« March 2005 | Main | May 2005 »

April 11, 2005

Towards documenting deployment best practices

Marc Brooks writes up an exhaustive account of the tools and practices he uses for enterprise development and deployment. The development tools will be familiar to anyone who's been tracking .NET best practices over the last four years, but the notes on the deployment environment, while inevitably specific to the author's particular deployment needs (for example, my boss-to-be, shaking his greying locks at my youthful zeal, has warned me that BizTalk deployment issues may well thwart my naive strategies for continuous integration and test), are well worth a look.

It will be interesting to see if deployment designers such as Whitehorse help to build a culture of common deployment best practices and supporting tools, similar to what we are already seeing in the development space (source code management, unit testing, automated build, etc.). Of course Whitehorse is only a static, physical architecture designer but what if we had a language for expressing deployment practices, like a content management system but for code instead of data?

Of course, they do say that code is data...

April 11, 2005 in Software | Permalink | Comments (1) | TrackBack

April 06, 2005

Vincent Baker, Dogs in the Vineyard

Is it unfair to judge a RPG based on two hours' experience with a GM who's never run the game before? Well, Dogs in the Vineyard outright encourages such quick-draw judgements, and heck, boy, he who lives by the sword, dies by the sword.

Actually, my overall impression of the game was pretty positive. Mike Sands, the GM, has posted a session account and some of his reflections, but I'll put it in my own terms anyway.

Dogs in the Vineyard is set in a religious Wild West environment, with the player characters as the enforcers of the religion. It's based on the early settlement of Utah by the Mormon Church, but carefully decoupled from any real-world religion. This is both an advantage, as it allows the group to be flexible about the teachings of the religion (and saves doctrinal character players from having to know any real religion), and a disadvantage, as the teachings the characters are meant to be enforcing are meant to be fundamental but, except for trivial cases, actually end up having to be worked out on the spur of the moment. (Our GM was wise enough to stick to trivial cases. And, to be fair, the source book does provide enough structure that a group can improvise around it without wrecking the feel of the thing.)

The combination of Western setting and fundamentalist religion works really well. There's a strong Johnny Cash vibe to the whole thing, something morally uncompromising amidst stark, dusty landscapes. It's a striking game world, possibly the most original and attention-grabbing I've seen since the original Deadlands. It's very easy to get into the spirit of the thing.

However, the details of the Dogs setup are potentially problematic. The characters, as I understand it, have essentially divine authority: whatever they judge, that's the Word of God. Whoa back. Even Paranoia has more checks and balances than this. Seems to me if our group had "solved" the problem by proclaiming the entire village riddled with sin and massacring every single person within it, we could still have just walked away without a blemish on our records, insisting that this was God's Judgment made manifest to us.

This problem seems to be exacerbated by a game mechanism called "demonic influence." Our game ended with us interrogating a suspect. The interrogation ended when one PC opened fire. We then learned that the more time we gave the villain to argue his case, the more demonic influence he would have had in the final conflict. If this was the author's intention, the only sane policy for a Dog is to shoot first and ask questions afterwards: trying to make sure you've correctly identified the guilty party is unnecessary at best (since your judgments are unquestionable) and harmful at worst.

In fact the whole thing had a whiff of Paranoia about it. The source book describes how corruption progresses in an individual or community: from the simple flaw of pride at first, through to sinfulness and then to outright heresy and consorting with demons. It is very easy to find examples of pride or sin in NPCs', or indeed fellow PCs', ordinary behaviour, if that's what you choose to do; and having caught a whiff of pride, why risk allowing it to fester and develop? Better to shoot it quick and claim divine authority. Our group certainly did not altogether avoid the urge to issue wild accusations, though we did manage to avoid addressing each other as "Citizen" or urging NPCs to "report for termination immediately."

On the plus side, and again like Paranoia or Werewolf, the "troubleshooting team" setup makes it very easy to get characters together and keep them together no matter what.

Finally, the system is innovative but appears fundamentally broken. The basic idea is that each side rolls a bucket of dice for all their applicable traits. The first side then "raises" any two dice. The second side has to "see" (match) that total from its dice: if it can match it with one die, it wins, with two dice, there's no effect, with three or more dice, it incurs that many dice of "fallout." The two sides then swap and another "raise" is carried out from the remaining dice. This is pretty simple and pretty neat.

The problem is that "all their applicable traits" clause. Firstly, this encourages abusively wide applicability. Any sane munchkin will throw all their character generation points into a trait called something like "determined to succeed at anything I do." Then whenever anything is going against them they can go, "Ooh, I'm determined to hit him with my blunderbuss / win the Eagle Falls Branch waltzing competition / ram a cucumber up Satan's butt -- can I roll my extra dice now please?"

Secondly, and more seriously, it requires people to cast around for ways of bringing new dice into a conflict. This appears to be an explicit aim of the design ("escalation"), but for me it is deeply distorting. Suppose you're sat up a tree sniping at a bad guy. Once you've fired once, you've got all your stealth and gunplay dice: you don't need to keep sniping to hang on to them. Instead, you're better off leaping out of the tree and hitting the guy with a stick, because then you not only get to keep your stealth and gunplay dice, you get to add your leaping and drubbing dice. And once you've had a go with the stick, throw it away quick as you can, so you can get your unarmed combat dice to bear as well. Similarly -- and I've seen this problem in HeroQuest too -- the system tempts players to waste time trying to bring irrelevancies to bear: "if I quote the Book of Jeremiah while shooting him, can I add my rhetoric dice please?" The negotiation slows things down and the introduction of irrelevancies for skill reasons makes conflicts artificial.

Overall? The system problems need fixing, because they interfere with the game, but as with most system problems, this is easily done by tweaking the system or by GM clampdown. The Paranoia-without-the-checks-and-balances-isms may or may not prove to be a problem long term: one trial game is not enough to say. It would certainly be a shame if the game gave too much encouragement to the "shoot first and ask questions later" approach: the setting is wonderfully atmospheric, imaginative and innovative, and it deserves to be explored in depth, not just used for target practice. Well worth a look.

April 6, 2005 in Games | Permalink | Comments (4) | TrackBack

April 03, 2005

Prevention of Terrorism Act 2005. Again

Woolamaloo Gazette: "Why are the government trying to force through a new Terror Bill? Well, put basically the existing Terror legislation brought in in the wake of 9-11 is about to expire. If a new bill is not passed quickyl [sic] the United Kingdom may well run out of Terror. Naturally this would be a national disaster since the electorate may then raise their heads from their government-supplied ostrich holes and have a look around."

April 3, 2005 in Current Affairs | Permalink | Comments (4) | TrackBack

April 01, 2005

Blogging the Prevention of Terrorism Act

The Last Ditch is "a blog about the death of liberty in Britain." The author is an expat living in Russia. He nails why I left the UK. "I am back in England today - for the first time since habeas corpus was abolished. It feels odd. Everyone is going about their business as normal, as if nothing had happened... This does not feel like my country any more."

April 1, 2005 in Current Affairs | Permalink | Comments (1) | TrackBack