In article <blake-927869.06434124062003[at]news.bellatlantic.net>,
Blake Patterson <blake[at]blakespot.com> wrote:

> Does "Xcode" replace current Project Builder? They never mentioned the
> words "Project Builder" or "Cocoa" for that matter in the WWDC keynote /
> demo. And looking at the Xcode site pages:
>
> http://www.apple.com/macosx/panther/xcode.html
>
> ...the page says:
>
> "Whether you¹re writing your app in C, C++ or Java, scripting in
> AppleScript or looking to migrate code from another legacy tool, you¹ll
> find that Xcode is as quick to forgive as it is to compile. Fast and
> easy does it every time."
>
> How about Objective-C?? No mention of what Apple had been pushing to be
> the premier dev language/framework for OS X. What gives?
>
> Thanks.
>
>
>
>
> bp

Yes, XCode is the successor. PB was basically a straight port from Next,
so its been rewritten now as part of Apple's crusade to revolutionize
all the important apps. And don't worry, cocoa/obj-c is going to be a a
central part of the OS X platform probably for the life of the OS.

Did you watch the keynote webcast? The part where Chris Espinoza (I
think!) demos XCode's rapid compile times and "Fix and Continue"
features blew me away. He made a bug fix to an application that was
already running, the application started compiling in the background
before he touched it, so that when he hit "Build" it was built almost
instantly, and the running application immediately reflected the code
patch. It was unbelievable--nothing short of it. This is going to
revolutionize the way we write code--and speed the hell out of the
process too.

--
|\/| /-| |2 |<
mehaase at sas dot upenn dot edu

"Mark Haase" <no[at]spam.please> wrote in message
news:no-347F79.14412524062003[at]netnews.upenn.edu...
> Did you watch the keynote webcast? The part where Chris Espinoza (I
> think!) demos XCode's rapid compile times and "Fix and Continue"
> features blew me away. He made a bug fix to an application that was
> already running, the application started compiling in the background
> before he touched it, so that when he hit "Build" it was built almost
> instantly, and the running application immediately reflected the code
> patch. It was unbelievable--nothing short of it. This is going to
> revolutionize the way we write code--and speed the hell out of the
> process too.

I remember using that feature in Microsoft Visual C++ 6 years
ago. It's nice, but not revolutionary.

I did not, however, use it all that extensively. You could not introduce
new types or methods while the program was running, which put a
fundamental limit on how much you could do.

I suspect Apple's effort will have similar limitations. I have a hard
time imagining how to get C++ to tolerate more extensive
changes at runtime.

"Dan Johnson" <danieljohnson[at]vzavenue.net> wrote in message news:<vfhhnl7lvr8r49[at]news.supernews.com>...
> "Mark Haase" <no[at]spam.please> wrote in message
> news:no-347F79.14412524062003[at]netnews.upenn.edu...
> > Did you watch the keynote webcast? The part where Chris Espinoza (I
> > think!) demos XCode's rapid compile times and "Fix and Continue"
> > features blew me away. He made a bug fix to an application that was
> > already running, the application started compiling in the background
> > before he touched it, so that when he hit "Build" it was built almost
> > instantly, and the running application immediately reflected the code
> > patch. It was unbelievable--nothing short of it. This is going to
> > revolutionize the way we write code--and speed the hell out of the
> > process too.
>
> I remember using that feature in Microsoft Visual C++ 6 years
> ago. It's nice, but not revolutionary.
>
> I did not, however, use it all that extensively. You could not introduce
> new types or methods while the program was running, which put a
> fundamental limit on how much you could do.
>
> I suspect Apple's effort will have similar limitations. I have a hard
> time imagining how to get C++ to tolerate more extensive
> changes at runtime.

Dunno. They're doing something tricky with dyld (dynamic code loading)
though.

It looks like Building 1 has been cranking the code on the compiler.
(It helps to have a campus full of people bitching at you every day.)

I re-upped my ADC membership, looking forward to the beta.
ProjectBuilder was just too frickin' sucky for daily use.

=Heywood=

"Dave Seaman" <dseaman[at]no.such.host> wrote in message
news:bdca4f$1dt$4[at]mozo.cc.purdue.edu...
> On Tue, 24 Jun 2003 17:47:29 -0400, Dan Johnson wrote:
> > I suspect Apple's effort will have similar limitations. I have a hard
> > time imagining how to get C++ to tolerate more extensive
> > changes at runtime.
>
> That might be true of C++, but Objective-C is an altogether different
> matter. I see no reason why you couldn't introduce new types or methods
> while the program is running.

Well, I suppose that is true for classes, though you could not, I think,
change a class type's members very easily.

