Board index » kylix » Re: The future..

Re: The future..


2004-01-14 12:08:21 AM
kylix1
"pNichols" wrote:
[snip]
Quote
C/C++ will always be around, since though maybe not as
pretty as a more strongly typed language, C/C++ are ideal for low level
programming.
I agree with your message except for the last part. C and C++ are two
distinct languages. C++ doesn't cover the C99 specs. C is weak typed and C++
is strong typed. You can use C++ perfectly well for application development
due to the existence of abstraction layers like STL, boost, VCL, wxWindows,
Qt, MFC and Nokia Series60 or UIQ (for Symbian) etc.
Peter
 
 

Re:Re: The future..

..Network computing..
A example business model in the public domain that could make this
succeed could be to have an organisation (even governmental) that
leases/rents out gadgets with software or users pay for usage. This
Organisation could also rent out super computing in a grid computing
acrchitechture using this same hardware/gadgets.
or
A enterprise can have a similar setup internally - no PC's only a gadget
per employee - savings could be enourmous!
Some universities are already employing grid computing over the
internet. SETI is also an example.
Does it make sense? I think this should be feasible in a few years from now.
Siegs
 

Re:Re: The future..

JQP wrote:
Quote
Maybe there will be primarily one framework, maybe two, maybe more. This is
all relatively far off in the future. We'll just have to wait and see what
happens.
The gadget manufacturers may hold the key as to which framework to ship
use on their gadgets.. Users won't care to much. It does not help MS to
compete against the main gadget producers with XBox <g>We might see MS
bringing out their own gadgets to compensate.
Time will tell, I just feel we are at the doorstep of a revolution in
computing..
Siegs
 

{smallsort}

Re:Re: The future..

pNichols wrote:
Quote

It is certainly true that embedded devices will be an important part of the
IT structure, but Java and NET are positioning themselves for that market
as well.
I would refer to this as the intermediate market (2-3 years?) i.e. in
the current climate of uncertainty and Gadget hardware immaturity..
Quote
I would definitely give Java the nod here, as they are better
positioned (xplatform and j2ME has improved dramaticlly), and C++. I think
Borland is pursuing the right course with Builder X.
Agreed
Quote
I like Delphi and I like Kylix concept. But we all have to be realistic.
Lets wait for the Builder X GUI developer - that would be a good
alternative to Kylix. So long as Builder X also implements a model
driven development feature. QT developer also looks interesting.
Delphi still has a enough life to cover the 'intermediate' period I
refer to. A lot can happen during. I feel sorry for people who have to
start new large, long lifecycle, commercial projects right now. C++ or
Java seem to make the most sense for me...otherwise Delphi (current
win32)/Kylix still makes 100% sense for the rest (depending on your
target as well).
Quote
.. I see it as nothing more than MS' Platform specific Java;-- more of a J++ iteration 2. But I do
not see any evidence whatsoever, of anything on the horizon now, toppling
either Java or NET. C/C++ will always be around, since though maybe not as
pretty as a more strongly typed language, C/C++ are ideal for low level
programming.
Agreed
 

Re:Re: The future..

"SiegfriedN" < XXXX@XXXXX.COM >wrote in message
Quote
Time will tell, I just feel we are at the doorstep of a revolution in
computing..
That a pretty radical thought, particularly for this group.
Most people here seem to be waiting for the marketplace to digress and adopt
an OS whose design dates back to the 1960s.
 

Re:Re: The future..

Peter Agricola wrote:
Quote

"pNichols" wrote:
[snip]
>C/C++ will always be around, since though maybe not as
>pretty as a more strongly typed language, C/C++ are ideal for low level
>programming.

I agree with your message except for the last part. C and C++ are two
distinct languages. C++ doesn't cover the C99 specs. C is weak typed and
C++ is strong typed. You can use C++ perfectly well for application
development due to the existence of abstraction layers like STL, boost,
VCL, wxWindows, Qt, MFC and Nokia Series60 or UIQ (for Symbian) etc.

I would agree with that. I lumped them together because that is how they are
primarily thought of and unfortunately used <G>.
Hopefully that will change and more developers will start using C++ they way
it was intended to be used, not as an additional layer to C.
We primarily program in Java these days, but there are times when I need
some low level functionality that C or C++ affords us quite well. C++ is
not nearly as strongly typed however as say Java or Delphi. It can be, if
one uses Classes/Objects and STL, but how many C+ programs have we seen
that are not interspersed with structs and the like?
 

