Re: The Great Debate V. What have changed ?

From: Kain (kahlinor_at_yahoo.com)
Date: 03/04/04 

Date: 4 Mar 2004 11:51:25 -0800

"The Half A Wannabee" <ShakainZulu_AT(Spetakkel - Innsikt over makt og Spetakkel Rocker).com> wrote in message news:<4046f745$1@news.broadpark.no>...
> "Kain" <kahlinor@yahoo.com> wrote in message
> news:cd77bdb5.0403032009.3c671661@posting.google.com...
> > "Beth" <BethStone21@hotmail.NOSPICEDHAM.com> wrote in message
> news:<vAk1c.64$2F6.63@newsfep3-gui.server.ntli.net>...
>
> (*
> The problem with Wannabee is that he has been conditioned to be
> anti-Randy.
> *)
>
> No. This has been a slow progress, evolution if you will. I used to love
> Randalls AoA, until I discovered RosAsm. I started with Delphi3.0 in 1996. I
> had since long bought into the flawed idea of OOP, polymorphism, I fell prey
> to it. I programmed using the Delphi VCL for years. Slowly I discovered
> flaws in some of the objects, that carried subtle bugs. After numerous
> problems using the Objects library, I came to the concusion, that if I would
> ever be able to write a bugfree program (while allways pushing/expanding
> functionality), I at least had to make sure that the codebase I was blindly
> using would be errorfree. So I through away my VCL code, with some tears,
> and started developing my own equivalent. Duing development, I found I
> wanted more memory statistics than Delphi was giving me. They have an
> AllocMemCount and a variable showing how much memory you use, and how much
> is free, thats about it. So I rewrote the classes.pas file, and intercepted
> the objects tree, to insert a call to my own written memorymanager. During
> this fase I found a bug in the allocation routines that Delphi uses, that
> will make inserting an override call in the top objects to get memory from
> somewhere else buggy. I fixed the bugs. After a while I found that my memory
> manager, while written in Delphi only performed half as good as the default
> memorymanager. I had to perform a major operation of testing it, and
> evaluating it, and rewrite. During this work, which was an immensly idiotic
> operation, compared to writing my own memorymanager in RosAsm a few weeks
> ago, I began to see, that allthough Delphi allow me to replace the
> memorymanager, it does not "allow" me to replace HOW, the overall HOW of how
> memory is allocated. The sequence if you will. The objects would still call
> into the memorymanager in the same order, and for just as many iterations.
> This was irritating. Allthough my memorymanager was slower than delphis, on
> deallocations it was doubled the speed. Now my point. All of this was a lot
> of work, and so involving ( I had to first understand Delphis approach -
> from debugging through it ) that deep down I was allready on the lookout for
> a whole diffrent answer. More control. If my programs shall be crashing on
> me, I want to know that I was the once creating those bugs, as then I can
> remove them, without having to study the Delphi implementation better then
> Borland knows it themselfes. Then One day I found a link to Spasm. I loaded
> one of the Izcellion tutorials, infact, in the same day, I loaded all of
> them, and for each and everyone, my month just opened wider and wider. You
> know, if you have been looking for something for a long time, unknowingly,
> and when you finally find it to allready have implemented 99% of all you
> have ever envisioned, then Love at first site.
>
> Not taking this seriously, as it ment all of my newly written code would
> have to be deprecated, to program in VCL, I, lazily, kept away for some
> time. Some years. I followed the Spasm project from time to time, but was
> unwilling to give up on my now 200 000 delphi codelines. But to do the
> things I really want(tobee), through and though, about a month ago, late
> january, I just pulled the plug on delphi. And allready RosAsm has proved
> itself to be every bit as wonderful as I though back then and more. I have
> learn a lot of stuff, I even could not conceive of beeing so wonderfully
> easy to do in an assembler. I am still very fresh to asm, but this will not
> prevail.
>

That's all well and good, but it doesn't explain why your anti-Randy.
It seems to actually prove my point. You fell in with Rene, liked his
assembler and baught part-and-parcel the whole Rene idiology.

