In Search of Elegance
04-17-2000, 01:43 AM
I agree that s/w complexity grows with size; I think this is self evident
to anyone who has written a non-trivial s/w application. The real problem
is not this however. The real problem is that WinNT (written by some very
accomplished OS designers) had to be reasonably backward compatible with
other MS OSes.
OS technology is quite old now and many of the core requirements of a solid
OS are well known. The art of s/w design has also evolved quite alot, allowing
skilled individuals to develop s/w which scales better. So, don't be arrogant
to say that an accomplished team of designers could not make a cleaner operating
system. But I bet you they could not produce one that was better and backward
compatible with the other Microsoft stuff. Further more they could not sell
it because of the backward compatibility issues. This is not to say that
the MS team is not top notch, just their stuff is often crap because of constraints
other than those you mentioned.
Linux? I wonder why you claim it is not even in the same league as Windows
NT? I gather that security is a big strength of NT. I've heard that LINUX
crashes much less by people who have developed servers using both technologies.
I have seen first hand how easy it is to crash NT and I would not trust my
life to this system. I honestly don't know whether or not LINUX is better.
I do find that the Microsoft standards I've seen so far are often quite ugly
but well documented. Sometimes though I have seen some elegant Microsoft
stuff but ... I can't think of any examples at the moment though. I think
I may be getting used to MS ... (Eeeek. This could be scary. I do love MSDN
though; except I hate that I cannot view it on my GFs SUN workstation!!!
I don't think that anything really complicated is involved in this WWW page
that could not be handled by a simple standard ...? Is it such a big deal
to expect industry standards? The other problem is that most of the desktop
market is not made of people who realize that it is technically feasible
and good to have standards. They just want something that works that they
don't need a Ph.d to use.)
"Dave" <hardtosay@usa.net> wrote:
>Oh great, another so called great programmer who thinks they could do a
better
>job of writing a computer OS than MS - if they only had a little more time.
> If you've ever developed for a system with more than a hundred thousand
>lines of code (and Windows 2000 has about 2 million I think) you would know
>that as the system grows the complexity increases exponentially, not linearly.
> Just because you can develop a nice little 10,000 line
Mabye this is a reasonable assumption but for all you know the
author may indeed have the same or more experience than your
self! Mabye the person has a Ph.d is OS design? (I doubt this
but you never know.)
>application doesn't
>mean you can develop one even 10 times that size. Whenever >someone says
>how they could write a better OS than Microsoft I just laugh to myself at
>their ignorance. And please, don't start comparing Linux to
>Windows 2000
>as a better OS. Compare Linux to DOS 6.x and I will listen.
>
>Like many other developers I develop Intranet applications. All my users
>(2500) have the same OS, browser (IE 4.1) etc.. The problem is that my
users
>are used to feature rich user interfaces that HTML 3.2 can't deliver. By
>taking advantage of IE advanced features (i.e. DHTML, ActiveX) I am able
>to deliver to my users what the want and need. I also have to write Internet
>applications but the user's expectations are much lower so coding to HTML
>3.2 works fine without having to worry about browser compatibility issues.
>
>Standards are great and should be followed when they make sense. When they
>don't or are lacking then the people that actually make the software need
>to decide what goes and what stays. In the end it is their jobs on the
line,
>not the standard committees.
>
>
to anyone who has written a non-trivial s/w application. The real problem
is not this however. The real problem is that WinNT (written by some very
accomplished OS designers) had to be reasonably backward compatible with
other MS OSes.
OS technology is quite old now and many of the core requirements of a solid
OS are well known. The art of s/w design has also evolved quite alot, allowing
skilled individuals to develop s/w which scales better. So, don't be arrogant
to say that an accomplished team of designers could not make a cleaner operating
system. But I bet you they could not produce one that was better and backward
compatible with the other Microsoft stuff. Further more they could not sell
it because of the backward compatibility issues. This is not to say that
the MS team is not top notch, just their stuff is often crap because of constraints
other than those you mentioned.
Linux? I wonder why you claim it is not even in the same league as Windows
NT? I gather that security is a big strength of NT. I've heard that LINUX
crashes much less by people who have developed servers using both technologies.
I have seen first hand how easy it is to crash NT and I would not trust my
life to this system. I honestly don't know whether or not LINUX is better.
I do find that the Microsoft standards I've seen so far are often quite ugly
but well documented. Sometimes though I have seen some elegant Microsoft
stuff but ... I can't think of any examples at the moment though. I think
I may be getting used to MS ... (Eeeek. This could be scary. I do love MSDN
though; except I hate that I cannot view it on my GFs SUN workstation!!!
I don't think that anything really complicated is involved in this WWW page
that could not be handled by a simple standard ...? Is it such a big deal
to expect industry standards? The other problem is that most of the desktop
market is not made of people who realize that it is technically feasible
and good to have standards. They just want something that works that they
don't need a Ph.d to use.)
"Dave" <hardtosay@usa.net> wrote:
>Oh great, another so called great programmer who thinks they could do a
better
>job of writing a computer OS than MS - if they only had a little more time.
> If you've ever developed for a system with more than a hundred thousand
>lines of code (and Windows 2000 has about 2 million I think) you would know
>that as the system grows the complexity increases exponentially, not linearly.
> Just because you can develop a nice little 10,000 line
Mabye this is a reasonable assumption but for all you know the
author may indeed have the same or more experience than your
self! Mabye the person has a Ph.d is OS design? (I doubt this
but you never know.)
>application doesn't
>mean you can develop one even 10 times that size. Whenever >someone says
>how they could write a better OS than Microsoft I just laugh to myself at
>their ignorance. And please, don't start comparing Linux to
>Windows 2000
>as a better OS. Compare Linux to DOS 6.x and I will listen.
>
>Like many other developers I develop Intranet applications. All my users
>(2500) have the same OS, browser (IE 4.1) etc.. The problem is that my
users
>are used to feature rich user interfaces that HTML 3.2 can't deliver. By
>taking advantage of IE advanced features (i.e. DHTML, ActiveX) I am able
>to deliver to my users what the want and need. I also have to write Internet
>applications but the user's expectations are much lower so coding to HTML
>3.2 works fine without having to worry about browser compatibility issues.
>
>Standards are great and should be followed when they make sense. When they
>don't or are lacking then the people that actually make the software need
>to decide what goes and what stays. In the end it is their jobs on the
line,
>not the standard committees.
>
>