Re:Re: The future..

SiegfriedN wrote:
Quote
pNichols wrote:
>
>On where the future truly is, at least for the forseeable future anyways:
>
>(1) Java (2) NET (3) C/C++

Yes, maybe for the next 2-3 years?

Thats the problem with prognostication of course, we can only realistically
look at where we are now, to try and make a prediction for the near future.
Right now, I do not see NET or Java going anywhere but up. If one fails, I
would think it to be NET, not Java. Java has too strong a user base and
maturity on its side. NET has less of a chance, IMHO, due to the fact it is
basically an MS Java. A case has to be made of why NET is better than Java,
and I fail to see any reasons why it is better or accomplishes any mission
that Java was designed to accomplish, other than GC, which is not main
impetus for Java in the first place. I think NET is good for Java in a
competition type phase, but until I see a useful NET application that will
run on many platforms from a common code base, NET is nothing more than a
MS attempt to slow Java migration.
Quote
I would put C/C++/Kylix as (1) after +- 3 years. (Delphi language is
more RAD)

How many people actually use Pascal these days? Kylix did not do well on
Linux for that main reason, IMHO. That doesn't mean Pascal is not a very
good language, but I do not see the halls of IT lined with people wanting
to learn Pascal or OOP Pascal. I do see them line the hallways learning
Java, C++, and some new interest in learning C#.
Quote

The 'gadget producers' are the best placed for the future i.m.o. This is
with whom Borland should flirt...not MS. MS will stop creating OS's,
servers, etc. They would be creating end user apps to run on these
gadgets - unless they produce their own hardware as well. Borland can
provide the compilers/ dev tools. They will have a hard time competing
with OSS tools however!

MS will not stop creating an OS for a long time. It would be the death of a
large monolithic MS to lose the OS market. That is their {*word*241}hold on
the industry. They could lose it in say 8 years, but if they do, then they
will have to find new markets.
Without Windows, what good is MS? They would have to play with everyone else
then and compete with the industry on the basis of developing superior
products. They are not known for this, they are know for ease of use, and
being cheaper, but not superiority, nor are they known for innovation (even
though they like to claim otherwise). In fact, what has MS ever invented
other than Visual Basic?
 

Re:Re: The future..

"pNichols" wrote:
Quote
We primarily program in Java these days, but there are times when I need
some low level functionality that C or C++ affords us quite well. C++ is
not nearly as strongly typed however as say Java or Delphi. It can be, if
one uses Classes/Objects and STL, but how many C+ programs have we seen
that are not interspersed with structs and the like?
I don't see C++ is less strongly typed than Delphi. I don't see how a
library can make a weak language strong typed. There is no difference
between a struct and a class, except the members of a struct are public by
default where the members of a class are private by default.
Peter
 

Re:Re: The future..

"SiegfriedN" wrote:
[snip]
I don't feel the need to google how many times the dead of the dinosaurs
(mainframes) is predicted.
Peter
 

Re:Re: The future..

JQP wrote:
Quote
"SiegfriedN" < XXXX@XXXXX.COM >wrote in message
news: XXXX@XXXXX.COM ...
>Time will tell, I just feel we are at the doorstep of a revolution in
>computing..

That a pretty radical thought, particularly for this group.

Most people here seem to be waiting for the marketplace to digress and
adopt an OS whose design dates back to the 1960s.
A system that, BTW, runs the largest computers in the world, reliably.
Many changes have been instituted to the core of Unix and Linux over the
years. If we use that rationale, then Windows XP is nothing more than a
rehash of Windows 1.0.
 

Re:Re: The future..

"pNichols" < XXXX@XXXXX.COM >wrote in message
Quote
A system that, BTW, runs the largest computers in the world, reliably.
Now all you've got to do is convince people to put "the largest computers in
the world" on their desktops. Then we'll all have jobs doing Linux work<g>.
 

Re:Re: The future..

JQP wrote:
Quote
"pNichols" < XXXX@XXXXX.COM >wrote in message
news: XXXX@XXXXX.COM ...
>A system that, BTW, runs the largest computers in the world, reliably.

Now all you've got to do is convince people to put "the largest computers
in
the world" on their desktops. Then we'll all have jobs doing Linux
work<g>.
I do not have to convince anyone of anything. They are discovering this for
thsemlves. Aren't you the same one that stated a few years ago, that Linux
would never replace Windows servers, BTW?
And what of all of the new cellphones and PDAs that are running Linux?
 

