recent?
Well, there's https://tools.ietf.org/html/rfc8174 (https://tools.ietf.org/html/rfc8174).

 ;D
Title: Re: Requirement priority
Post by: qwerty on May 07, 2020, 06:37:55 pm
Or basically you can stick to Wittgenstein's  Tractatus logico-philosophicus.

q.
Title: Re: Requirement priority
Post by: Paolo F Cantoni on May 07, 2020, 11:55:53 pm
Or basically you can stick to Wittgenstein's  Tractatus logico-philosophicus.

q.
I manage to get some way in, but I had to give up.  TOO dense...

Paolo
Title: Re: Requirement priority
Post by: Uffe on May 08, 2020, 12:04:57 am
The book, or.....?  ;D
Title: Re: Requirement priority
Post by: qwerty on May 08, 2020, 12:31:12 am
The reason why we're using UML is because words don't match their semantics. But what I see is that UML models are also not much better in making a connection. UML is just another language. But somehow we must feed all the lawyers being involved with requirements management.

q.
Title: Re: Requirement priority
Post by: skiwi on May 08, 2020, 06:17:42 am
Didn't even know that RFC 2119 existed. Published back in March 1997.
https://tools.ietf.org/html/rfc2119 (https://tools.ietf.org/html/rfc2119)
Is there nothing more recent?
Well, there's https://tools.ietf.org/html/rfc8174 (https://tools.ietf.org/html/rfc8174).

 ;D

and https://tools.ietf.org/html/rfc6919 (lol)
Title: Re: Requirement priority
Post by: skiwi on May 08, 2020, 06:19:47 am
Didn't even know that RFC 2119 existed. Published back in March 1997.
https://tools.ietf.org/html/rfc2119 (https://tools.ietf.org/html/rfc2119)
Is there nothing more recent?

Does their need to be?
The fact that this is well defined, widely used, makes it a better fit than some sort of MOSCOW varient, or an organisation specific 1 2 3 4, or low med hi IMHO
Title: Re: Requirement priority
Post by: Paolo F Cantoni on May 08, 2020, 08:22:35 am
Didn't even know that RFC 2119 existed. Published back in March 1997.
https://tools.ietf.org/html/rfc2119 (https://tools.ietf.org/html/rfc2119)
Is there nothing more recent?
Well, there's https://tools.ietf.org/html/rfc8174 (https://tools.ietf.org/html/rfc8174).

 ;D

and https://tools.ietf.org/html/rfc6919 (lol)
RFC6919 is very cute.

But on a more thoughtful note, and somewhat related to RFC6919 is the notion of the evolution of Desired to Actual requirements.  It seems to me that what is ALWAYS missing is...  "This is the desired set of requirements we went to RFP with" and "This is the actual set of requirements we used to finally select the successful proposal."  There is always a (usually undocumented) change in the requirements from commencement to the termination.  Understanding THAT process (and documenting WHY) is crucial, in my view.

Paolo
Title: Re: Requirement priority
Post by: Sunshine on May 08, 2020, 03:33:45 pm
...
Does their need to be?
The fact that this is well defined, widely used, makes it a better fit than some sort of MOSCOW varient, or an organisation specific 1 2 3 4, or low med hi IMHO
Yeah good point. Keeping it simple is always a good tactic.
Must admit the "W" part of MOSCOW always baffled me. Why on earth would you say want you don't want it to do. It could include everything in the universe including Won't tell in appropriate jokes :)
Title: Re: Requirement priority
Post by: Uffe on May 09, 2020, 07:32:23 pm
Must admit the "W" part of MOSCOW always baffled me. Why on earth would you say want you don't want it to do. It could include everything in the universe including Won't tell in appropriate jokes :)
Not right out of the gate, no. But at a later stage, if a certain feature has been agreed or is otherwise expected, but won't in fact be built (ever, or just in this release) -- due to renegotiations, time or budget constraints, or whatever -- it makes sense to explicitly say so, and it makes sense to have standardized language for that situation too.

/U
Title: Re: Requirement priority
Post by: Paolo F Cantoni on May 11, 2020, 08:22:07 am
Must admit the "W" part of MOSCOW always baffled me. Why on earth would you say want you don't want it to do. It could include everything in the universe including Won't tell inappropriate jokes :)
Not right out of the gate, no. But at a later stage, if a certain feature has been agreed or is otherwise expected, but won't, in fact, be built (ever, or just in this release) -- due to renegotiations, time or budget constraints, or whatever -- it makes sense to explicitly say so, and it makes sense to have standardized language for that situation too.

/U
You can also use a modified "W" to require that some features NOT be created.  Something along the lines of "just because you can, doesn't mean you should".  I'm a great believer in saying what systems must not do as well as what they can/should. 

Paolo
Title: Re: Requirement priority
Post by: skiwi on May 11, 2020, 08:52:36 am
I'm a great believer in saying what systems must not do as well as what they can/should. 

Paolo

I agree with that, it is important to capture decisions about what is out of scope, and why.
I do that in the scope section not in the requrements (although if I did I could see the need for requirements levels such as: not needed, too expensive, don't do this, too modern, costs jobs, whatever you do don't do this; for example)
Title: Re: Requirement priority
Post by: Paolo F Cantoni on May 11, 2020, 09:56:10 am
I'm a great believer in saying what systems must not do as well as what they can/should. 

Paolo

I agree with that, it is important to capture decisions about what is out of scope, and why.
I do that in the scope section, not in the requirements (although if I did I could see the need for requirements levels such as: not needed, too expensive, don't do this, too modern, costs jobs, whatever you do don't do this; for example)
As in "Whatever you do, don' smoke" - Yul Brynner   ;)

Paolo