help-hurd
[Top][All Lists]
Advanced

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

Re: Design Decisions and Hurd/L4 work (was: Re: Improving Hurd)


From: Thomas Bushnell, BSG
Subject: Re: Design Decisions and Hurd/L4 work (was: Re: Improving Hurd)
Date: 21 Apr 2002 15:05:19 -0700
User-agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2

Jan Atle Ramsli <trollet@skynet.be> writes:

> I asked you if you thought it was possible that any Hurd server could be
> implemented with only 5 functions.

Whe cares how many functions it takes?

> (And yes, I said ADT, and NO! ADT is not a buzzword, and if my old
> software enineering tutor got hold of the the one who said that, he
> would pull him by one ear, forcing him to repeat "ADT is not a buzzword!
> ADT is a fundamental concept! ADT is not ..." to the end of the street.
> He would grab him by the other ear and drag him back and force him to
> repeat "ADT is the way in which programs can be mathematically
> justified. ADT is ..."

The fixation on abstract data types was a fad in computer science of
about ten or fifteen years ago.  We now see that abstract data types
are a useful concept, but not something to organize all your thinking
around.

Certainly if you poke around the Hurd source code, you will see some
things that look mighty similar to abstract data type programming--for
it is indeed a useful tool to have in one's toolbox.  But it is
pointless to throw out many useful tools in some zealous desire to use
only one.

There are--horror of horrors--data type problems that abstract data
types are unable to adequately describe, and this is precisely what
led to their demise as a fundamental organizing principle about ten
years ago.




reply via email to

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