Re: Should Inline be private in the private part of a package spec?

[Georg Bauhaus <rm.dash-bauhaus@futureapps.de>]

On 15.08.12 20:53, Dmitry A. Kazakov wrote:
> On Wed, 15 Aug 2012 18:53:09 +0200, Georg Bauhaus wrote:
>
>> On 15.08.12 18:24, Dmitry A. Kazakov wrote:
>>> On Wed, 15 Aug 2012 17:49:01 +0200, Georg Bauhaus wrote:
>>>
>>>> On 15.08.12 13:55, Dmitry A. Kazakov wrote:
>>>>> On Wed, 15 Aug 2012 13:08:37 +0200, Georg Bauhaus wrote:
>>>>>
>>>>>> I would say that a network connecting the electrical devices
>>>>>> of a car does not counts as heterogeneous.
>>>>>
>>>>> How are the ECU, navigation system, ABS, windshield wiper homogenous?
>>>>
>>>> They are homogeneous in that the 4 suppliers agree to be supplying
>>>> for the same 1 system, under the "umbrella-notion" of car
>>>> electronics.
>>>
>>> I see, not to forget that the gravitational constant is same too. That
>>> evidently makes it homogenous...
>>
>> Dropping the blinders will make you see how people decide
>> which data format is adequate under the respective circumstances.
>>
>> I have tried to supply hooks for formalization of these
>> situational variables. Apply fuzzy logic.
>
> Formal by which measure?

They are formal by any measure that relevant sciences have
produced from material evidence, and that can be tested by
observing reality. (They allow theorems using If-Then,
TheMore-TheLess, ceteris paribus, ...)

> And how does it make car's devices homogenous?

Homogeneous by all important points of view, see below.

> And
> how these false premises could justify any application of XML even if per
> some miracle they happened true?

The point of the exercise is to make apparent the variables
that help decide when, and when not, to use XML, based on
a measure of homogeneity *of* systems, not a measure of
perceived homogeneity or heterogeneity of devices *within*
*a* system ("system" as in "car's electronics").

:non-Ada.

Anyway, how can an ECU and a windshield wiper be homogeneous?
I'll try an elaborate answer although the answer seems obvious,
from a distance. And distance is a measure, needless to say!

Car electronics, by the very fact that this classification exists,
establishes a class of things that are of the same kind (homogeneous).
This requires a measure of sameness. There are several variables.
This is somewhat ad-hoc, but it leads to ways of measuring sameness.

With regard to the format of data exchange in a network, I imagine
there are at least these:

1) The equivalence class of car engineers establishes expectations.
These are likely sensible, based on technical knowledge and
experience, modulo personal preferences.

For example, making the car's wires transport Word documents
to the windshield wipers is likely considered absurd by every
engineer, XML is likely considered pointless by almost all
engineers, and then there will be interesting descriptions of
data formats by several engineers. [-All, -Almost_All, +Several;
more are of the same opinion (against Word or XML)]

2a) locality of networks,
2b) properties of connections,
2c)control of wires
establish, and limit, technical possibilities, [Length, Quality,
Predictability; e.g. every device is connected within the same
thing of 2x2x5 units of length, and unlike a web server of
a public library, the stop lamps cannot be offline between
1:00 AM and 3:00 AM]

3) all parties involved my communicate the data formats to be
used by others [E.g., REST services are of a different kind]

4) swiftness of data processing [Latency, Number of layers]