Hi

I am currently thinking to come with some solution related to
Processor Obsolescence. To start with, it's always wise to come with a
document which should have abundant information and queries to solve
it related to Processor Obsolescence. I think the embedded community
members will appreciate in sharing some knowledge of theirs and guide
me to some extents, so that I can at least move in right direction
with having mamooth queries to solve it and also improve the document
contents.

To solve Processor Obsolescence problem, the easy solution which comes
in my mind is building a new construct called Meta Language. Here, I
don't mean re-writing C or C++ HLL, but simply defining constructs
which can serve the purpose in parallel to C/C++ HLL. Also, the focus
is currently a single processor (probably, one from the series of m68k
family) to start with. To solve this problem I don't wish to think
preferably with HAL concepts.

I have also listed some queries below to brain-storm myself -

(1) Since processor obsolescence goes at processor level or
architectural level, is coming of New language called Meta Language
sufficient to support or choosing a operating system in parallel which
can be modified at both processor and application level. Which
operating system can support above things, is MINIX or Linux or any
other operating systems a right answer?
Note: Normally, the application software relies on the services of
Operating System, the set of operating system service calls or system
calls maybe considered as defining the boundary betweenthe application
software and operating-system. Also, an operating-system interacts
with the processor, the processor instruction set may be considered as
defining the boundary between theapplication software and operating-
system.

(2) When the question of Language comes, grammar is mandatory. Is
grammar written for new language (ML) should have manual lexer or the
lexer as available for C/C++? Should attempting to write parser for
New Language Semantics would be a prudent approach?

(3) Is coming with New Language only solution for resolving Processor
obsolescence? Re-writing or coming with New Language isn't a time-
consuming as language-to-language translation will happen?

(4) The New language which I am thinking to write for the processor
specific which has become obsolete, is considerable study required
with original design of processor and recent technology update. Are
the factors required to be noted - how to fill the gap?

(5) Do I need to think that the obsolete processor hardware is to be
tested rigorously so that New Language implementation software could
be partitioned into number of components in having well-defined
interactions between them?

(6) Does New Language should take care of all instructions as defined
with obsolete processor? Also, is the modified Compiler of New
Language going to taken care of memory consumption and memory
management for obsolete processor?

(7) Do I need to think that the New Language should have some abstract
concept of having Hardware Abstraction Layer(HAL) for the obsolete
processor? Or can earlier HAL if defined for the processor be used
with New Language?

(8) Since applications exhibit high degree of dependency on hardware,
and graphical displays fall in this category. Do the New Language be
capable of manipulating graphics memory and execute graphics
primitives (i.e., points, lines, circles, etc.).

(9) Is the obsolete processor working on 8-bit or 16-bit data and is
the New Language taking care of indexing the data structures or
locating data structures within processor memory?

(10) Do the concept of SMP on Processor Obsolescence can make to
worry?

The above queries though many won't be required but thought to put in,
wish to have more queries from the embedded community members, so that
finally the document and it's implementation becomes more worth.

Looking forward for some & guidance. Having reference of persons
or companies or links who are already working in these areas will
really be of great too.

BR
Mukesh K Srivastava

<srimks11@gmail.com> wrote in message
news:1192866786.601111.224180@e9g2000prf.googlegro ups.com...
CPU development is certainly not obsolete. Perhaps your programming style
wrt said processors is out of band...

srimks11@gmail.com wrote:

> Hi
>
> I am currently thinking to come with some solution related to
> Processor Obsolescence. To start with, it's always wise to come with a
> document which should have abundant information and queries to solve
> it related to Processor Obsolescence. I think the embedded community
> members will appreciate in sharing some knowledge of theirs and guide
> me to some extents, so that I can at least move in right direction
> with having mamooth queries to solve it and also improve the document
> contents.
>
> To solve Processor Obsolescence problem, the easy solution which comes
> in my mind is building a new construct called Meta Language. Here, I
> don't mean re-writing C or C++ HLL, but simply defining constructs
> which can serve the purpose in parallel to C/C++ HLL. Also, the focus
> is currently a single processor (probably, one from the series of m68k
> family) to start with. To solve this problem I don't wish to think
> preferably with HAL concepts.
>
> I have also listed some queries below to brain-storm myself -

Welcome Mukesh,

Obsolescence is a factor of long term supported products that should rightly
be factored into the business risk assessment, budgets and planning stages
of a project. However, there is only so far one can foresee and the
unexpected curtailing of production of vital parts does occur.

One of the best protections against obsolescence is to design your systems
in a highly modular form where the interfaces (surfaces) of the modules are
well documented and fully specified. Such specification will include a
record of performance, behaviour and limitations.

Programming language and operating systems are only a manifestation of the
environment presented to the application. There are currently a wide spread
of single and multi-core processors which use the same operating systems
and can run the same applications no matter what the underlying hardware
actually is (so long as the right sort of facilities are provided).

All systems break down into a user interface, a communications channel and
some storage area. The mix of type and quantity of each may vary from
system to system but applications will still run. As a system developer you
need to ensure you build to standards (internal and external) that will
allow that to continue to happen.

--
************************************************** ******************
Paul E. Bennett...............<email://Paul_E.Bennett@topmail.co.uk>
Forth based HIDECS Consultancy
Mob: +44 (0)7811-639972
Tel: +44 (0)1235-811095
Going Forth Safely ..... EBA. www.electric-boat-association.org.uk..
************************************************** ******************

"Chris Thomasson" <cristom@comcast.net> wrote in message
news:4r6dncBDm72DI4TanZ2dnUVZ_rqlnZ2d@comcast.com. ..
> <srimks11@gmail.com> wrote in message
> news:1192866786.601111.224180@e9g2000prf.googlegro ups.com...
> CPU development is certainly not obsolete. Perhaps your programming style
> wrt said processors is out of band...

Oh shit. Sorry about that, I misread your post!


I would listen to Joes advise wrt. ccNUMA.

On Sat, 20 Oct 2007 10:36:22 -0700, "Chris Thomasson"
<cristom@comcast.net> wrote:

>"Chris Thomasson" <cristom@comcast.net> wrote in message
>news:4r6dncBDm72DI4TanZ2dnUVZ_rqlnZ2d@comcast.com ...
>> <srimks11@gmail.com> wrote in message
>> news:1192866786.601111.224180@e9g2000prf.googlegro ups.com...
>> CPU development is certainly not obsolete. Perhaps your programming style
>> wrt said processors is out of band...
>
>Oh shit. Sorry about that, I misread your post!

Watch out, Thomasson. You're getting more senile every day. ahahaha...

Louis Savain

Why Software Is Bad and What We Can Do to Fix It:
http://www.rebelscience.org/Cosas/Reliability.htm

"Traveler" <traveler@noasskissers.net> wrote in message
news:mefkh3t5m69r7onf5ihrj4vm1dcqtp6vgk@4ax.com...
> On Sat, 20 Oct 2007 10:36:22 -0700, "Chris Thomasson"
> <cristom@comcast.net> wrote:
>
>>"Chris Thomasson" <cristom@comcast.net> wrote in message
>>news:4r6dncBDm72DI4TanZ2dnUVZ_rqlnZ2d@comcast.co m...
>>> <srimks11@gmail.com> wrote in message
>>> news:1192866786.601111.224180@e9g2000prf.googlegro ups.com...
>>> CPU development is certainly not obsolete. Perhaps your programming
>>> style
>>> wrt said processors is out of band...
>>
>>Oh shit. Sorry about that, I misread your post!
>
> Watch out, Thomasson. You're getting more senile every day. ahahaha...

lol.