Beautiful Code

Greg Wilson

University of Toronto

gvwilson@cs.utoronto.ca

http://www.cs.utoronto.ca/~gvwilson

June 19, 2008

Once Upon a Time...


My wife and I decided to start a family	Plus, I had a promise to keep...

Plan A

Write some rilly cool software
But I'm getting old...
...and I've always been better at Version 2

Plan B

I was about to start teaching software architecture
- I wasn't satisfied with any of the books I could find
I'd dabbled enough in open source to know that people will thank you for scratching your itches in public
- They might even help you
Maybe I could just organize something?
- "Just"...

Everyone Else Uses Examples

Ours Are Few and Far Between

May I Have Your Attention Please?

    From: Greg Wilson
    Subject: Beautiful Code
    Date: May 17, 2006

    I hope you don't mind mail out of the blue, but I'm working on a
    new book project with O'Reilly called "Beautiful Code" and would
    like to ask you to contribute an article-length section. Profits
    from the book will be donated to Amnesty International.

    The book will be a collection of master classes in software design.
    In each chapter, a well-known software developer will present one
    of his or her favorite pieces of code, then explain what makes it
    particularly appealing.  The aim is to "think aloud" while walking
    through its design and implementation, so that junior developers
    can learn to see through more experienced developers’ eyes.

    Thanks,
    Greg

Surprises (Many Pleasant)

What I Expected:

    From: [name withheld]
    Subject: re: Beautiful Code

    Sorry, I'm too busy.  Good luck.

What Most People Said:

    From: [name withheld]
    Subject: re: Beautiful Code

    I don't think I've ever written any code I'd call beautiful,
    but I'd really like to read the book when it's done.

But Quite A Few Responded With:

    From: [name withheld]
    Subject: re: Beautiful Code

    Sure, count me in!

Contributors

Tim Bray	Jim Kent	TV Raman
Karl Fogel	Ronald Mak	Arun Mehta
Jon Bentley	Adam Kolawa	Kent Dybvig
Brian Hayes	Lincoln Stein	Andrew Patzer
Alberto Savoia	Ashish Gulhati	Bryan Cantrill
Andreas Zeller	Brian Kernighan	Charles Petzold
Henry S. Warren	Andrew Kuchling	Michael Feathers
Douglas Crockford	Greg Kroah-Hartman	Diomidis Spinellis
Travis E. Oliphant	Simon Peyton Jones	Yukihiro Matsumoto
Elliotte Rusty Harold	Jeff Dean & Sanjay Ghemawat	Jack Dongarra & Piotr Luszczek
William Otte & Douglas C. Schmidt	Laura Wingerd & Christopher Seiwald	Rogerio de Carvalho & Rafael Monnerat

Beauty Takes Many Forms

This is not one of them

    int j=0;
    for (int i=0; i < a.length; ++i) {
        j++;
    }        
    System.out.println("array size=" + j);

Other than that, little agreement about where to look
Or rather, so many possibilities that everyone found one of their own

Beauty in the Small (Hayes)

Beauty in the Medium (Fogel)

...Beauty in the Medium (Spinellis)

Beauty In The Large (Dean & Ghemawat)

Beauty In The Even Larger (Mak)


What do you do when a day is 24 hr 39 min long?

The Beauty of the Classics (Kernighan)

    int match(char *regexp, char *text)
    {
        if (regexp[0] == '^')
            return matchhere(regexp+1, text);
        do {
            if (matchhere(regexp, text))
                return 1;
        } while (*text++ != '\0');
        return 0;
    }

    int matchstar(int c, char *regexp, char *text)
    {
        do {
            if (matchhere(regexp, text))
                return 1;
        } while (*text != '\0' && (*text++ == c || c == '.'));
        return 0;
    }

c	any literal character
.	any character
^	matches the beginning of the input string
$	matches the end of the input string
ccc	any sequence of characters or .'s
*	closure: zero or more of the previous character

    int matchhere(char *regexp, char *text)
    {
        if (regexp[0] == '\0')
            return 1;
        if (regexp[1] == '*')
            return matchstar(regexp[0], regexp+2, text);
        if (regexp[0] == '$' && regexp[1] == '\0')
            return *text == '\0';
        if (*text!='\0' && (regexp[0]=='.' || regexp[0]==*text))
            return matchhere(regexp+1, text+1);
        return 0;
    }

