Re: Real time Java? Huh?

Re: Real time Java? Huh?

[Pat Rogers]

David Chase wrote in message <364224FB.7FCF@world.std.com>...

<snip>

>Most of that commercial software is written in "unsafe" languages.
>C++ is not safe.  Pascal is not safe.  Ada, as typically used,
>is not safe.

Detailed, hard evidence, please, regarding Ada.

Re: Real time Java? Huh?

[Joseph]

I missed way too much of this thread.   Java as a real time language?
Tell me that this is a hoax, please.

Joseph

Jay O'Connor wrote:

> On Thu, 5 Nov 1998 21:21:27 -0600, "Pat Rogers"
> <progers@NOclasswideSPAM.com> wrote:
>
> >David Chase wrote in message <364224FB.7FCF@world.std.com>...
> >
> ><snip>
> >
> >>Most of that commercial software is written in "unsafe" languages.
> >>C++ is not safe.  Pascal is not safe.  Ada, as typically used,
> >>is not safe.
> >
> >Detailed, hard evidence, please, regarding Ada.
>
> That one kinda surprised me as well because Ada was designed to be
> safe but also designed to support low level hardware access if needed.
>
> I mean, the language was originally targetted for embedded weapons
> systems and such,
>
> BTW - The onboard systems on the Boeing 777 are written in Ada.  As
> are most of the train systems in Europe.
>
> Jay O'Connor
> joconnor@roadrunner.com
> http://www.roadrunner.com/~joconnor
>
> "God himself plays on the bass strings first, when he tunes the soul"

Re: Real time Java? Huh?

[Jean-Pierre Rosen]

Joseph <**********@*******.**> a �crit dans le message
<36428b55.0@blushng.jps.net>...
>I missed way too much of this thread.   Java as a real time language?
>Tell me that this is a hoax, please.
>
At an AFNOR-Ada (French standardization body) we had a discussion
about this yesterday.
It seems that by real-time Java, people really mean "embedded Java",
i.e. in SmartCards for example. When I asked how extensive dynamic
allocation and garbage collector could be used in smart-cards,
people told me that there were restrictions, like only final classes
with only final methods... When I asked why use Java rather than
C, and why not call such "classes" packages, I had no answer...
sigh.
----------------------------------------------------------------------------
                  J-P. Rosen (Rosen.Adalog@wanadoo.fr)
      Visit Adalog's web site at http://perso.wanadoo.fr/adalog

Re: Real time Java? Huh?

[dewar]

In article <36428b55.0@blushng.jps.net>,
  Joseph <**********@*******.**> wrote:
> I missed way too much of this thread.
> Java as a real time language?
> Tell me that this is a hoax, please.

Sorry, no one will tell you that, because it is most
certainly the case that there are JVM systems that are
suitable for real-time use.

For an interesting demo, please stop by the Ada Core
Technologies booth at the Washington Sig-Ada meeting next
week, where you will see some real-time Java in action
(well actually most of what you see will be Ada targetted
to the JVM for real time use :-)

Robert Dewar
Ada Core Technologies

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: Real time Java? Huh?

[dewar]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1279 bytes --]

In article <71uc8n$lu1$1@platane.wanadoo.fr>,
  "Jean-Pierre Rosen" <rosen.adalog@wanadoo.fr> wrote:
>
> Joseph <**********@*******.**> a �crit dans le message
> <36428b55.0@blushng.jps.net>...
> >I missed way too much of this thread.   Java as a real
time language?
> >Tell me that this is a hoax, please.
> >
> At an AFNOR-Ada (French standardization body) we had a
discussion
> about this yesterday.
> It seems that by real-time Java, people really mean
"embedded Java",
> i.e. in SmartCards for example. When I asked how
extensive dynamic
> allocation and garbage collector could be used in
smart-cards,
> people told me that there were restrictions, like only
final classes
> with only final methods... When I asked why use Java
rather than
> C, and why not call such "classes" packages, I had no
answer...
> sigh.



There are very good reasons for using JVM based approaches
for real time embedded systems, and you should not just
assume that garbage collection is out of the question in
such systems. This is why it is so important that we have
Ada 95 systems that target the JVM!

Robert Dewar
Ada Core Technologies

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: Real time Java? Huh?

[Jean-Pierre Rosen]

dewar@gnat.com a �crit dans le message <71usgj$uv4$1@nnrp1.dejanews.com>...
>There are very good reasons for using JVM based approaches
>for real time embedded systems, and you should not just
>assume that garbage collection is out of the question in
>such systems. This is why it is so important that we have
>Ada 95 systems that target the JVM!
>
Sure. Garbage collection is not out of the question for some
systems, but there are certainly many RT systems where it
is an issue.