Re:Re: The future..

Peter Agricola wrote:
Quote

"pNichols" wrote:
>We primarily program in Java these days, but there are times when I need
>some low level functionality that C or C++ affords us quite well. C++ is
>not nearly as strongly typed however as say Java or Delphi. It can be, if
>one uses Classes/Objects and STL, but how many C+ programs have we seen
>that are not interspersed with structs and the like?

I don't see C++ is less strongly typed than Delphi. I don't see how a
library can make a weak language strong typed. There is no difference
between a struct and a class, except the members of a struct are public by
default where the members of a class are private by default.


Peter
Struts cannot contain methods for one thing, another is the way that they
are instantiated and destroyed. Objects have scope and
constructors/destructors which defines instantiation and the cleanup of
that Object. If you use a struct, then the construction and deallocation is
left up to however it is referenced and in whatever units it is used. It
is very easy to pass a struct around in serveral units and lose track of
where the original pointer was allocated to initialize the structure for
use. Classes or Objects can be abused in like manner (pointers to Pointers
of Objects), but this is a no-no from the world of OOP and Class/Object
structure.
The fact that structs is not self contained and has no constructor nor
deallocation mechanims within the structure, makes it quite different from
a class/object, IMHO.
There is alot of free hand allowed in C++ that would not be allowed in a
more OOP paradigm. The freedom to mix C and C++ is exactly what I have in
mind. I just used a struct as an example, BTW. But it fits as it seems you
too are using it interchangably with a class (NOTE I am not saying this is
a bad thing either, but it is not the purist C++ way).
The library has nothing to do with the language being strongly typed, you
are absolutely correct. It was the Object paradigm that I had in mind here,
which STL follows.
C++ can be as OOP as the developer wants it to be or does not want it to be.
It is the ability to mix C and C++ that gives C++ great flexibility, but it
also this license of mixture that can make it {*word*193}, to say the least. That
is why most clump C and C++ together, since most C++ programs are a mixture
of the two.
 

Re:Re: The future..

"pNichols" < XXXX@XXXXX.COM >wrote in message
Quote
I do not have to convince anyone of anything. They are discovering this
for
thsemlves. Aren't you the same one that stated a few years ago, that Linux
would never replace Windows servers, BTW?
I don't recall making such a statement.
But since you brought it up, do you have any reliable data to show that this
is indeed happening? Everything I've seen shows both Windows and Linux
growing on the server --- at the expensive of proprietary Unix.
Quote
And what of all of the new cellphones and PDAs that are running Linux?
What about them? Are they taking over the market just any decade now?
 

Re:Re: The future..

"pNichols" wrote:
Quote
Struts cannot contain methods for one thing, another is the way that they
are instantiated and destroyed. Objects have scope and
constructors/destructors which defines instantiation and the cleanup of
that Object.
Structs can have methods, constructors and destructors.
Although structs can be initialized like arrays it is better to use a
constructor for that.
StroustrupIII 5.7.
A struct is by definition a class where members are public by default.
struct s {...
is the same as
class s { public: ...
StroustrupIII 10.2.8.
Quote
There is alot of free hand allowed in C++ that would not be allowed in a
more OOP paradigm.
Agreed. In Java and .NET everything is and has to be a class. This doesn't
count for C++ nor for Delphi.
Quote
The freedom to mix C and C++ is exactly what I have in
mind. I just used a struct as an example, BTW. But it fits as it seems you
too are using it interchangably with a class (NOTE I am not saying this is
a bad thing either, but it is not the purist C++ way).
You can not mix C and C++. You can use C syntax and paradigms in C++ but it
still is C++. A struct in C is different from a struct in C++. When
compiling a 'C-program' with a C++ compiler the C++ rules will be applied.
This is confusing and should only be used for legacy code.
Quote
The library has nothing to do with the language being strongly typed, you
are absolutely correct. It was the Object paradigm that I had in mind
here,
which STL follows.

C++ can be as OOP as the developer wants it to be or does not want it to
be.
Agreed. This counts also for Delphi.
Quote
It is the ability to mix C and C++ that gives C++ great flexibility, but
it
also this license of mixture that can make it {*word*193}, to say the least.
That
is why most clump C and C++ together, since most C++ programs are a
mixture
of the two.
See above.
Peter