But do we even know this fix & continue deal will support
Objective-C? I observe that Microsoft's equivalent does not
work for anything but C++.

"Dan Johnson" <danieljohnson[at]vzavenue.net> wrote in message
news:vfk321lc74015a[at]news.supernews.com...
> "Dave Seaman" <dseaman[at]no.such.host> wrote in message
> news:bdca4f$1dt$4[at]mozo.cc.purdue.edu...
> > On Tue, 24 Jun 2003 17:47:29 -0400, Dan Johnson wrote:
> > > I suspect Apple's effort will have similar limitations. I have a hard
> > > time imagining how to get C++ to tolerate more extensive
> > > changes at runtime.
> >
> > That might be true of C++, but Objective-C is an altogether different
> > matter. I see no reason why you couldn't introduce new types or methods
> > while the program is running.
>
> Well, I suppose that is true for classes, though you could not, I think,
> change a class type's members very easily.
>
> But do we even know this fix & continue deal will support
> Objective-C? I observe that Microsoft's equivalent does not
> work for anything but C++.

Which makes since, since Objective-C isn't used on their platorm. C++ is.
Actually,
I'll have to check my copy of Digital Fortran....

JCS

On Wed, 25 Jun 2003 16:55:23 -0400, Dan Johnson wrote:
> "Dave Seaman" <dseaman[at]no.such.host> wrote in message
> news:bdca4f$1dt$4[at]mozo.cc.purdue.edu...
>> On Tue, 24 Jun 2003 17:47:29 -0400, Dan Johnson wrote:
>> > I suspect Apple's effort will have similar limitations. I have a hard
>> > time imagining how to get C++ to tolerate more extensive
>> > changes at runtime.
>>
>> That might be true of C++, but Objective-C is an altogether different
>> matter. I see no reason why you couldn't introduce new types or methods
>> while the program is running.

> Well, I suppose that is true for classes, though you could not, I think,
> change a class type's members very easily.

Macintosh Common Lisp can do it. I don't see why Objective-C can't.
There are some similarities in the way objects are treated in the two
languages.

> But do we even know this fix & continue deal will support
> Objective-C? I observe that Microsoft's equivalent does not
> work for anything but C++.

Objective-C is the primary development language for the Mac. Apple
wouldn't spend time developing features in support of other languages
while forgetting their mainstay.


--
Dave Seaman
Judge Yohn's mistakes revealed in Mumia Abu-Jamal ruling.
<http://www.commoncouragepress.com/index.cfm?action=book&bookid=228>

In article <vfk321lc74015a[at]news.supernews.com>,
"Dan Johnson" <danieljohnson[at]vzavenue.net> wrote:

> "Dave Seaman" <dseaman[at]no.such.host> wrote in message
> news:bdca4f$1dt$4[at]mozo.cc.purdue.edu...
> > On Tue, 24 Jun 2003 17:47:29 -0400, Dan Johnson wrote:
> > > I suspect Apple's effort will have similar limitations. I have a hard
> > > time imagining how to get C++ to tolerate more extensive
> > > changes at runtime.
> >
> > That might be true of C++, but Objective-C is an altogether different
> > matter. I see no reason why you couldn't introduce new types or methods
> > while the program is running.
>
> Well, I suppose that is true for classes, though you could not, I think,
> change a class type's members very easily.
>
> But do we even know this fix & continue deal will support
> Objective-C? I observe that Microsoft's equivalent does not
> work for anything but C++.
>
>

Why would they talk about it if it didn't support Objective-C? Why in
the world wouldn't they at least release Xcode to all of us?

"James Stutts" <stuttjc[at]knology.net> wrote in message news:<vfkj4hka1di308[at]corp.supernews.com>...
> "Dan Johnson" <danieljohnson[at]vzavenue.net> wrote in message
> news:vfk321lc74015a[at]news.supernews.com...
> > "Dave Seaman" <dseaman[at]no.such.host> wrote in message
> > news:bdca4f$1dt$4[at]mozo.cc.purdue.edu...
> > > On Tue, 24 Jun 2003 17:47:29 -0400, Dan Johnson wrote:
> > > > I suspect Apple's effort will have similar limitations. I have a hard
> > > > time imagining how to get C++ to tolerate more extensive
> > > > changes at runtime.
> > >
> > > That might be true of C++, but Objective-C is an altogether different
> > > matter. I see no reason why you couldn't introduce new types or methods
> > > while the program is running.
> >
> > Well, I suppose that is true for classes, though you could not, I think,
> > change a class type's members very easily.
> >
> > But do we even know this fix & continue deal will support
> > Objective-C? I observe that Microsoft's equivalent does not
> > work for anything but C++.
>
> Which makes since, since Objective-C isn't used on their platorm.