Anyway, the issue was not about the JVM, but about the
Java language. I keep trying to separate these issues
clearly - Guess why!

BTW, can we deduce from your message that the
soon-to-be-delivered GNAT 3.11p will include
GNAT-for-Java ?
----------------------------------------------------------------------------
                  J-P. Rosen (Rosen.Adalog@wanadoo.fr)
      Visit Adalog's web site at http://perso.wanadoo.fr/adalog

Re: Real time Java? Huh?

[oopster]

In article <3642a327.38883600@news.nmia.com>,
  joconnor@roadrunner.com wrote:
> On Thu, 05 Nov 1998 21:36:10 -0800, Joseph <**********@*******.**>
> wrote:
>
> >I missed way too much of this thread.   Java as a real time language?
> >Tell me that this is a hoax, please.
>
> That gist of it is that some companies are breaking with Sun and
> trying to establish a standard for Java in real time systems.

More accurately, the gist is that Sun has a proposed standard for real-time
Java, and a group of other companies has also formed a consortium to promote
a somewhat different realtime Java standard.  The main issue seems to be the
licensing terms for Sun Embedded Java.

HP has already produced a competing embedded Java VM, which they are also
licensing to other companies.

> I remember being at a Alsys demonstration once where one of the Ada
> compiler writers warned against using exception handling in hard real
> time systems because you couldn't guarantee how long it would take to
> unwind to find the handler for the raised exception.  I wonder if such
> conditions esist for exception handling in Java?  (Or dynamic
> dispatch, for that matter)

Is it still the case that Ada exceptions are to be avoided for real-time?
Not a good thing if true...

It seems clear that deterministic exception handling will be critical to
real-time Java efforts.  I'm sure you can find out more about this at the
various real-time Java websites if you're interested.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: Real time Java? Huh?

[Jay O'Connor]

On Thu, 5 Nov 1998 21:21:27 -0600, "Pat Rogers"
<progers@NOclasswideSPAM.com> wrote:

>David Chase wrote in message <364224FB.7FCF@world.std.com>...
>
><snip>
>
>>Most of that commercial software is written in "unsafe" languages.
>>C++ is not safe.  Pascal is not safe.  Ada, as typically used,
>>is not safe.
>
>Detailed, hard evidence, please, regarding Ada.

That one kinda surprised me as well because Ada was designed to be
safe but also designed to support low level hardware access if needed.

I mean, the language was originally targetted for embedded weapons
systems and such,

BTW - The onboard systems on the Boeing 777 are written in Ada.  As
are most of the train systems in Europe.

Jay O'Connor
joconnor@roadrunner.com
http://www.roadrunner.com/~joconnor

"God himself plays on the bass strings first, when he tunes the soul"

Re: Real time Java? Huh?

[Jay O'Connor]

On Thu, 05 Nov 1998 21:36:10 -0800, Joseph <**********@*******.**>
wrote:

>I missed way too much of this thread.   Java as a real time language?
>Tell me that this is a hoax, please.


That gist of it is that some companies are breaking with Sun and
trying to establish a standard for Java in real time systems.

I remember being at a Alsys demonstration once where one of the Ada
compiler writers warned against using exception handling in hard real
time systems because you couldn't guarantee how long it would take to
unwind to find the handler for the raised exception.  I wonder if such
conditions esist for exception handling in Java?  (Or dynamic
dispatch, for that matter)



>Joseph
>
>Jay O'Connor wrote:
>
>> On Thu, 5 Nov 1998 21:21:27 -0600, "Pat Rogers"
>> <progers@NOclasswideSPAM.com> wrote:
>>
>> >David Chase wrote in message <364224FB.7FCF@world.std.com>...
>> >
>> ><snip>
>> >
>> >>Most of that commercial software is written in "unsafe" languages.
>> >>C++ is not safe.  Pascal is not safe.  Ada, as typically used,
>> >>is not safe.
>> >
>> >Detailed, hard evidence, please, regarding Ada.
>>
>> That one kinda surprised me as well because Ada was designed to be
>> safe but also designed to support low level hardware access if needed.
>>
>> I mean, the language was originally targetted for embedded weapons
>> systems and such,
>>
>> BTW - The onboard systems on the Boeing 777 are written in Ada.  As
>> are most of the train systems in Europe.
>>
>> Jay O'Connor
>> joconnor@roadrunner.com
>> http://www.roadrunner.com/~joconnor
>>
>> "God himself plays on the bass strings first, when he tunes the soul"
>

Jay O'Connor
joconnor@roadrunner.com
http://www.roadrunner.com/~joconnor

"God himself plays on the bass strings first, when he tunes the soul"

Re: Real time Java? Huh?

