Re: Digest Number 1840

View: New views
6 Messages — Rating Filter:   Alert me  

Parent Message unknown Re: Digest Number 1840

by Michael O'Dell-2 :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

One thing LLVM has in its favor is adoption by Apple

Apple *really* needs a programming language other than
Objectionable-C for writing MacApps. to this end, they
have created MacRuby which offers painless, transparent native access
to all the Obj-C frameworks, and with LLVM, it is well on its
way to generating machine code (AOT) using LLVM.

I think they also want a C/Obj-C compiler that is more stable.
it seems that with every new version of GCC, it takes 3 months
to get the FreeBSD/Darwin kernel to compile and run.
(that beggars my imagination but the people i know grinching
about it are not lightweights.)

in terms of run-time traversal of the AST to allow changes
to, for example, eval/uneval arg semantics, it brings to
mind the Lispers discussing "hygenic macros". it seems to me
that some of the reasons for doing run-time AST flanging
might could be extracted and canonified, so there was a way
for the compiler to know that some construct needs the full
generality of io, so no short-cuts need apply. but in other
cases, the compiler could be more aggressive about the optimization
it could do. i think it was one of the StrongTalk papers (or maybe
another paper on a duck-typed language? - i'm home with the flu
and am more addled than usual) that talked about optimizations
which could be "unoptimized" by back-patching more general code.
that way the compiler doesn't have to be perfectly right - the
runtime just needs to be able to tell the compiler guessed wrong.

not claiming for certain there's any value in the observed
similarity of concerns, but maybe there is in terms of sorting
through and identifying the issues.

        -mo




Re: Digest Number 1840

by Samuel A. Falvo II :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

On Thu, Oct 29, 2009 at 6:35 AM, Mike O'Dell <mo@...> wrote:
> Apple *really* needs a programming language other than
> Objectionable-C for writing MacApps. to this end, they

I think you mean "Objective-C".

As a C coder who is equally happy with Smalltalk, I find little to
complain about Objective-C, and really, REALLY, wish this kind of
slander would stop.

You've got beefs with the language?  Fine.  But, please respect those who don't.


--
Samuel A. Falvo II

Re: Digest Number 1840

by Guy Hulbert :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

On Thu, 2009-29-10 at 08:18 -0700, Samuel A. Falvo II wrote:
>  
> On Thu, Oct 29, 2009 at 6:35 AM, Mike O'Dell <mo@...> wrote:
> > Apple *really* needs a programming language other than
> > Objectionable-C for writing MacApps. to this end, they
>
> I think you mean "Objective-C".

Perhaps he means C++ ;-)

>
> As a C coder who is equally happy with Smalltalk, I find little to
> complain about Objective-C, and really, REALLY, wish this kind of
> slander would stop.

It was new and funny to me ...

... I've seen at least one article, which says that Obj-C is what C++
should have been.

>
> You've got beefs with the language? Fine. But, please respect those
> who don't.

I won't repeat the joke on this list then.

>
> --
> Samuel A. Falvo II
>
>
>
>
--
Guy Hulbert
gwhulbert@... (preferred)
work: (416) 391-2051 (no voicemail)
cell: (416) 738-6257 (voicemail)



Re: Digest Number 1840

by Friedrich :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

Guy Hulbert <gwhulbert@...> writes:

> On Thu, 2009-29-10 at 08:18 -0700, Samuel A. Falvo II wrote:
>>  
>> On Thu, Oct 29, 2009 at 6:35 AM, Mike O'Dell <mo@...> wrote:
>> > Apple *really* needs a programming language other than
>> > Objectionable-C for writing MacApps. to this end, they
>>
>> I think you mean "Objective-C".
I guess not, because the mail shows a dislike for Objective-C. Which I
can't understand. It's on of the really good extensions to C and it just
feels "right"
>
> Perhaps he means C++ ;-)
I wished, he did.

>
>>
>> As a C coder who is equally happy with Smalltalk, I find little to
>> complain about Objective-C, and really, REALLY, wish this kind of
>> slander would stop.
>
> It was new and funny to me ...
>
> ... I've seen at least one article, which says that Obj-C is what C++
> should have been.
That's absolutly something I would subscribe immediatly.

A "dream-duo" would be some IO/Objective-C I know there should something
in Io for that but I guess I never have been able to build it. Just
imagine instead of AppleScript we had Apple-Io. As we have MacRuby and
the like. I don't know if anyone has worked or tried automation with Io
on Apple, but I'm sure it would be nice.

Agreed Io has it's own kind of problems but it's really one of the most
beatiful languages out there. I wish I had known it 11 years ago, while
we were working on our own language. If we had known Io we surely could
have tried to make it "compilable", and it would have been nice. I'm
sure.

Regards
Friedrich


--
Q-Software Solutions GmbH; Sitz: Bruchsal; Registergericht: Mannheim
Registriernummer: HRB232138; Geschaeftsfuehrer: Friedrich Dominicus

Re: Digest Number 1840

by Steve Dekorte :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message


On 1 Nov 2009, at 12:40 am, Friedrich Dominicus wrote:
>> ... I've seen at least one article, which says that Obj-C is what C++
>> should have been.
> That's absolutly something I would subscribe immediatly.

Me too. I actually would have implemented Io in Objective-C if  
Objective-C was cross platform. (Btw, I know about the gnu Objective-
C but it really isn't compatible with Apple's and I doubt it ever  
will be.)

> Agreed Io has it's own kind of problems but it's really one of the  
> most
> beatiful languages out there. I wish I had known it 11 years ago,  
> while
> we were working on our own language. If we had known Io we surely  
> could
> have tried to make it "compilable", and it would have been nice. I'm
> sure.

Thanks! What is that language you were working on?

Re: Digest Number 1840

by Friedrich :: Rate this Message:

Reply to Author | View Threaded | Show Only this Message

Steve Dekorte <steve@...> writes:

> On 1 Nov 2009, at 12:40 am, Friedrich Dominicus wrote:
>>> ... I've seen at least one article, which says that Obj-C is what C++
>>> should have been.
>> That's absolutly something I would subscribe immediatly.
>
> Me too. I actually would have implemented Io in Objective-C if  
> Objective-C was cross platform. (Btw, I know about the gnu Objective-
> C but it really isn't compatible with Apple's and I doubt it ever  
> will be.)
Yes I'm afraid that's the case. But gnus objective C seems to work quite
similiar on every platform. But (this probably has to be written BUT),
the IDE support outside the Mac OS X is "near" non-existant. Please
don't tell me ProjectManager is the "answer".

>
>> Agreed Io has it's own kind of problems but it's really one of the  
>> most
>> beatiful languages out there. I wish I had known it 11 years ago,  
>> while
>> we were working on our own language. If we had known Io we surely  
>> could
>> have tried to make it "compilable", and it would have been nice. I'm
>> sure.
>
> Thanks! What is that language you were working on?
It was named Q
http://www.q-software-solutions.de/q/index.shtml

It was thought as a follow-up to Eiffel with streng separation of
commands and queries. The queries should have been pure functional.
To that time I thought OO is the best invention since sliced bread and
that Eiffel is the "right" thing. To my shame I have to conclude, I'd
better had been fluent in Smalltalk, Common Lisp. Yes I've looked into
many languages but I ignored the "dynamically" typed ones. That wasn't
very clever. But that's not the most unclever point. My idea was that
people would be willing to buy software on Linux also. Oh boy, how could
one be so mistaken...

Regards
Friedrich