Actually it is, they're just calling it C# ;)

=Heywood=

"James Stutts" <stuttjc[at]knology.net> wrote in message
news:vfkj4hka1di308[at]corp.supernews.com...
> > But do we even know this fix & continue deal will support
> > Objective-C? I observe that Microsoft's equivalent does not
> > work for anything but C++.
>
> Which makes since, since Objective-C isn't used on their platorm. C++ is.
> Actually, I'll have to check my copy of Digital Fortran....

Other languages are used on their platform; VB and recently C# for
two. But this fix and continue business only works with C++, at
least with MS's tools.

"Dave Seaman" <dseaman[at]no.such.host> wrote in message
news:bddteq$o0l$2[at]mozo.cc.purdue.edu...
> On Wed, 25 Jun 2003 16:55:23 -0400, Dan Johnson wrote:
> > Well, I suppose that is true for classes, though you could not, I think,
> > change a class type's members very easily.
>
> Macintosh Common Lisp can do it. I don't see why Objective-C can't.
> There are some similarities in the way objects are treated in the two
> languages.

Objective-C is a much more static languages than Lisp; it bakes
offsets to members directly into the code. Any change to the
field layout would mean updating every method that used that
field.

Not impossible, but not easy.

> > But do we even know this fix & continue deal will support
> > Objective-C? I observe that Microsoft's equivalent does not
> > work for anything but C++.
>
> Objective-C is the primary development language for the Mac. Apple
> wouldn't spend time developing features in support of other languages
> while forgetting their mainstay.

You may be right. I hope you are; but there may be technical obstacles
that neither of us knows about.

We'll see.

"Heywood Mogroot" <imouttahere[at]mac.com> wrote in message
news:dd5de929.0306260005.61ebfef2[at]posting.google.com...
> > > But do we even know this fix & continue deal will support
> > > Objective-C? I observe that Microsoft's equivalent does not
> > > work for anything but C++.
> >
> > Which makes since, since Objective-C isn't used on their platorm.
>
> Actually it is, they're just calling it C# ;)

C# is very different from Objective-C.

But it may be that if C# produced native code they'd
have made their own fix & continue work for it; if so
we may suppose Apple will make it work for Objective-C.

"Dan Johnson" <danieljohnson[at]vzavenue.net> wrote in message news:<vfliuvenjqta53[at]news.supernews.com>...
> "Heywood Mogroot" <imouttahere[at]mac.com> wrote in message
> news:dd5de929.0306260005.61ebfef2[at]posting.google.com...
> > > > But do we even know this fix & continue deal will support
> > > > Objective-C? I observe that Microsoft's equivalent does not
> > > > work for anything but C++.
> > >
> > > Which makes since, since Objective-C isn't used on their platorm.
> >
> > Actually it is, they're just calling it C# ;)
>
> C# is very different from Objective-C.

Metaphysically, C# was inspired by Java, which was inspired by
Objective C.

Syntactically, they're nothing alike, but deep down inside there's a
lot of philosophical similarity.

> But it may be that if C# produced native code

implementation detail.

> they'd
> have made their own fix & continue work for it; if so
> we may suppose Apple will make it work for Objective-C.

At apple, the compiler people and the system people used to work on
the same floor of AC2. Apple has no excuses for not producing a
kick-ass runtime; a small number of engineers own it top to bottom.

Unlike Micorsoft, where probably 3,000 people are on the cc list for
engineering change talk.

=Heywood=

In article <dd5de929.0306260445.71ccbd54[at]posting.google.com>,
imouttahere[at]mac.com (Heywood Mogroot) wrote:

> "Dan Johnson" <danieljohnson[at]vzavenue.net> wrote in message
> news:<vfliuvenjqta53[at]news.supernews.com>...
> > "Heywood Mogroot" <imouttahere[at]mac.com> wrote in message
> > news:dd5de929.0306260005.61ebfef2[at]posting.google.com...
> > > > > But do we even know this fix & continue deal will support
> > > > > Objective-C? I observe that Microsoft's equivalent does not
> > > > > work for anything but C++.
> > > >
> > > > Which makes since, since Objective-C isn't used on their platorm.
> > >
> > > Actually it is, they're just calling it C# ;)
> >
> > C# is very different from Objective-C.
>
> Metaphysically, C# was inspired by Java, which was inspired by
> Objective C.
>
> Syntactically, they're nothing alike, but deep down inside there's a
> lot of philosophical similarity.
>
> > But it may be that if C# produced native code
>
> implementation detail.
>
> > they'd
> > have made their own fix & continue work for it; if so
> > we may suppose Apple will make it work for Objective-C.
>
> At apple, the compiler people and the system people used to work on
> the same floor of AC2. Apple has no excuses for not producing a
> kick-ass runtime; a small number of engineers own it top to bottom.