[Roga Danar]

Timing is a better (or another) reason why one should look twice at a Java
for a real-time system must instead of Ada.  Real-Time embedded systems
that frequently require persice processing in a narrow time margin would be
better off with Ada, IHO.

dewar@gnat.com wrote:

> In article <71uc8n$lu1$1@platane.wanadoo.fr>,
>   "Jean-Pierre Rosen" <rosen.adalog@wanadoo.fr> wrote:
> >
> > Joseph <**********@*******.**> a �crit dans le message
> > <36428b55.0@blushng.jps.net>...
> > >I missed way too much of this thread.   Java as a real
> time language?
> > >Tell me that this is a hoax, please.
> > >
> > At an AFNOR-Ada (French standardization body) we had a
> discussion
> > about this yesterday.
> > It seems that by real-time Java, people really mean
> "embedded Java",
> > i.e. in SmartCards for example. When I asked how
> extensive dynamic
> > allocation and garbage collector could be used in
> smart-cards,
> > people told me that there were restrictions, like only
> final classes
> > with only final methods... When I asked why use Java
> rather than
> > C, and why not call such "classes" packages, I had no
> answer...
> > sigh.
>
> There are very good reasons for using JVM based approaches
> for real time embedded systems, and you should not just
> assume that garbage collection is out of the question in
> such systems. This is why it is so important that we have
> Ada 95 systems that target the JVM!
>
> Robert Dewar
> Ada Core Technologies
>
> -----------== Posted via Deja News, The Discussion Network ==----------
> http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own

Re: Real time Java? Huh?

[Roga Danar]

dewar@gnat.com wrote:

> In article <71uc8n$lu1$1@platane.wanadoo.fr>,
>   "Jean-Pierre Rosen" <rosen.adalog@wanadoo.fr> wrote:
> >
> > Joseph <**********@*******.**> a �crit dans le message
> > <36428b55.0@blushng.jps.net>...
> > >I missed way too much of this thread.   Java as a real
> time language?
> > >Tell me that this is a hoax, please.
> > >
> > At an AFNOR-Ada (French standardization body) we had a
> discussion
> > about this yesterday.
> > It seems that by real-time Java, people really mean
> "embedded Java",
> > i.e. in SmartCards for example. When I asked how
> extensive dynamic
> > allocation and garbage collector could be used in
> smart-cards,
> > people told me that there were restrictions, like only
> final classes
> > with only final methods... When I asked why use Java
> rather than
> > C, and why not call such "classes" packages, I had no
> answer...
> > sigh.
>
> There are very good reasons for using JVM based approaches
> for real time embedded systems, and you should not just
> assume that garbage collection is out of the question in
> such systems. This is why it is so important that we have
> Ada 95 systems that target the JVM!
>
> Robert Dewar
> Ada Core Technologies
>
> -----------== Posted via Deja News, The Discussion Network ==----------

> http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own

>

Timing is a better (or another) reason why one should look twice at a Java
for a real-time system must instead of Ada.  Real-Time embedded systems
that frequently require persice processing in a narrow time margin would be
better off with Ada, IHO.

Re: Real time Java? Huh?

[Jay O'Connor]

On Fri, 06 Nov 1998 19:11:46 GMT, oopster@my-dejanews.com wrote:


>Is it still the case that Ada exceptions are to be avoided for real-time?
>Not a good thing if true...

Not sure.  That was for Ada '83 about 5 or 6 years ago.

Jay O'Connor
joconnor@roadrunner.com
http://www.roadrunner.com/~joconnor

"God himself plays on the bass strings first, when he tunes the soul"

Re: Real time Java? Huh?

[dewarr]

In article <71ve5m$iv2$1@platane.wanadoo.fr>,
  "Jean-Pierre Rosen" <rosen.adalog@wanadoo.fr> wrote:

 BTW, can we deduce from your message that the
> soon-to-be-delivered GNAT 3.11p will include
> GNAT-for-Java ?

No, not yet, GNAT-for-Java is still very much in the
testing phase, but it is getting closer, we have some
customers doing serious evaluations, and as I mentioned you
will be able to see an interesting real-time demo at
SigAda.


-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: Real time Java? Huh?

