[Info-vax] [OT] Wirth style languages, was: Re: Obscure Ada compiler vendors?

[Stephen Hoffman]

On 2013-04-04 19:52:23 +0000, Craig A. Berry said:

> In article <kjkbh6$bac$1 at dont-email.me>,
>  Stephen Hoffman <seaohveh at hoffmanlabs.invalid> wrote:
> 
>> As good as it is, Haskell wouldn't be at the top of my list for an
>> operating system.  Same with Scala or Clojure.  Good languages both,
>> but you probably don't want the JVM underneath an operating system, if
>> you can avoid it.
> 
> Unlike the other functional languages, Haskell, as far as I can tell,
> is a true compiled language and does not run on a virtual machine like
> the others do.

The antecedent to "both" was Scala and Clojure.   Not Haskell.

AFAIK, Haskell only fairly recently acquired abilities around 
mixed-language programming work, and that'd be a fair-sized issue with 
an operating system, and from what little I've encountered of it, the 
foreign function stuff is still seems a little too limited.  While it's 
an entirely viable and good choice for many apps, Haskell "wouldn't be 
at the top of my list for an operating system".

> The very size of the code base is one of the reasons I suggested 
> something more concise, like Haskell.  Still not in any way realistic, 
> but not quite as hopeless.

If the design and features move significantly forward, a number of the 
current and some potential future users can see the benefits of 
migrating.  If you're going to be making an investment of this scale, 
you need to provide folks with good reasons to move, and you also want 
or need to remove at least some of the existing roadblocks 
(performance, scaling, features, etc) lurking within the current 
implementations.

Without the incentive to move, a number of folks simply won't move 
forward, and folks won't target the environment.

This whole discussion is replete with trade-offs, though.

If the results of the investment and the programming effort aren't 
sufficiently compelling, there's little incentive to move forward, or 
to migrate to the new environment.

Change too much, and some folks don't port.

Don't change enough, and some other folks don't port.

Don't swap out enough of the existing "broken" stuff, and you will need 
to expend significantly more effort to make those changes, or you won't 
be able to reasonably make the changes.

Swap out too much of the "broken" stuff, and you'll spend even more 
time and effort sorting out the secondary effects.

These days, I'd be looking more for a Rhapsody-scale 
<http://en.wikipedia.org/wiki/Rhapsody_(operating_system)> change, and 
not a "straight" port, or an open-source clone.  But that's a very 
large effort.

But I don't see any of this happening.

-- 
Pure Personal Opinion | HoffmanLabs LLC