>
> So. This is not "conditioned" at all, if it is, then its conditioned by
> experience. Randall should maybe program MORE in HLL, so that he will learn
> the same. You know, when you find you have been lied to your whole life,
> even by you parents, even by your best friends, even by yourself, then when
> you find out, you're likly to not accept any bull*** from anyone, if it
> even be God himself. RosAsm is very advanced. Randy, while claiming 35 years
> in the programming area, cant see it. So then he makes himself the laughing
> stock to the hardcore programmers.

Actually, my observation shows the opposite of what you claim here.

> Cause they, maybe not beeing any smarter,
> just more focused than he is, does the same things that reallife programmers
> have allways done. Feel sorry for the old professor, that even though his
> heart is in the right place perhaps (dont know) - doesnt know as much as he
> thinks he does. That even a newbie, to asm like me, can easily cause his
> selfimage to shake. But the people who give testimony that Betov is fake,
> like Randy, Hutch, Donkey and now, lately Beth, are fully makeing an ass out
> of themself completly. To thoose with more than 5 years programming
> experience with solving real world problems, theese people come out looking
> like a bunch of rather pityfull retards. I am frankly sometimes, afraid,
> that when I pick on them, I pick on some gang of people, I should rather
> help, then pick on. They make me feel like a bulley, just for answering to
> the fullest of my simple skills. The four, Beth, Randy, Hutch and Donkey,
> even though diffrent, when it comes to what they say about programming, I
> get the feeling that to them its not about anything but theire higly
> overestimated egos.

Listen to yourself here. "overestimated egos"? Hello? Have you
listened to Rene's holy-war-assembly-rebirth-microsoft/america evil
speils? I know an inflated ego when I see one. It's usually the one
who can't take any criticism while dishing out criticism left and
right.

> IS there really anyone, still thinking LuxAsm or ConvInc
> will have a chance ? Last time I looked the last message there was
> http://sourceforge.net/forum/forum.php?thread_id=1016998&forum_id=346957
> posted 7. feb 2004, allmost a month ago.
>

I think "LuxAsm" is just a mean joke on Rene. They have found that
Rene's ego is his weakness and that's what the "LuxAsm" project is
designed to attack. I doubt we'll every see any LuxAsm code.

> To those of us with a heart for programming, we know that you dont get it
> right the first time, nor the second time, nor the third. Maybe you can get
> it right the 20th time around. THIS is what it takes to make world class
> programs theese days, because you fight agains big dollars, and they can
> easily crush you. The only logical consequence for those who loves
> programming is to do fully asm, nothing left to chance. Beth word put into
> action. Words are so cheap, that the more of them the less worth they have.
> This of course apply to me as well;-) Thats why I end this now.I feel I have
> said what I came here to say, a long time ago. I may not have the weigth of
> 68K to back me, but I _do_ really feel that what Betov says should not be
> taken lightly, but should be considered very carefully. From now on I will
> sidestep this debate. It wear me down, and I have work to do. If you keep at
> programming, and are enough interessted, you will most likely arrive at
> something similar to theese views, when you've gone the mile. The main
> reason why Spesific is better way, is that HLL can allways answer just a few
> of the problems we will face as programmers. asm can answer allways to your
> current needs, and as I have shown with my example code, with an assembler
> like RosAsm there is no need to write unreadable code. Thoose who do, also
> have no clue what it takes to write a full part project. When they emit
> unreadable code, that you have to spend minutes to get the picture, then
> this should be proof enough that they will head nowhere by that approach.
> Bye.

It's commendable that you want out of the debate, but I have a feeling
it will suck you back in, even though it is going nowhere fast. I bet
when Rene stops his baseless accusations, the debate will end as well.
 But it's kind of hard for a hard-head like him to stop once a
holy-war aginst reason has been declared.