[bill]

In article <720mvr$gg4$1@nnrp1.dejanews.com>, dewarr@my-dejanews.com says...
>
>
>No, not yet, GNAT-for-Java is still very much in the
>testing phase, but it is getting closer, we have some
>customers doing serious evaluations, and as I mentioned you
>will be able to see an interesting real-time demo at
>SigAda.
>
 
Cool stuff !

go Ada !

Bill.

Re: Real time Java? Huh?

[whiter5195]

In article <36435a47.1894177@news.nmia.com>,
  joconnor@roadrunner.com wrote:
> On Fri, 06 Nov 1998 19:11:46 GMT, oopster@my-dejanews.com wrote:
>
> >Is it still the case that Ada exceptions are to be avoided for real-time?
> >Not a good thing if true...
>
> Not sure.  That was for Ada '83 about 5 or 6 years ago.

  I know that we use as a code inspection checklist item that
exceptions are not to be used for normal control flow.  We do HRT.
At "outside world" input interfaces, values are checked with normal
logic comparisons and a completely deterministic response is given
(often very helpful since the offending parameter is found in a
very localized fashion).  Could this be done in a more blanket fashion
by enabling full Ada checks and having a local exception handler?
Perhaps, but IMHO a system that is running properly _should_ not
have any exceptions being raised.
_____________________________________________________________________
Robert S. White

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: Real time Java? Huh?

[Charles R. Lyttle]

oopster@my-dejanews.com wrote:
> 
> In article <3642a327.38883600@news.nmia.com>,
>   joconnor@roadrunner.com wrote:
> > On Thu, 05 Nov 1998 21:36:10 -0800, Joseph <**********@*******.**>
> > wrote:
> >
> > >I missed way too much of this thread.   Java as a real time language?
> > >Tell me that this is a hoax, please.
> >
> > That gist of it is that some companies are breaking with Sun and
> > trying to establish a standard for Java in real time systems.
> 
> More accurately, the gist is that Sun has a proposed standard for real-time
> Java, and a group of other companies has also formed a consortium to promote
> a somewhat different realtime Java standard.  The main issue seems to be the
> licensing terms for Sun Embedded Java.
> 
> HP has already produced a competing embedded Java VM, which they are also
> licensing to other companies.
> 
> > I remember being at a Alsys demonstration once where one of the Ada
> > compiler writers warned against using exception handling in hard real
> > time systems because you couldn't guarantee how long it would take to
> > unwind to find the handler for the raised exception.  I wonder if such
> > conditions esist for exception handling in Java?  (Or dynamic
> > dispatch, for that matter)
> 
> Is it still the case that Ada exceptions are to be avoided for real-time?
> Not a good thing if true...
> 
> It seems clear that deterministic exception handling will be critical to
> real-time Java efforts.  I'm sure you can find out more about this at the
> various real-time Java websites if you're interested.
> 
> -----------== Posted via Deja News, The Discussion Network ==----------
> http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own

Exceptions are never to be used in Ada ( or Java) for any reason other
than to catch totaly unpredictable disasters. Java IOStreams will have
to be completely re-written if Java is to be used in hard real-time
systems. Until then don't use ^D or ^Z to  end an input stream.

-- 
Russ Lyttle, PE
<http://www.flash.net/~lyttlec>

Re: Real time Java? Huh?

[JP Thornley]

In article: <71vhli$hul$1@nnrp1.dejanews.com>  oopster@my-dejanews.com 
writes:
> 
> Is it still the case that Ada exceptions are to be avoided for 
real-time?
> Not a good thing if true...
> 

For high-integrity Ada, I've always been fascinated by the existence of 
the two opposed camps on exceptions - one group taking the view that 
having exceptions in the language makes it *more* suitable for 
high-integrity applications and the other regarding them as, not quite 
the work of the devil, but definitely not something that self-respecting 
software engineers would want to get involved with.

The Annex H Rapporteur Group spent a long time discussing the 
appropriate advice to give for exceptions in the (currently Draft) 
"Guide for the Use of the Ada Programming Language in High Integrity 
Systems" with the result that it describes three different approaches:-

1. Catch and handle all exceptions locally - so there is no exception 
propagation; program state must be well-defined after the handler has 
executed.

2. Use a single catch-all handler at the top level that makes almost no 
assumptions about program state - it probably does a restart.

3. Prove that no run-time errors can occur (so code can be compiled with 
all checks suppressed). [But you often still include the 'catch-all' 
handler to cope with hardware glitches/memory corruptions.]

