Re: HRT-HOOD

[Niklas Holsti <niklas.holsti@tidorum.invalid>]

On 19-03-14 21:36 , russ lyttle wrote:
> Are there any diagramming tools supporting HRT-HOOD?

Ellidiss has some tools, https://www.ellidiss.com/.

See also 
http://www.hurray.isep.ipp.pt/ae2006/pdfs/Tuesday/Vendor_Sessions/ellidiss.pdf.

Their STOOD tool is interesting, as it uses Prolog. It is not often one 
sees a real-life Prolog application.

> Google returns lots of academic papers referencing HRT-HOOD, but nothing
> about tools. UML and HRT-HOOD seem to have issues.

The Ellidiss presentation linked to above speaks of a tool called 
HRT-UML, which seems to be an evolution from (or a replacement of) an 
earlier HOOD tool called HoodNICE, of which I have experience from some 
decades ago (I would not recommend _that_ version of HoodNICE).

Regarding the HOOD method itself, do note that the HOOD object hierarchy 
is not equivalent to an Ada child package hierarchy, because in HOOD a 
parent object depends on its child objects, also through the public 
specifications. HOOD was designed in the Ada-83 era, before child 
packages existed.

The real-time aspects of HOOD itself always seemed baroque to me; 
associating real-time behaviour with operations departed too much from 
Ada concepts. Nor did the code generator in HoodNICE support those 
real-time properties. (Well, it was called a "code generator", but most 
of the Ada code had to be manually written into the model's operation 
specifications and was just textually copied into the "generated" Ada 
sources.)

The real-time concepts in HRT-HOOD felt OK, when I last saw it (a decade 
or so ago). As I recall, they were roughly equivalent to the Ravenscar 
profile.

-- 
Niklas Holsti
Tidorum Ltd
niklas holsti tidorum fi
       .      @       .