97 Things Every Programmer Should Know
Create an account by clicking here. You don't need an email invitation to create an account.
Welcome to the home page for developing the initial phase of 97 Things Every Programmer Should Know.
97 Things Every Programmer Should Know is intended to be a book of wisdom for programmers collected from leading practitioners. There is no overarching narrative: it is intended simply to contain multiple and varied perspectives on what it is that you feel programmers should know. This can be anything from code-focused advice to culture, from algorithm usage to agile thinking, from implementation know-how to professionalism, from style to substance, etc.
Now that we have a suitable number of complete and high-quality contributions, we're looking to move to the next phase of the project, where we open up the existing items to the public for comment and further contributions. This looks like it will be July 2009. Following that we will move to the final phase of the project, publishing a book with a selection of the 97 contributions that work best together.
Thank you for taking the time and making the effort to contribute.
What Will Come of All This?
This current site is for the initial phase of the project. In the next phase of the project, starting in July, O'Reilly will publish the contents of this wiki in a public and free web site off the O'Reilly properties. It will be free for anyone to access but you'll have to register to contribute or comment. Users (that's everyone who is registered) will be able to comment on other peoples contributions and create, edit, and improve their own contributions. Anyone and everyone be able to view the material without requiring registration. The web site will be strongly promoted by O'Reilly and all contributors will get full attribution.
After this, O'Reilly will consider taking the next step, which is to publish a 97 Things Every Programmer Should Know book. If a book is to be published, the best contributions and the contributions that fit best together will be selected and edited by me and Michael Loukides, an editor from O'Reilly. The book will sell in book stores and on-line. It will be listed as edited by Kevlin Henney. If your contribution is chosen, any recommended edits recommended will be contributed back to the 97 Things project for everyone to enjoy. Every contributor whose contribution goes into the book will be fully acknowledged in the book and will get a complementary copy of the book when it is published.
To get a feel for this type of project take a look at another book in the series called 97 Things Every Software Architect Should Know. It's the flagship project for the 97 Things series and is the first book in the series.
How to Make a Contribution
- Each contributor is asked to provide one or more items (tips or bits of wisdom) that each have a title and associated discussion. The title should only be a 2 to 10 words long if possible and should summarize or capture the essence of the advice. In print, we want each contribution to fit on a two-page spread. Keep your discussion between 400 and 500 words. Any contribution under 400 words is unlikely to make it to the next phase of the project. And much more than 500 words will need to be edited down.
- Create an account and author page. To create and edit a page you will need to create an account. You can then contribute your items and provide an author page. For more details on this see How to Get Started.
- Please read the first contribution. I've added an example contribution in The Contribution section below to provide further guidelines on content style and to show you an example of what you will see when your are ready to add your own tip/axiom/pearl/guideline/contribution. Reading the initial contribution won't take long — it's not much more than 500 words!
Rules of Engagement
- Contributors need to have an account and to create an author page. Instructions for doing this can be found here.
- Minimize work in progress and work suggested but not started. Although it can be useful to put a place holder for an item, such as just its title or a couple of lines of content that are notes, please try to keep this to a minimum. It is more valuable to have submitted a few items that are complete and are of high quality than a long list of suggestions or partial submissions. Reducing work in progress makes it easier for you to see your own progress and for others to see the progress of the whole project. So, ideally, try to have no more than a couple of incomplete items at any one time.
- Nominate others. Contribution is by invitation only, but you can nominate others for inclusion by contacting me with your suggestions.
- Editing ethics. You have the ability to add or change your contributions at any time. To be a good participant, please edit your own contributions only. Be very careful that you don't accidentally alter someone else's work. As editor, I will limit my editorial changes to basic copy editing (spelling, punctuation, grammar, and formatting). I will discuss any other suggestions or comments on a contributed item directly with its author.
- Protect the privacy of our site. Please keep this URL private sharing it only with people you invite personally to contribute. Don't link to it, digg it, put it on your web pages, send it out to a mailing list, etc. First, it's only temporary. This project will not live within O'Reilly commons indefinitely. Second, we'd like to keep this under wraps until we have a decent block of material to release.
- Free of commercials. Please keep contributions free from references to specific products or technologies that compare their worth, or paint them in a positive or negative light.
