[erlang-questions] Erlang vs Clojure

[G Bulmer]

On 28 Nov 2007, at 16:31, Alex Arnon wrote:
> I remember seeing some posts about inroads being made with regard  
> to realtime performance on JVMs lately, on Lambda the Ultimate.
> On the other hand, listening to Software Engineering Radio ( this  
> one I think), the expert being interviewed said that Java is  
> practically nonexistent in the realtime space. So... grains of salt  
> with everything, as usual... :)
I would like pointers to realtime any-VM stuff. For me, that is the  
last major piece of my jigsaw.
Were you thinking of the IBM paper: http://domino.research.ibm.com/ 
comm/research_people.nsf/pages/dgrove.ecoop07.html
which popped up at LtU as http://lambda-the-ultimate.org/node/2393 ?  
That seemed to be an impressive piece of work.

I feel realtime Java suffers from the classic 'chicken and the egg'  
syndrome - process control is pretty conservative about technologies  
because the systems are usually safety critical. So there will be few/ 
no volunteers to use a realtime-anything-new until some others have  
successfully deployed. This will take a few years due to the  
deployment environment, e.g. an oil refinery or power plant, (by  
which time the hype will have died down on that technology, and  
everyone will be evangelising and 'dis-ing' the next over-hyped  
technology ;-)

I know *I* feel the same way about highly available and fault- 
tolerant systems, and their initial implementation can be much  
quicker than process control.

This inherent conservatism as systems become more critical is why, to  
me, Erlang is way ahead of the field; there are real systems out  
there working!

GB

PS -
I do expect the 'tweaked' versions of Java, and .NET, to become  
usable*, but they have a lot of ground to cover before they catch up  
to Erlang.
*usable - ill defined, and depends on many factors, some of which are  
not rational ;-)

> On Nov 28, 2007 4:23 AM, G Bulmer < gbulmer@REDACTED> wrote:
>
> ... James Gosling ... was very excited about a research JVM which  
> was being built and tested
> for use in hard realtime, the project he talked about as specifically
> for process control in a power plant.
>
> Anyway, in his initial presentation ...
> showed stats from the research JVM that indicated good support for
> many thousands of threads (I accept, still well behind the Erlang VM,
> but I'd need more detail to judge how far behind) with interrupt
> response times with a jitter of less than 4 micro-seconds  (I am sure
> this was at least response to timer events, and may have been to
> other interrupt sources).
>
> I am certain I asked about garbage collection, and he said that the 4
> micro-seconds was not an overall average, but was 'realtime', and it
> included garbage collection.
>
> ...
> GB
> ...
> > Date: Mon, 26 Nov 2007 09:43:18 +0000
> > From: Charles Forsyth < forsyth@REDACTED>
> > Subject: Re: [erlang-questions] Erlang vs Clojure
> >> The thought of running on the JVM makes me feel both sick and
> >> bloated...but strangely I am warming up to the idea.
> ...
> >
> > if you expect to get high-performance concurrency from a JVM
> > i expect you to be disappointed.  the Erlang implementation
> > guarantees cheap processes everywhere it exists.  by contrast,
> > you get whatever your particular JVM gives you, which might not
> > be much, and it will vary from platform to platform.
> > if it works for you, though ...
> > i'd be particularly interested if you find JVMs that
> > offer ``massive local concurrency''....