Wi-Fi N is still in the research phase.  ;)

Doing complete round-tripping is not a viable goal, but you can do
enough stuff to make it useful, even though it's not a complete
implementation.  So it's both.  Hell, I read a few weeks ago that some
academics created a new prime number sieve 4-5 years ago.  That
problem has been "solved" for thousands of years, but people are still
doing research on it.

On Tue, Feb 10, 2009 at 11:40 AM, Henry Ho <[email protected]> wrote:
> This reminds me of the problem of what WYSIWYG HTML editor had in the last
> decade.  Once some custom HTML is added, the editor can no longer edit the
> page.
>
> So if you're already using a tool (the French one) that does round tripping,
> why did Peter make it sound like it is just a concept in the research phase?
>
>
> Henry Ho
>
>
> On Tue, Feb 10, 2009 at 11:28 AM, Barney Boisvert <[email protected]>
> wrote:
>>
>> That is the model->code piece, which pretty much every significant UML
>> tool will do.  The round tripping comes when you add a new type and a
>> few methods to the Java code and the UML tool detects that and updates
>> the UML model for you. We used a tool from a French company that I
>> simply cannot remember the name of.  Started with an "O", I believe.
>> It worked reasonably well, but like I said, it was all too easy to get
>> stuck in a corner that the tool didnt support and have to do a lot of
>> extra work.  That was a few years ago though.
>>
>> cheers,
>> barneyb
>>
>> On Tue, Feb 10, 2009 at 11:17 AM, Henry <[email protected]> wrote:
>> >
>> > So it is like, Rational Rose ( http://www-01.ibm.com/software/rational/
>> > ) that works both way? I've used it before in a java course in
>> > college.  It lets you specify the UML and then it generates the java
>> > classes for you.  Is round tripping something like, an ultra smart
>> > thing like Rational Rose that can read your code and update the model?
>> >
>> > Henry Ho
>> >
>> > On Feb 10, 11:02 am, Barney Boisvert <[email protected]> wrote:
>> >> When you can take your model and generate the code for it, and then
>> >> change the code and update the model from the modified source code.
>> >> I.e. it goes both ways.  Doing model->code is easy, doing code->model
>> >> is harder, doing both is crazy difficult.  So the idea is to get your
>> >> model as rich as possible so that you don't have to do round tripping,
>> >> you only do code generation.  But to get there, you start programming
>> >> in your modeling language, which just flattens the process back out
>> >> (i.e. you're only doing programming, you're not doing modeling).  So
>> >> it's a delicate balancing act.
>> >>
>> >> In the real world, the model is often expressed in UML, and then
>> >> translated into a "real" langauge to be executed.  In the ideal world,
>> >> that happens without further modification by the developer.  If you're
>> >> not in an ideal world, you modify the generated code, and then need to
>> >> round trip that information back to the UML so that next time you
>> >> generate the code it doesn't have to have the same mods reapplied.
>> >>
>> >> cheers,
>> >> barneyb
>> >>
>> >> On Tue, Feb 10, 2009 at 10:51 AM, Henry <[email protected]> wrote:
>> >>
>> >> > What is round tripping?
>> >>
>> >> --
>> >> Barney Boisvert
>> >> [email protected]http://www.barneyb.com/
>> > >
>> >
>>
>>
>>
>> --
>> Barney Boisvert
>> [email protected]
>> http://www.barneyb.com/
>>
>>
>
>
> >
>



-- 
Barney Boisvert
[email protected]
http://www.barneyb.com/

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"CFCDev" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/cfcdev?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to