Phil Thornley

-- 
------------------------------------------------------------------------
| JP Thornley    EMail jpt@diphi.demon.co.uk                           |
|                      phil.thornley@acm.org                           |
------------------------------------------------------------------------

Re: Real time Java? Huh?

[Richard IRVINE]

I see so many raging discussions on this newsgroup
(comp.lang.java.advocacy) about templates or named parameters that it is
refreshing to see contributions from the somewhat calmer Ada world, where
many of these problems were well thought out years if not decades ago.

GNAT for Java sounds interesting. I wonder whether GNAT for Java implies
coding in a subset of Ada, without pointers or access types and without
manual memory allocation/deallocation, or will we be free to carry across
dangling pointer and memory leakage problems from Ada into Java?

dewarr@my-dejanews.com wrote:

> In article <71ve5m$iv2$1@platane.wanadoo.fr>,
>   "Jean-Pierre Rosen" <rosen.adalog@wanadoo.fr> wrote:
>
>  BTW, can we deduce from your message that the
> > soon-to-be-delivered GNAT 3.11p will include
> > GNAT-for-Java ?
>
> No, not yet, GNAT-for-Java is still very much in the
> testing phase, but it is getting closer, we have some
> customers doing serious evaluations, and as I mentioned you
> will be able to see an interesting real-time demo at
> SigAda.
>
> -----------== Posted via Deja News, The Discussion Network ==----------
> http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own

Re: Real time Java? Huh?

[Stephen Leake]

Richard IRVINE <richard.irvine@eurocontrol.fr> writes:

> I see so many raging discussions on this newsgroup
> (comp.lang.java.advocacy) about templates or named parameters that it is
> refreshing to see contributions from the somewhat calmer Ada world, where
> many of these problems were well thought out years if not decades ago.
> 
> GNAT for Java sounds interesting. I wonder whether GNAT for Java implies
> coding in a subset of Ada, without pointers or access types and without
> manual memory allocation/deallocation, or will we be free to carry across
> dangling pointer and memory leakage problems from Ada into Java?

It's not "GNAT-for-Java", but rather "GNAT-for-JVM" (Java Virtual
Machine). Since the JVM allows pointers and access types, GNAT-for-JVM
will too (or at least I hope so - I don't have it yet !).
ObjectAda-for-JVM allows pointers and access types now; it supports
all of Ada (barring the many bugs, and most of the annexes :).

-- Stephe

Re: Real time Java? Huh?

[dennison]

In article <364874D1.D788CE9C@eurocontrol.fr>,
  Richard IRVINE <richard.irvine@eurocontrol.fr> wrote:
>
> GNAT for Java sounds interesting. I wonder whether GNAT for Java implies
> coding in a subset of Ada, without pointers or access types and without
> manual memory allocation/deallocation, or will we be free to carry across
> dangling pointer and memory leakage problems from Ada into Java?

Actually, I don't believe it is possible to have dangling pointers in Ada w/o
using one of the "unchecked" programming constructs. From what I know of
Java, I suspect the Java Virtual Machine wouldn't support Ada unchecked
programming constructs anyway.

--
T.E.D.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: Real time Java? Huh?

[dewarr]

In article <364874D1.D788CE9C@eurocontrol.fr>,
  Richard IRVINE <richard.irvine@eurocontrol.fr> wrote:
>
> I see so many raging discussions on this newsgroup
> (comp.lang.java.advocacy) about templates or named
parameters that it is
> refreshing to see contributions from the somewhat calmer
Ada world, where
> many of these problems were well thought out years if not
decades ago.
>
> GNAT for Java sounds interesting. I wonder whether GNAT
for Java implies
> coding in a subset of Ada, without pointers or access
types and without
> manual memory allocation/deallocation, or will we be free
to carry across
> dangling pointer and memory leakage problems from Ada
into Java?


GNAT-for-Java will support the entire Ada language, with
the exception of certain representation clauses that do not
make sense in the JVM environment.


-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: Real time Java? Huh?

[Jon S Anthony]

> GNAT-for-Java will support the entire Ada language, with the
> exception of certain representation clauses that do not make sense
> in the JVM environment.

This doesn't mean you are not going to use the GC in the JVM, right?
You will support completely GC style programming, right?  Admittedly
the current JVM GC is a complete piece of trash, but the claim is that
final 1.2 will have a "decent" generational scavenging collector.


/Jon

-- 
Jon Anthony
Synquiry Technologies, Ltd. Belmont, MA 02478, 617.484.3383
"Nightmares - Ha!  The way my life's been going lately,
 Who'd notice?"  -- Londo Mollari