Re: dip Notions 2 Major Errors

From: Mark Nicholls (Nicholls.Mark_at_mtvne.com)
Date: 09/15/04 

Date: 15 Sep 2004 05:37:46 -0700

"Roger L. Cauvin" <roger@deadspam.com> wrote in message news:<2qmuk3Fv0ae6U1@uni-berlin.de>...
> "Universe" <universe@tAkEcovad.OuT.net> wrote in message
> news:3n0ck0pjnf07ljt8kte67uk1l7dtur3qim@4ax.com...
> >
> > But Cauvin is skirting the fact that even though RCM has
> > recently backed away from doing so, in RCM's earlier
> > posts in the dip threads, the design/structuring of
> > compilation units whether called modules or packages
> > was INTEGRAL to the practice of dip.
>
> No, I reviewed his messages in these threads. Robert did mention packages
> (sets of classes) as particularly illustrative of the impact of dependency
> inversion, but he did not portray them as essential to DIP.
>
> And when he has used the term "module" synonymously with "package", it has
> only been in direct response to such usage on Mark's part.

Mr Martin wrote in his post.

> |------module A------| |------module B------|
>
> |P|-------------------->|IQ|<|------|Q|
>
> And then move the interface out of the depended upon module.
>
> |------module A------| |------module B------|
>
> |P|----->|IQ|<|----------------|Q|

....

Mr Martin wrote in his post.....

> Then the inverted dependency was stretched
> across the module boundary so that B now depends upon A.

note the use of the word module and how he has drawn two layers,
logical and physical.

I agree with him about the definition of module and we both agree with
Booch.

Can we agree to disagree? I can.

Relevant Pages

  • Re: dip Notions 2 Major Errors
    ... >Mr Martin of that. ... >> show a logical inversion of dependancies. ... Now imagine C which uses A. There is a "logical dependency" ... You can't apply DIP to every dependency. ...
    (comp.object)
  • Re: dip Notions 2 Major Errors
    ... > show a logical inversion of dependancies. ... this is a 'logical dependency' ... > B. Abstraction should not depend on details. ... break DIP, or more than 2 layers you will find your physical module ...
    (comp.object)
  • Re: dip Notions 2 Major Errors
    ... primary arguments against the DIP are as follows: ... applying the DIP never changes the dependency so that 'y ... To the extent that the DIP inverts dependencies at the package level, ... That's a form of inversion, ...
    (comp.object)
  • Re: To understand OOP better...
    ... > making callers independent from callees, ... if you sketch the dependency graphs of procedural programs and OO ... I can assert that DIP has been more useful ... > Most of them have cycles. ...
    (comp.object)
  • su -c yum update
    ... 2752/2752 Added 56 new packages, deleted 29 old in 7.44 seconds ... ---> Downloading header for abiword to pack into transaction set. ... Importing Additional filelist information for dependency resolution ...
    (Fedora)