Board index » delphi » Re: Delphi - does catastrophe loom ? (long)

Re: Delphi - does catastrophe loom ? (long)


2005-04-12 06:59:03 PM
delphi278
"Dave Jewell" <XXXX@XXXXX.COM>writes
Quote
"Tim Anderson" <XXXX@XXXXX.COM>writes
news:XXXX@XXXXX.COM...

>Your statement about unsuitability for desktop apps also seems to be
>undermined by the popularity of a number of apps written with the .NET
>Framework. For example RSS clients - RSS Bandit, SharpReader, Omea
>Reader, others too.

... and Paint.NET which you know I have mentioned in "Another Place". :-)
Paint.NET is different though. it is a nice example, but I don't know that it
is popular as a real-world paint application. Whereas the RSS readers are in
constant use.
Quote
However, I'd still like to see evidence that Microsoft are themselves
committed to serious desktop app development with .NET.
Understood, but there's nevertheless substantial use of .NET at Microsoft.
But not in Office - apart from VSTO of course.
Quote
I'm sure others saw no evidence back in the days of OS/2. :-)
Personally, I wouldn't be dogmatic either way, but I wouldn't exclude the
possibility either. What *is* certain is that Microsoft definitely didn't
"bet the farm" on .NET, regardless of what they might have said to the
contrary.
Hmm, so what did Microsoft say? Would be interesting to have some quotes.
I do remember it saying that Windows was a transitional thing on the way to
OS/2 :-)
Tim
 
 

Re: Delphi - does catastrophe loom ? (long)

Dave Jewell writes:
Quote
As with Kylix,
I've never been convinced that the Kylix idea was such a turkey.
For whatever reason it never quite worked though. For me the biggest
problem is that Kylix applications always looked 'slightly strange' -
at least on KDE. Non antialiased fonts, not quite 'standard' behaviour
of common controls, etc.
If it had been 'perfect' it could have romped away with the client, gui
side of things (there's still no other proper RAD language for Linux,
and doing non-rad UI development in Linux is even worse than on win32!).
And once it had got established on the client/gui side it would have
osmosed onto the server side too.
--
Colin
 

Re: Delphi - does catastrophe loom ? (long)

Mat Ballard writes:
Quote
I believe that .NET has been grossly over-hyped, in typical Microsoft
fashion. While the original intent seemed for it to be a Java
replacement, it is now pushed as the general programming language for
future versions of Windows.
.NET is at most a platform, not a language. Like everything that is
rather new, there are issues, like the speed of graphics, etc. I'm
surethat eventually such things can and will be solved.
Java was also grossly overhyped, a few years ago.
In the meantime, I think it would be foolish to ignore .NET completely.
--
Rudy Velthuis [TeamB] rvelthuis.bei.t-online.de
"I'm not a member of any organized political party, I am a Democrat!"
-- Will Rogers (1879-1935)
 

Re: Delphi - does catastrophe loom ? (long)

Chris Burrows writes:
Quote
.NET desktop apps that I have written compare
well with their Win32 equivalents. However, I have taken care to
rewrite them specifically for .NET rather than just do a blind port
from Win32.
Yes, indeed. I noticed the same. But also non-desktop apps compare
favourably, AFAICS.
--
Rudy Velthuis [TeamB] rvelthuis.bei.t-online.de
"Physics is not a religion. If it were, we'd have a much easier time
raising money." -- Leon Lenderman
 

Re: Delphi - does catastrophe loom ? (long)

Jim Cooper wrote in <XXXX@XXXXX.COM>:
Quote
You fail to acknowledge the business side of
things for some reason, though.
That's not a failure on my part-- I would accept the "business side"
argument if it made the least bit of sense.
Will
--
Want native support in Delphi for AMD64/EM64T? Vote here--
qc.borland.com/wc/qcmain.aspx
 

Re: Delphi - does catastrophe loom ? (long)

Quote
I've never been convinced that the Kylix idea was such a turkey.
Indeed. Kylix wasn't a bad idea, it was just very badly executed.
 

Re: Delphi - does catastrophe loom ? (long)

"marc hoffman" <XXXX@XXXXX.COM>writes
Quote
>I've never been convinced that the Kylix idea was such a turkey.

...wasn't a bad idea, it was just very badly executed.
Hmm, where have I heard that term before - oh yes, it was during last
Thursday's conference call that led to Friday's tumultuous selloff.
 

Re: Delphi - does catastrophe loom ? (long)

"Atle Smelvær" <XXXX@XXXXX.COM>writes
Quote
Maybe the reason they have .NET is for them to rule the software market.
Because Microsoft will not use .NET, they will have faster and better
applications. And it will also be very easy for them to dissassemble
competitors software, because basic .NET software is so wide open.

.NET to rule them all...
..and in the IL bind them? ;-)
Best,
Dave
 

Re: Delphi - does catastrophe loom ? (long)

Quote
Eric Grange in <425b7bde$XXXX@XXXXX.COM>writes:
>FB-111A is the fastest bomber.

