OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

regrep message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [regrep] Final draft recommendations for CC's and Registry -UML/UMM Profile for CCTS




Anders W. Tell wrote:

> The statements such as "A machine cannot parse the internal structure 
> of a *.jpg diagram."
> "UML is a two dimensional syntax concerned with rendering (upon which 
> implications of structure can be interpreted by a human viewing the 
> UML) and cannot be parsed accurately by an application"
> ....did throw me of track.

The main point is that what you see with your eyes (boxes and lines) and 
how an application sees something (1's and 0's) is quite different.  I 
have had people tell me that they see applications parse UML and render 
it as java code or XMI.  They are often fooled by their eyes and don't 
understand the concept that is working behind the scenes.  What is 
really happening is some binary (and often proprietary) persistence 
mechanism is being used as the basis to generate both the java code or 
XMI  and the UML diagram they see on their screen.  XMI would be 
notoriously innefficient as a storage mechanism for UML projects.  Its' 
use as such is noted along with the shortcomings of it concerned with 
graphical representation (not an issue for the subject we are discussing).

>
> As well as: "We have successfully demonstrated it for machine 
> automated assembly and deriviation of BIE's from a set of bound core 
> components and a context declaration statement.  No one else has done 
> that yet to my knowledge. "
>
> For you info this can  be done with a UML profile for CCTS, which is 
> what we have created.

Can you send the UML as an attachment to me in a separate email please?

Duane

>
> /anders
>
>>
>> ;-)
>>
>> Anders W. Tell wrote:
>>
>>> Duane Nickull wrote:
>>>
>>>> I am not disputing that XMI can be used.  I am sure it can be 
>>>> used.  I am disputing that UML itself (without XMI) can be used 
>>>
>>>
>>>
>>>
>>> But this is the point, the recommended mechanism to serialize a "UML 
>>> models" is using the MOF model of UML X.y and applying the XMI 
>>> principles. This is what UML vendors have been doing for years?!
>>>
>>>> and noting that whatever we use, UN/CEFACT should standardize that 
>>>> thing.
>>>
>>>
>>>
>>>
>>> This is exactly what we are doing, we even got a projectplan and an 
>>> very active project for it. We have also scheduled a conference in 
>>> Stockholm where we are about to discuss UML in a UN/CEFACT 
>>> environment. More info later.
>>>
>>>> The OMG itself did note that using XMI had some issues albeit most 
>>>> of them are probably irrelevant to our purpose since they are 
>>>> concerned with consistent graphical representation.
>>>>
>>>> UML cannot be parsed by applications unless it is conveyed in some 
>>>> form of electronic format.  
>>>
>>>
>>>
>>>
>>> Recomended Serialization is through using XMI based on the MOF meta 
>>> model for a specific UML version.
>>>
>>>> UML itself does not (in the UML specification) constrain the exact 
>>>> format for serialization for items like object persistence, object 
>>>> serialization or interchange.
>>>
>>>
>>>
>>>
>>> The UML "framework" consists of several integrated parts including 
>>> XMI, MOF and a MOF meta model of UML
>>>
>>> Here is the MOF meta model for UML 
>>> <http://www.omg.org/cgi-bin/doc?ad/01-02-15>
>>> so this file + XMI principles = a XML schema for transporting UML 
>>> models.
>>>
>>>
>>> /anders
>>>
>>>
>>
>
>

-- 
***********
Senior Standards Strategist - Adobe Systems, Inc. - http://www.adobe.com
Vice Chair - UN/CEFACT Bureau Plenary - http://www.unece.org/cefact/
Chair - OASIS eb SOA TC - http://www.oasis-open.org/committees/tc_home.php?wg_abbrev=ebsoa
***********



[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]