Juergen:

Thanks for the high quality and pragmatic response.  Yes, I did mean UML
Class diagrams.  As the teens say "my bad".  

As to the tools and code generation:
During my days of writing protocol software or managing protocol software
writers, I found it often took 3+ rewrites to get efficient code after
working good.   If tools create inefficient code but working code - at least
a prototype to test against comes up quickly.  If tools improve the
prototype may get closer and the re-writes less. 

As to descending into detail with UML:  that's actually a plus for me.  The
purpose behind UML is to quicken the pace of the process from high level
agreement to DM.  We can standardize the high level and then use the UML to
provide quicken layers of agreement.  

Sue 

 

-----Original Message-----
From: i2rs [mailto:[email protected]] On Behalf Of Juergen Schoenwaelder
Sent: Wednesday, April 30, 2014 1:49 PM
To: Susan Hares
Cc: 'Nitin Bahadur'; 'Joel Halpern Direct'; 'Mach Chen';
[email protected]; [email protected]
Subject: Re: [i2rs] Some comments on draft-ietf-i2rs-rib-info-model-01

On Wed, Apr 30, 2014 at 10:49:21AM -0400, Susan Hares wrote:
 
> I'm trying to get a workable alternative to RBNF to discuss the actual 
> im-tree.  UML is best, but I'm not sure If I can it in the documents.  
> 5 pages of UML covers all the RIB models.  UML tools can create yang data
> models and forces (In my understanding).    However, some human beings
seem
> to need the yang tree models to see the tree and links to yang. 

I assume people mean UML class diagrams when they say UML. That said, yes
tools can turn UML into other stuff. However, whether the result is 'nice'
or 'implementable are reasonable pain' is a different question (and I
understand that 'nice' and 'pain' have no clear definition). UML class
diagrams allow the representation of arbitrary relationships between
classes. If you have to turn this into a hierarchy for a protocol using a
hierarchical namespace, then tools most likely only create _a_ solution not
a _good_ solution.

If you believe tools do a good job translating IMs to DMs, you will likely
soon find that this is generally not the case and in order to make it work
better you will start to augment you IM with lots of additional detail for
the translator to do a reasonably good job, in which case your IM
deteriorates to some extend into a DM.

/js

-- 
Juergen Schoenwaelder           Jacobs University Bremen gGmbH
Phone: +49 421 200 3587         Campus Ring 1, 28759 Bremen, Germany
Fax:   +49 421 200 3103         <http://www.jacobs-university.de/>

_______________________________________________
i2rs mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/i2rs

_______________________________________________
i2rs mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/i2rs

Reply via email to