Ada and Distributed Objects - Seeking Success Stories

Ada and Distributed Objects - Seeking Success Stories

[Barry and Jackie Schiff]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1035 bytes --]



I am currently involved in a large real-time Ada project (1,000,000
SLOC). We have successfully developed and are using a Distributed Object
Manager that implements a client/server model and an OO paradigm. The
basic concept is that clients subscribe/register for passive objects
that they are interested in and object managers at the servers
automatically distribute applicable updates to clients. All updates are
based on transactions. When a commit is performed at the server it
evaluates the changes against the outstanding subscriptions and updates
the object-space at the impacted clients. Clients can optionally
register notification handlers which act as callback for specified
update events.

What I am looking for is a list of other projects that have used similar
techniques and have been successfully. Thanks in advance for any
information you may provide. I would prefer direct email for a response
if you don�t mind. Feel free to also post back to the group if you think
others might be interested.

>Barry Schiff

Re: Ada and Distributed Objects - Seeking Success Stories

[Michael F Brenner]

    > I am currently involved in a large real-time Ada project (1,000,000
    > SLOC) ... Distributed Object Manager ... client/server model ...
    > clients subscribe/register for passive objects that they are 
    > interested in and object managers at the servers automatically
    > distribute applicable updates to clients. All updates are
    > based on transactions. When a commit is performed at the server it
    > ... updates the object-space at the impacted clients. 

This sounds like an interesting project, but I am curious why it is called
real-time, which usually means that each transaction must be completed by
a specific deadline. This description seems to have unlimited delays in
each remote procedure call, each client server interaction, and each
transaction, making it impossible to meet any particular deadline set in 
advance, for example, when the clients or servers are turned off or break
for a few days?

Re: Ada and Distributed Objects - Seeking Success Stories

[Robert Dewar]

Michael says

<<This sounds like an interesting project, but I am curious why it is called
real-time, which usually means that each transaction must be completed by
a specific deadline>>

In the most commonly used terminology, what you describe is called *hard*
realtime. The designation realtime without the hard does not imply that
thee are deadlines that must be met.

Re: Ada and Distributed Objects - Seeking Success Stories

[Samuel Tardieu]

>>>>> "Robert" == Robert Dewar <dewar@merv.cs.nyu.edu> writes:

Robert> Michael says <<This sounds like an interesting project, but I
Robert> am curious why it is called real-time, which usually means
Robert> that each transaction must be completed by a specific
Robert> deadline>>

Robert> In the most commonly used terminology, what you describe is
Robert> called *hard* realtime. The designation realtime without the
Robert> hard does not imply that thee are deadlines that must be met.

To rephrase what you said, *soft* realtime may usually be seen as
*best effort*.

  Sam
-- 
Samuel Tardieu -- sam@ada.eu.org

Re: Ada and Distributed Objects - Seeking Success Stories

[Robert Dewar]

Sam said (replying to me)

<<Robert> In the most commonly used terminology, what you describe is
Robert> called *hard* realtime. The designation realtime without the
Robert> hard does not imply that thee are deadlines that must be met.

To rephrase what you said, *soft* realtime may usually be seen as
*best effort*..>>

Well it may be more than this, you may have absolute requirements for
average service performance, even if there are no specific hard deadllines
to be met...

Re: Ada and Distributed Objects - Seeking Success Stories

[W. Wesley Groleau (Wes)]

> To rephrase what you said, *soft* realtime may usually be seen as
> *best effort*.

Or as one text-book author wrote,
   " 'real-time' means 'quick as a bunny' "
----------------------------------------------------------------------
    Wes Groleau, Hughes Defense Communications, Fort Wayne, IN USA
Senior Software Engineer - AFATDS                  Tool-smith Wanna-be

  w w g r o l  at  p s e u l t 0 1  dot  f w  dot  h a c  dot  c o m

     SPAM should be sent to   I.want.one@mailbombs.for.idiots.org
 If you don't want to pay $500 (see 47 USC 227), don't send it here.
----------------------------------------------------------------------

Re: Ada and Distributed Objects - Seeking Success Stories

[Barry and Jackie Schiff]

Michael F Brenner wrote:

> This sounds like an interesting project, but I am curious why it is called
> real-time, which usually means that each transaction must be completed by
> a specific deadline. This description seems to have unlimited delays in
> each remote procedure call, each client server interaction, and each
> transaction, making it impossible to meet any particular deadline set in
> advance, for example, when the clients or servers are turned off or break
> for a few days?


I left out a lot details one of which is that an optimistic
async transaction model is used (post commit version check
not locking). No RPCs just async communication.