Re: Users of the BON notation among Ada users ?

[Nicholas Collin Paul Gloucester <Colin_Paul_Gloster@ACM.org>]

On 2009-01-13, Jean-Pierre Rosen <rosen@adalog.fr> wrote:

|----------------------------------------------------------------------------|
|"Martyn Pike a ?crit :                                                      |
|                                                                            |
|> I have often wondered why HOOD and HRT-HOOD have not remained popular     |
|> within the Ada community.  Or perhaps they have - anyone care to comment ?|
|                                                                            |
|Because, unlike UML, they are not simply notations, but true design         |
|*methods*. They drive the design process through a precise engineering      |
|methodology. [..]                                                           |
|                                                                            |
|[..]"                                                                       |
|----------------------------------------------------------------------------|

They do not.

Not only do they lack rules for proceeding step by step, but even
Burns and Wellings or another coauthor (who took HOOD and produced
HRT-HOOD) admitted that HOOD and HRT-HOOD are not the best (though
this is not why they are not popular): on Hardcopy Page 3-4 of Volume
1 of a preliminary defintion of HRT-HOOD it was indicated that formal
methods are better. It was claimed however in that preliminary
definition:
"Structured methods often use a graphical representation, but unlike
the informal diagrams these graphs are well defined"
which is quite a boast. The inconsistency applied to layering (and
hence to decomposition) in depicting use relationships to uncles as on
for example Page 11 of their technical report on how HRT-HOOD could
have been applied to the Olympus satellite (a technical report in
which they excused their inability to use HRT-HOOD to satisfy the
requirements which had been satisfied in the real mission without
HRT-HOOD as describing the launched design which accomplished what was
required as being overengineered) is reminiscent of problems with
inconsistent levels of abstraction in UML.

Note that in the large definition of RAVENSCAR coauthored by Burns and
Wellings and Dr. T. V. (another proponent of HRT-HOOD), HRT-HOOD was
not expressive enough for a diagram so a diagram like (but
incompatible with) a diagram which might appear in HRT-HOOD was
used. That is not a good advertisement.

On Hardcopy Pages 2-18 and 2-25 of Volume 2 of the aforementioned
HRT-HOOD document it was strongly recommended to add the Ada package
Calendar but more recently this banned in another doucment with an
intersecting set of coauthors (Hardcopy Page 19 of the aforementioned
RAVENSCAR document). Coherency, anyone?

The HRT-HOOD tool from Intecs-HRT was reputed to be very buggy. That
would not have helped to make it popular.

Yours sincerely,
Nicholas Collin Paul Gloucester