The Beauty of Utility (Kolawa)

      SUBROUTINE SGBSV( N, KL, KU, NRHS, AB, LDAB, IPIV, B, LDB, INFO )
*
*  -- LAPACK driver routine (version 2.0) --
*     ..
*     .. Scalar Arguments ..
      INTEGER            INFO, KL, KU, LDAB, LDB, N, NRHS
*     ..
*     .. Array Arguments ..
      INTEGER            IPIV( * )
      REAL               AB( LDAB, * ), B( LDB, * )

Easy to mock, but how much code has been in daily use for 35 years?
On dozens of novel architectures?

Utility Takes Many Forms (Mehta)

The Beauty of Compromise (Crockford)

     The turnstile hash table is partitioned into two halves: the lower half
     is used for upimutextab[] locks, the upper half for everything else.
     The reason for the distinction is that SOBJ_USER_PI locks present a
     unique problem: the upimutextab[] lock passed to turnstile_block()
     cannot be dropped until the calling thread has blocked on its
     SOBJ_USER_PI lock and willed its priority down the blocking chain.
     At that point, the caller's t_lockp will be one of the turnstile locks.
     If mutex_exit() discovers that the upimutextab[] lock has waiters, it
     must wake them, which forces a lock ordering on us: the turnstile lock
     for the upimutextab[] lock will be acquired in mutex_vector_exit(),
     which will eventually call into turnstile_pi_waive(), which will then
     acquire the caller's thread lock, which in this case is the turnstile
     lock for the SOBJ_USER_PI lock.  In general, when two turnstile locks
     must be held at the same time, the lock order must be the address order.
     Therefore, to prevent deadlock in turnstile_pi_waive(), we must ensure
     that upimutextab[] locks *always* hash to lower addresses than any
     other locks.  You think this is cheesy?  Let's see you do better.

Theory Made Practice

Kent Dybvig on Scheme's syntax-case expander
Simon Peyton Jones on software transactional memory
Jon Bentley on Quicksort
Yukihiro Matsumoto on language design


    Simplicity is one of the most misunderstood
    concepts in programming.  People who design
    languages frequently want to keep those
    languages simple and clean.  While the
    sentiment is noble, doing this can make
    programs written in that language more complex.

Beauty As A Process

Alberto Savoia on beautiful testing
Andreas Zeller on beautiful debugging
Laura Wingerd & Christopher Seiwald on beautiful version control
- How do you design code so that it will survive decades of branching and merging?

Reactions

"'Beauty' is the wrong way to think about software."
- St. Paul's is beautiful
- So are the blueprints for it
"Everyone has their own standards for beauty."
- Existence of recording encouraged musicians to strive for something other than perfection
- We can define any esthetic for software we want

2008-06-19

Beautiful Code

Beautiful Code

Greg Wilson

University of Toronto

gvwilson@cs.utoronto.ca

http://www.cs.utoronto.ca/~gvwilson

June 19, 2008

Once Upon a Time...

Plan A

Plan B

Everyone Else Uses Examples

Ours Are Few and Far Between

May I Have Your Attention Please?

Surprises (Many Pleasant)

What I Expected:

What Most People Said:

But Quite A Few Responded With:

Contributors

Beauty Takes Many Forms

Beauty in the Small (Hayes)

Beauty in the Medium (Fogel)

...Beauty in the Medium (Spinellis)

Beauty In The Large (Dean & Ghemawat)

Beauty In The Even Larger (Mak)

The Beauty of the Classics (Kernighan)

The Beauty of Utility (Kolawa)

Utility Takes Many Forms (Mehta)

The Beauty of Compromise (Crockford)

Theory Made Practice

Beauty As A Process

Reactions

Why Should We Care?

Where Next?

Thank You


"Beauty is truth, truth beauty."	Pride