The "fastest bomber" title might rather go to the Tu-22M3 "Backfire",
which has a similar top speed, but maintains that speed over
longer distances and delivers heavier payloads than the F-111.
IIRC the fastest "bomber" would have to be the B-58.
www.aviation-history.com/convair/b58.html
 

Re: Delphi - does catastrophe loom ? (long)

"Mat Ballard" <XXXX@XXXXX.COM>writes
Quote
g'day Chris,

however, the main project i am working on at the moment involves complex
manipulations on images that have now reached>100M in size, and require
minutes per algorithm. no amount of optimization or re-writing will make
.NET viable. but a 64-bit Delphi will.
Are you using fully managed code or have you have you tried using unsafe
code in performance-critical sections? Most image processing type code
that I have seen in .net is done with unsafe code in critical sections, and
then you can get the benefit of using pointers.
Mike
 

Re: Delphi - does catastrophe loom ? (long)

"Mat Ballard" <XXXX@XXXXX.COM>writes
Quote
I believe that .NET has been grossly over-hyped, in typical Microsoft
fashion. While the original intent seemed for it to be a Java replacement,
it is now pushed as the general programming language for future versions
of Windows. However, recent posts on its performance have confirmed its
Java-like performance (memory x 5, speed x 0.02-0.2), which makes it
unsuitable for general desktop applications.
If it is so unsuitable for general desktop applications, then why is MS
developing their next Office App in .NET (Small Business Accounting). While
I don't expect Word or Excel to be ported to .net any time soon, if ever, MS
is at least eating their own dog food.
Mike
 

Re: Delphi - does catastrophe loom ? (long)

Quote

I've never been convinced that the Kylix idea was such a turkey.

For whatever reason it never quite worked though. For me the biggest
problem is that Kylix applications always looked 'slightly strange' -
at least on KDE. Non antialiased fonts, not quite 'standard' behaviour
of common controls, etc.

If it had been 'perfect' it could have romped away with the client, gui
side of things (there's still no other proper RAD language for Linux,
and doing non-rad UI development in Linux is even worse than on win32!).

And once it had got established on the client/gui side it would have
osmosed onto the server side too.

+1
I still think Kylix is a great idea.
But unfortunately it isn't (wasn't?) only the quality (Wine-dependent
IDE, CLX Bugs) which is responsible for its non-success (they say).
Things aren't that easy.
In other words: If you were Borland an you could start again with a
project like this, how would do it?
- How can a closed-source solution be attractive enough in a open-source
world?
- Which Widgetset to choose? GTK is ugly and for QT you need a license.
Andreas Hausladen is doing a Qt3-port
unvclx.sourceforge.net/other/ClxQt3.png
But you will need a license to use it for commercial Apps.
- Would you introduce a CLX or just try to mimic the VCL on Linux
without the need for a CLX for Delphi?
- If you kept it closed-source, would you offer a service-pack every 6
months to make it work with the newest Distros?
If I was Borland, I would keep a foot in the Linux-market (?) by selling a
compiler which is in sync with the Win32 (64?) Compiler and a RTL.
I'd leave the rest to the opensource-community (IDE, VCL..). Too much work.
I (as a consumer) would be happy with this and would pay for it.
I know about FPC. But the Borland compiler is better.
 

Re: Delphi - does catastrophe loom ? (long)

Quote
That's not a failure on my part-- I would accept the "business side"
argument if it made the least bit of sense.
You mean you don't see the business reasons why MS have not rewritten
SQL Server?
Cheers,
Jim Cooper
__________________________________________
Jim Cooper XXXX@XXXXX.COM
Tabdee Ltd www.tabdee.ltd.uk
TurboSync - Connecting Delphi to your Palm
__________________________________________
 

Re: Delphi - does catastrophe loom ? (long)

Dave Jewell <XXXX@XXXXX.COM>writes
<425ba05c$XXXX@XXXXX.COM>
Quote
That's an interesting observation. I would like to know how that is achieved,
bearing in mind that the .NET frameworks are built on top of the Win32 API.
The same way the VCL has better support for GUI apps than the Win32 API
alone. You see, there are these things called frameworks that add
functionality and power to things.
--
***Free Your Mind***
Posted with JSNewsreader-BETA 0.9.4.516
 

Re: Delphi - does catastrophe loom ? (long)

Quote
That's *your* opinion, which hardly makes it a fact or the truth.
How do you know it is just my opinion? ;-)
Quote
Who's to say that the only real apps are those you write?
I've never maintained that. You're the one who's complained before that
any apps that don't have a performance issue on .NET aren't "real" apps
Quote
To encounter that, you only have to allocate medium-sized
objects, or get close to the swapping hell, which with the GC's
extra memory consumption isn't very hard.
Why have I never encountered it then?
Quote
For *some* of you, that may be true. :)
Yes indeed, for many of us it is.
Quote
Yet you keep thinking that the kind of application
*you* are doing with Delphi is representative of the majority
of Delphi applications out there
And as it happens...
Quote
and from that point on,you tell us that thus Delphi doesn't need to
be fast.
No I don't. I point out that your performance issues don't necessarily
affect many others.
Quote
Maybe pointing us to well known, widespread software made
by you would give more weight to that opinion...
Why would the software have to be well-known and widespread? By
definition most people don't work on that sort of software.
Quote
but 'till then, it is just your opinion and your perception
Take note of the ;-), Eric
Quote
as there are many Delphi applications out there that are recognized
for their
Quote
performance, efficiency and richness-without-needing-a-supercomputer ;)
I'm not arguing with that. Just pointing out .NET performance is
perfectly adequate for many applications.
Quote
Yes, but we aren't being told the VCL/RTL is the epitome
of good design
By whom? I don't think even the people who wrote it would claim that :-)
Quote
How can you obsolete MS's code?
Why would you mark other people's code as obsolete? *MS* can mark things
as obsolete.
Quote
You have to make do with whatever MS delivered
No you don't, any more than you have to with Delphi. Use something else.
Quote
Strings have issues with concatenation
Yes, strings being immutable means you should do things differently
Quote
and all operations as soon as their length goes a bit over 64-90 kB
Do you mean speed issues?
Cheers,
Jim Cooper
__________________________________________
Jim Cooper XXXX@XXXXX.COM
Tabdee Ltd www.tabdee.ltd.uk
TurboSync - Connecting Delphi to your Palm
__________________________________________