- Legal stuff. All contributions made to this site are required to be made under the Creative Commons Attribution 3.0 license. This means that by making a content contribution, you are agreeing that it is licensed to us and to others under this license. If you do not want your content to be available under this license, you should not contribute it.
- Volunteers only. Contributions are made on a volunteer basis -- in other words, contributors are not paid for their contributions. The contributions will be made easily available to everyone on the World Wide Web for free. However, remember that those of you whose tips are chosen for publication will get your name attached to your work, your bio published next to it, and a free copy of the published book. Any item you contribute you can also reuse in any form you wish, such as in a blog posting.
- Submit only your own work. You warrant that all work that you contribute to this site is your original work, except for material that is in the public domain or for which you have obtained permission. Feel free to draw from your own existing work (blogs, articles, talks, etc.), so long as you are happy with the Creative Commons licence.
Please add your contributions in the subsections below, placing them and moving them to the subsection that best fits the state of a contribution. For guidance on the mechanics of how to contribute an item, please see How to Get Started. The following is an example contribution you may find useful as a guideline:
Contributions in this section are effectively complete from the author's point of view. Items placed in this section must meet the word-count requirements (at least 400 words and not wildly over 500 words) and the associated author bio must also be complete. Authors may continue to edit their own items in this section, and items will also be copy-edited, but the idea of completeness is that the items are potentially releasable.
- Fulfill Your Ambitions with Open Source by Richard Monson-Haefel
- Comment Only What the Code Cannot Say by Kevlin Henney
- Restrict Mutability of State by Kevlin Henney
- Speed Kills by Uncle Bob
- Encapsulate Behavior, not Just State by Einar Landre
- Only the Code Tells the Truth by Peter Sommerlad
- Interfaces Should Reveal Intention by Einar Landre
- Inter-Process Communication Affects Application Response Time by Randy Stafford
- Test for Required Behavior, not Incidental Behavior by Kevlin Henney
- Test Precisely and Concretely by Kevlin Henney
- Verbose Logging Will Disturb your Sleep by Johannes Brodwall
- The Road to Performance Is Littered with Dirty Code Bombs by Kirk Pepperdine
- Keep the Build Clean by Johannes Brodwall
- Use the Aggregate Design Pattern to Reduce Coupling by Einar Landre
- WET Dilutes Performance Bottlenecks by Kirk Pepperdine
- Testing Is the Engineering Rigor of Software Development by Neal Ford
- Make Interfaces Easy to Use Correctly and Hard to Use Incorrectly by Scott Meyers
- Don't Just Learn the Language, Understand its Culture by Anders Norås
- Small! by Uncle Bob
- Don't Nail Your Program into the Upright Position by Verity Stob
- You Gotta Care about the Code by Pete Goodliffe
- Know Your Next Commit by Dan Bergh Johnsson
- The Professional Programmer by Uncle Bob
- The Three Laws of Test-Driven Development by Uncle Bob
- Programmers Who Write Tests Get More Time to Program by Johannes Brodwall
- The Single Responsibility Principle by Uncle Bob
- The Longevity of Interim Solutions by Klaus Marquardt
- Prefer Domain-Specific Types to Primitive Types by Einar Landre
- Distinguish Business Exceptions from Technical by Dan Bergh Johnsson
- Don't Ignore that Error! by Pete Goodliffe
- The Boy Scout Rule by Uncle Bob
- A Comment on Comments by Cal Evans
- Don't Touch that Code! by Cal Evans
- Own (and Refactor) the Build by Steve Berczuk
- Deploy Early and Often by Steve Berczuk
- Understand Principles behind Practices by Steve Berczuk
- Acknowledge (and Learn from) Failures by Steve Berczuk
- Hard work does not pay off by Olve Maudal
- Continuous Refactoring by Michael Hunger
- Scoping Methods by Michael Hunger
- Improve Code by Removing It by Pete Goodliffe
- Learn to Estimate by Giovanni Asproni
- Domain-Specific Languages by Michael Hunger
- Learn Foreign Languages by Klaus Marquardt
- Check Your Code First before Looking to Blame Others by Allan Kelly
- Two Wrongs Can Make a Right (and Are Difficult to Fix) by Allan Kelly
- Floating-point Numbers Aren't Real by Chuck Allison
- The Linker Is not a Magical Program by Walter Bright