Do you work for Apple?

Vareck Bostrom <v.bostrom[at]attbi.com> wrote in message news:<v.bostrom-285554.06111326062003[at]netnews.attbi.com>...
> In article <dd5de929.0306260445.71ccbd54[at]posting.google.com>,
> imouttahere[at]mac.com (Heywood Mogroot) wrote:
>
> > "Dan Johnson" <danieljohnson[at]vzavenue.net> wrote in message
> > news:<vfliuvenjqta53[at]news.supernews.com>...
> > > "Heywood Mogroot" <imouttahere[at]mac.com> wrote in message
> > > news:dd5de929.0306260005.61ebfef2[at]posting.google.com...
> > > > > > But do we even know this fix & continue deal will support
> > > > > > Objective-C? I observe that Microsoft's equivalent does not
> > > > > > work for anything but C++.
> > > > >
> > > > > Which makes since, since Objective-C isn't used on their platorm.
> > > >
> > > > Actually it is, they're just calling it C# ;)
> > >
> > > C# is very different from Objective-C.
> >
> > Metaphysically, C# was inspired by Java, which was inspired by
> > Objective C.
> >
> > Syntactically, they're nothing alike, but deep down inside there's a
> > lot of philosophical similarity.
> >
> > > But it may be that if C# produced native code
> >
> > implementation detail.
> >
> > > they'd
> > > have made their own fix & continue work for it; if so
> > > we may suppose Apple will make it work for Objective-C.
> >
> > At apple, the compiler people and the system people used to work on
> > the same floor of AC2. Apple has no excuses for not producing a
> > kick-ass runtime; a small number of engineers own it top to bottom.
>
> Do you work for Apple?

used to...

=Heywood=

"Heywood Mogroot" <imouttahere[at]mac.com> wrote in message
news:dd5de929.0306260445.71ccbd54[at]posting.google.com...
> > > Actually it is, they're just calling it C# ;)
> >
> > C# is very different from Objective-C.
>
> Metaphysically, C# was inspired by Java, which was inspired by
> Objective C.

Well, there may be something to that, but somehow or other
Java came out a lot like Simula.

> Syntactically, they're nothing alike, but deep down inside there's a
> lot of philosophical similarity.

I 'spose they are more alike than, say, C and Lisp.

> > But it may be that if C# produced native code
>
> implementation detail.

I wonder if that's true. Managed code is kinky stuff.

Suppose I use System.Reflection.Emit to dynamically
create a subclass for class Foo, and this subclass overrides
a method bar and accesses a field baz.

What is fix & continue gonna do if I delete the method or
the field?

The only answer I can think of is "give up and refuse to
apply the change"

> > they'd
> > have made their own fix & continue work for it; if so
> > we may suppose Apple will make it work for Objective-C.
>
> At apple, the compiler people and the system people used to work on
> the same floor of AC2.

What's "AC2"?

> Apple has no excuses for not producing a
> kick-ass runtime; a small number of engineers own it top to bottom.

They have an excuse: they inherited it from NeXT, which
did not have the resources to pull of something like .NET or
Sun's JVM.

But I hope to see improvements in future.

> Unlike Micorsoft, where probably 3,000 people are on the cc list for
> engineering change talk.

:D

"Dan Johnson" <danieljohnson[at]vzavenue.net> wrote in message
news:vflionj0eso9fd[at]news.supernews.com...
> "James Stutts" <stuttjc[at]knology.net> wrote in message
> news:vfkj4hka1di308[at]corp.supernews.com...
> > > But do we even know this fix & continue deal will support
> > > Objective-C? I observe that Microsoft's equivalent does not
> > > work for anything but C++.
> >
> > Which makes since, since Objective-C isn't used on their platorm. C++
is.
> > Actually, I'll have to check my copy of Digital Fortran....
>
> Other languages are used on their platform; VB and recently C# for
> two. But this fix and continue business only works with C++, at
> least with MS's tools.

I will have to try it with CVF.

JCS

>
>

In article <dd5de929.0306262144.64182151[at]posting.google.com>,
imouttahere[at]mac.com (Heywood Mogroot) wrote:

[at]reload the binary. We're only scratching the surface of fix &
[at]continue. The era of "recompiling" code changes is receding fast.

Yeah, we're almost on par with Smalltalk! (:

--
B.B. --I am not a goat! thegoat4 at airmail.net
Regime change begins at home.