Hints for Computer Systems Design
----------------------------------
This paper offers a number of helpful hints towards the design of computer
systems, focusing on three major areas: functionality, speed, and
fault-tolerance.  The amount of information presented in this paper is
substantial, and I don^t doubt that this will be a handy reference in the
future, but a number of ideas firmly lodged into my head while reading.

1) Designing a computer system is very different from designing an
algorithm. Although this may be obvious, it is useful to realize the way
they differ and what this means for computer systems designs.  For
example, it may pay off to be extremely clever when designing an
algorithm, but being clever can get you into trouble when designing a
computer system.

2) Keep it simple! This idea is one of the main tenets of this paper.  
Often, we bury ourselves in complexity in an attempt to be clever and
general when the simplest approach - while perhaps not being ^optimal^ -
can be more than sufficient.

3) Learn from the past. This paper (and the ideas contained within) exist
because Lampson decided to organize his years worth of systems experience
into a paper filled with useful hints.  It is important to look back on
our experience and see what went well/wrong and why.

Although this paper contains a lot of useful information, some of the
ideas seem contradictory.  Yes, I read the disclaimer stating that these
hints were not always consistent, but it leaves me feeling as if I need
more information.  For example, the suggestion to ^Keep secrets of the
implementation^ left me quite confused.  On one hand, it^s stated that
assumptions between components of a system are bad since changes to one
may require changes to the other, but assumptions can also improve
performance. Improving performance supports the idea of ^Make it fast.^

I see this paper as being a very useful reference for those involved in
systems design and implementation.  The majority of ideas make sense, with
examples to support them.  I couldn^t help but relate many of these ideas
to projects I have worked on in the past, and have no doubt that I will
relate these ideas to projects I will work on in the future.