Not Round, Not a Table
When I looked into it, I quickly learned that a Round Table isn't an event held at a specific time for the benefit of a select few. The Round Tables are open day and night. They're liable to be busy at all hours, and if they happen not to be busy, you can easily make them so. You don't have to sign up. Anyone can read the content, and all TopCoder members are invited to start a new thread or to augment an existing one whenever they please. If you have something to say about a new programming textbook or video card, if you think someone might be able to explain just how a heap is useful in Dijkstra's algorithm, or if you need tips on cleaning grape soda out of your keyboard, the Round Tables are a good bet. If you don't know exactly what you're looking for, but you're fond of coding and enjoy making good conversation with intelligent coders, the Round Tables are exactly where you want to go. Once you've spent a few hours browsing the topics, you may, like me, find yourself irremediably hooked.
If "threads" and "browsing" put you in mind of the interactive web thingamajigs that are generically known as bulletin boards or discussion forums, it's because that's what the Round Tables are. "Round Tables" is TopCoder's name for its web-based bulletin board. The content is publicly accessible and unobtrusively advertised on TopCoder's main page: look for a modest grey link at far left, in the Competition navigation menu, sandwiched between Events and Support/FAQs. The Round Table pages are visually appealing and the threads are more or less logically laid out. The user interface works pretty much the way you expect it to. The features that set the Round Tables apart, establishing them as a higher class of bulletin board, are the singularly talented TopCoder member base and the endless variety of programming-related subject matter they are willing to discuss.
Like Greta Garbo, the Round Tables are an attractive institution with a peculiar name. The historical-mythological echoes are unavoidable but misleading. In the fifth century AD, at the dawn of the middle ages, the Romans withdrew their garrisons and abandoned their attempt to colonize what they called Britannia. This was a land of Celts, today known as England, that had yet to be settled by the Germanic tribes of Angles, Jutes, and Saxons. Legend has it that the last great Celtic king was Arthur Pendragon, who held court at Camelot. In King Arthur's retinue were many great knights such as Sir Tristram and Sir Launcelot, who, though noble and invincible, liked to quarrel with one another. In order to maintain unity in the kingdom, Arthur decreed that his knights convene periodically in Camelot and sit around a certain table, the one and only Round Table, to reason out their differences.
The name of the TopCoder bulletin boards, then, rings false in some respects. These Round Tables are not held at preordained intervals, and no pedigree is required to join the deliberations. Swordplay is kept to a minimum. Not a stick of furniture is required. If you like, you can take part without a stitch of clothing on your back, and no one will be the wiser. But what wisdom you stand to gain! Clustered around the Round Tables are some of the swiftest coders and sharpest minds in the world. From Hangzhou to Waterloo, TopCoder's top coders use this venue to trade information and opinions, to reach accord or to express colorful disagreement. Discussion is by no means reserved to the elite. The masses of middling and aspiring coders are highly valued participants, posing fruitful questions and making trenchant observations in their own right.
Members often use the Round Tables in ways that supplement TopCoder's core mission. After the frenzy of a Single Round Match, once the Arena has gone quiet and everyone has slept on the problems for a night or two, players gravitate toward the Round Tables to hash over difficulties encountered during or after the match. "Would someone like to take a stab at explaining John Dethridge's Div 1, Level 3 submission?" asked mckavity in reference to a recent match, and the forbiddingly dense code was duly dissected by schveiguy. On another occasion, the latter guy began a thread about Monty's Dilemma by saying, "This issue is really confusing me." After methodical reasoning from vorthys and others, he attained a state of enlightened bliss. A thread by Hurd about a different probability puzzle proved wildly popular, eliciting exegeses from ChristopherH that made my head spin. Meanwhile, (Lx.xx)(Lx.xx) pointed out a significant off-by-one error in the match editorial.
In the aftermath of an older match, when an outcry arose in the Round Tables over the unexpected results of certain floating-point calculations, jms137 stepped in to identify a bug in the Java compiler. TopCoder administrators soon responded with a software update. Although pointed complaints are best directed to management via email, the Round Tables offer a theater for lamentations of a freewheeling kind. These sometimes result in prompt, palpable change. When dplass pondered why member handles weren't clickable in the Round Tables, dok added this feature in a jiffy. Changes in the calculation of country rankings were brought about after protracted Round Table discussion involving members from every hue of the rating spectrum. Administrators have also initiated threads to request commentary on proposed new features or to make announcements of a nature not suited to official channels. How else would we find out the identity of the staff member who appears, in a photograph published on the website, reclining in a peaceful stupor after his corn-chip binge in the depths of the TopCoder Open war room?
Some of the most playful and most cerebral Round Table content ensues from musings on the theoretical ramifications of a match problem. A master of this art is ChristopherH, who outlined a proof that an unbounded-integer version of the Unefunge machine featured in SRM 152 is Turing-complete. The problem author, leadhyena_inran, had earlier committed himself to eating his hat and shoes as the result of a challenge, boldly issued by him, that gepa managed to answer. In a different thread, ChristopherH and radeye had it out over their favorite parsing techniques: which tastes better, recursive descent or operator precedence? In connection with yet another match, dgarthur made a splash when he wrote a formal proof using polynomial expressions to model a circular chain of red and green magic beetles.
There has lately been a vogue for solving match problems using as few characters of code as possible. Some have participated avidly in the one-upmanship, while others have decried it as an exercise in futility. Whatever one's view of this recreation, there is no denying that it exemplifies the energy of the TopCoder member base. While hip journalists and future gurus talk about "smart mobs" as a revolution in the making, the Round Tables prove that the smarts have already arrived. Distributed human intelligence is here. Instant collaborative problem-solving is a reality.
Then again, the Round Tables are just a fun way to share opinions and information. You can increase the fun by taking part. Yes, you. If you've never seen the Round Tables, try surfing through them today and tomorrow. If you've been reading without writing, reluctant to post your nagging questions for fear of mockery, I urge you to break free of your inhibitions. Hit the keyboard right now. Make your message concise, try to spell most of your words correctly, and post away. Many will read what you have to say. Some may respond. No one will laugh at you. I won't, anyway.
Even as I write about the attractions of the Round Tables, I am reminded of interface details that bother me. Some of the thread categories are difficult to tell apart: what exactly is the difference between Contest Participation and Challenge Discussion, or between Challenge Discussion and Algorithm Discussion? Furthermore, I have found the Search facility nearly useless in looking up past threads. Yet I can recommend Round Table participation with a clear conscience because I know that the TopCoder experience has steadily improved through member feedback. Not only is TopCoder built on the idea that competitive collaboration makes for better software, but the organization itself welcomes criticism to a degree unmatched by any other private enterprise. I have seen ample evidence that members can directly improve TopCoder by speaking up or pitching in.
The Round Tables are both a medium for constructive criticism and a living example of it. Oddly named, sometimes awkward to work with, they constitute a good idea that will only get better with continued and increased participation. It may be true that our participation is a good deal for the company because we voluntarily contribute most of the content. On the other hand, it's an even better deal for the members, because we're building an interactive knowledge base for each other, a free resource without peer. No other site on the web offers a greater concentration of programming talent. No other bulletin board lets you talk to so many brilliant coders about so many things that matterand many amusing ones that don'tin the theory and practice of programming. Whether you're a knight or a knave, there's a seat reserved for you at the Round Tables.
Would you like to write a feature?