- Beware the Share by Udi Dahan
- Consider the Hardware by Jason P Sage
- Data Type Data Tips by Jason P Sage
- Reinvent the Wheel Often by Jason P Sage
- Economic Testability by George Brooke
- Consider Table-Driven Logic by George Brooke
- Put the mouse down and step away from the keyboard by BurkHufnagel
- Expect the unexpected by Pete Goodliffe
- Continuous Learning by Clint Shank
- Don't Be Cute With Your Test Data by Rod Begbie
- Choose Your Tools with Care by Giovanni Asproni
- Testing with Record / Replay by George Brooke
- Know Your Limits by Greg Colvin
- Do Lots of Deliberate Practice by Jon Jagger
- Code Is Hard to Read by Dave Anderson
- Simple Is not Simplistic by Giovanni Asproni
- Missing Opportunities for Polymorphism by Kirk Pepperdine
- Code in the Language of the Domain by Dan North
- Make the Invisible More Visible by Jon Jagger
- Ask "What Would the User Do?" (You Are not the User) by Giles Colborne
- Balance Duplication, Disruption, and Paralysis by Johannes Brodwall
- Methods Matter by Matthias Merdes
- The Golden Rule of API Design by Michael Feathers
- Don't Rely on "Magic Happens Here" by AlanGriffiths
- Prevent Errors by Giles Colborne
Contributions in this section are just short of being complete. The articles read as if complete but either just fall short of the minimum word count or the author bio is missing.
- One Binary by Steve Freeman
- Code Layout Matters by Steve Freeman
- Message Passing Leads to Better Scalability in Parallel Systems by Russel Winder
- Structure over Function by Peter Sommerlad
Contributions in this section are partially complete. They do not yet meet the word count requirement, or the author bios are incomplete, or the authors do not currently consider their items complete and potentially releasable.
- Duplicate to decouple by Klaus Marquardt
- Two hours of thinking can save two months of coding by Giovanni Asproni
- Know your IDE by Heinz Kabutz
- Hands on in all phases by Klaus Marquardt
- Implicit dependencies are also dependencies by Klaus Marquardt
- Learn reading and judging code, especially your own by Peter Sommerlad
- Understand SCM by Steve Berczuk
- Don't reinvent the wheel by Kai Tödter
- Use the same tools in a team by Kai Tödter
- Collection of Collections Is a Code Smell by Kirk Pepperdine
- Take Time to Read Other People's Good (and Bad) Code by Craig Larman
- Planning for performance is not a premature optimization by Kirk Pepperdine
- Reading Patterns by Klaus Marquardt
Contributions in this section only have a couple of sentences of suggestion or are just titles.
- Useful software is used longer than ever intended by Peter Sommerlad
- Measure, don't guess by Kirk Pepperdine
- Programming is a team sport by Pete Goodliffe
- Get the names right by Steve Freeman
- Plenty of domain types by Steve Freeman
- Logging is a user interface by Steve Freeman
- Reflection: Beauty or Horror? by Heinz Kabutz
- Don't handle errors, design them away by Michael Feathers
- Protection is a social problem, not a technical problem by Michael Feathers
- Start Small, Grow Big by Jørn Ølmheim
- Beauty is in Simplicity by Jørn Ølmheim
- Objects Want to be Asynchronous by Michael Feathers
- There is no Up or Down in Software by Michael Feathers
- Cohesion and Coupling matter by Tony Barrett-Powell
- Allow faults to be diagnosable by Tony Barrett-Powell
- Instrumentation for Quality Control by George Brooke
- Apply Functional Programming Principles by Edward Garson
This subsection is not for authored contributions, but for ideas you feel would benefit from being written about. You may not have the time, inclination, or background to write up a topic but still feel that it deserves to be covered. If so, please add a bullet below. And if you are looking for ideas for what to write about, please look at the list below for inspiration!
- Program to an interface, not an implementation
- Exception handling
- Don't put core application logic in the UI code
- Build tools
- Postel's Law and/or preconditions and postconditions
- Type conversions and type compatibility (languages have rules that can both help and hinder)
- Team and collaboration
- Algorithms and data structures (the importance of choosing the right ones)
- Hardware basics that can affect program performance (such as caching level and instruction pipelines)
- Memory usage
- The principle of least astonishment / law of least surprise
- Long argument lists
- Spikes, tracer bullets, and prototyping
- Design Patterns
- Code Metrics / Visualization