E2E: Erik Meijer and Burton Smith - Concurrency, Parallelism and Programming

Love these Charles!

If you ever need an understudy Charles...

This video was very enjoyable; lots of humor and insight

I love the idea of ultra-cheap cross-core/processor communication that facilitates extremely fine-grained parallelism. Meanwhile maybe some form of complexity and strictness analysis will help determine sensible concurrency granularity given pure semantics.

Quote of the talk: "dysfunctional programming" - a brilliant way to frame every other kind of programming. Not serious, just fun, heh

Always love the Burton videos.  They are right up there with the Beckman videos in the must watch category ... you wish they would just keep going for a few more hours. 

Good stuff.

In terms of exceptions as values, at the 500 foot level, it would seem if Object had a new property to "hold" exception, the type system would/could just work.  You could use normal try/catch or not as needed.  In a message passing model, I would tend to think all non-void functions need to return a type even if that type is an exception. 

you know, i dont think the awnser to charles question "will there be new languages" is that obvious.. arent java and c# really just c++ witch in turn is really just c? arent f# just ocaml witch in turn is ml? haskell and smalltalk are really old too.

i wonder if any completly new languages that arent based on anything exsisting will emerge in the forseeable future. sure, the ones we have will continue to evolve and fork, but will there be a completly new general purpose language? the awnser is less obvious at least

Actually, it's quite obvious. C#, Java, C++, C are sugar coated assembler. Reasoning about assembler, even sugar coated is a lost cause. Making those languages into something that can be reasoned about at compile and especially run-time would be practically impossible because of long hairy legacy that those languages carry around.

In order to run a program on a parallel hardware, run-time would have to reason about side effects to come up with some strategy to partition computational graph into work loads that have minimal interactions between each other.

If many core processors will have cores of different capabilities (which seems to be the case), run-time reasoning and JIT will be a necessity.

It seems like none of the existing imperative languages would survive transition to parallel era. Of course run-times are still be written in something that is sugar coated assembly, yet for general-purpose programming completely new languages would be required.

Declarative and richly typed presumably.

Also to the point of run-time reasoning and code generation, to provide fault tolerance computational graph might need to be re-evaluated if a computation node returns exceptional value or goes into non-termination state. That in theory would allow automatic remediation for run-away queries in databases and handling of non-responding services in the cloud (as well as mutating hardware - failed or hot plugged general and special purpose CPUs, failed or hot plugged memory and so on).

It probably will take another 10 to 20 years to get it right, but it looks like that's where things are going.

I think the language style that you are referring to is akin to the Flow-Based Programming languages of component processes. Since FBP dates back to the 1970's it would seem to backup Al_'s assertion that what we think are new programming languages are just nice facades on older ideas...  http://en.wikipedia.org/wiki/Flow-based_programming

Right, and old ideas are just facades on even older ideas and so on recursively till the big bang

In the end it will be about believes, whether one believe this or that language being "new" or not. Attempts to define "pure novelty" would end up nowhere.

On the subject of strict or linient evaluation.

It seems that an adanced enough run-time can and should use both, based on the accumulated "knowledge" (stats) about workloads being executed.

Expectation that something can be strictly evaluated in false in absolute sense, because each and every CPU instruction and/or memory read/write may fail because of faulty hardware. Yet, it can be statistically true. If hardware is somehow known to be 99.something% reliable, such assumption can be made safely (in statistical sense), otherwise nothing can be computed or done ever.

(I believe that proponents of strict evaluation are stuck because they base their reasoning on incorrect assumptions without explicitly stating what those assumptions are, which is a known issue that plagued physics for centuries, and most likely still does)

The same must apply to the algorithms as well. If algorithm is known to be predictable on a given workload (either statistically or by devine intervention of the mister human), it's OK to evaluate is strictly. If there is no prior knowledge, lazy evaluation is the way to go and please gather execution stats upon exit so it can be reused in the future evaluations/executions. And if it does not exit in the requested amount of time - abandon (preferrably kill first) the execution and and black list it (till the end of time or the next devine intervention).

From 10000 feet it looks like a nice logical schema with a feed back loop, which is statistically a necessity for each and every successful eco system (observe the nature).

"I think the language style that you are referring to is akin to the Flow-Based Programming languages of component processes"

It seems much of todays concurrent lineup (i.e. CCR, Axum, Erlan, TPL, functional programing, etc) have discovered or re-discovered the same things (i.e. black boxes w/ msgs).  At the base level, it seems this guy nailed it back in the 70s.  The right road seems to float around the FBP ideas.  Add hw support for efficient message passing (as Burton points out) and maybe even some kind of hw support for sw bounded queues, and things get interesting.  Add correct-by-construction language support (i.e. Axum and beyond) and it gets real interesting.  The syntax is not the important thing, it is the general model that must lead you down the correct path and makes the wrong path hard (i.e. the reverse of today).

Staceyw,

I totally agree, but the CCR is special in the list that you present since if you look beyond the CCR, in either the MSRS or DSS/CCR Toolkit, you will find a very nice VPL IDE that really nails down the graphical ideas of FBP..

Too bad that all this goodness from "BigTop" just slipped past most of the .NET world.. 

Could these points be manifest in Google's Go langage? It appears to implement pretty much every issue raised. Funny on the timing.

One final thing, I also find myself writing functional style in .net. Especially when doing recursion. Though I'm by no means a functional language programmer.

I don't quite follow the reason on the white board that Erik wrote.

He's saying that if you have a function

[code] F( x ) { return 13; }[/code]

(so basically, F always returns 13, regardless of what you pass in)

and then you call F like so:

F( E )

would it retrun 13?  One would say, "sure, because it doesn't matter what you pass in"

But his point is that, what if (for instance), the parameter E throws an exception.  That means F doesn't return 13 (or doesn't even get called)

That means, you cannot replace an arbitrary instance of "F( E )" with "13".

But I don't understand.   This is supposed to be purely-functional.  If I say that F returns 13 regardless of the parameter, I would want (and may be expect) that the compiler would not bother evaluating the parameter E to begin with.

That is, even if I were to call F( 1/0 ), I could make the argument that 13 should be returned because ultimately, it's about evaulating F, not evaluating the the parameter to F, whose own evaluation is there solely for the purpose of being passed to F.

That sounds right, but just suppose that F really was using x, then it would probably need to evaluate x and in that case the return type would be different if x threw an exception.