dotgnu-general
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [DotGNU]Dotgnu written in C#?


From: Fergus Henderson
Subject: Re: [DotGNU]Dotgnu written in C#?
Date: Tue, 12 Aug 2003 12:34:59 +1000
User-agent: Mutt/1.3.28i

On 10-Aug-2003, DrDiettrich <address@hidden> wrote:
> Fergus Henderson wrote:
> 
> > What happens if the software needs to do something which neither the IDE
> > nor the user know how to do?
> 
> Can you please give me any single concrete example of such a situation?

Sure, I could give you lots.  For example, getting pointers to the start
and end of the writable data segment (conservative garbage collectors
need this information, but current environments don't provide any
standard interface to it).

> A program, as source code, can require certain information about it's
> runtime environment, and it can request certain services from that
> environment. The only requirement is the availability of the services,
> which provide the information or perform specific tasks. All that is a
> matter of libraries, which may need a platform specific implementation.

Pushing the problem into "libraries" doesn't make it go away.
Those libraries still have to be implemented.
Each program may need to have its own corresponding library
for all the non-portable bits.

Of course it is good to structure programs so that the non-portable
parts are isolated.  But that doesn't necessarily mean putting all
those parts in separate packages.

> a program should never ever care about compiler bugs.

Programs which never care about compiler bugs will be out-competed by
programs which do care, because programs which do care make it easier
for users.

> Likewise a program should not rely on special
> non-standard compiler features. Such procedures may be appreciated by
> hackers, but not by software engineers.

By that definition, programs developed by software engineers will be
out-competed by programs developed by hackers, because the programs
developed by hackers will offer more features.

But I don't agree with that definition.  IMHO software engineers know
how to make appropriate trade-offs between portability, functionality,
and simplicity.  Choosing portability and simplicity at the expense of
functionality is not always the best trade-off.

-- 
Fergus Henderson <address@hidden>  |  "I have always known that the pursuit
The University of Melbourne         |  of excellence is a lethal habit"
WWW: <http://www.cs.mu.oz.au/~fjh>  |     -- the last words of T. S. Garp.


reply via email to

[Prev in Thread] Current Thread [Next in Thread]