entropia

I’m sure this information is already all over the place, but I couldn’t find a straightforward anwser with a simple search.

How do I get to the Management Studio after installing SQL Server with VWD 2008 Express?

Here’s the situation. I just installed Visual Web Developer 2008 Express, and checked the SQL Server box to install that too. The installation went smoothly, but I could find any tools for actually using SQL. 3 links were added to my start menu under Microsoft SQL Server 2008 -> Configuration Tools: SQL Server Configuration Manager, SQL Server Error and Usage Reporting, and SQL Server Installation Center.

None of them were what I wanted. I tried the installation center, which had an option to add features, but it didn’t give me any features to add.

So I decided to download and install SQL Server separately. It comes in 3 flavors - basic, with tools, and advanced. Naturally I chose advanced, because I didn’t want to miss out on anything I might need.

When I run it, it looks just like the installation center I already have. I tried installing a new instance, but it said I already had one. I tried repairing and adding features, but it didn’t let me add anything. Finally, I found Edition Update, under the Maintenance section.

When I tried it it seemed to upgrade to the “advanced” version I downloaded. Appearantly, installing SQL Server with Visual Web Developer installs only the most basic version, and the tool I needed are only available in the more advanced versions. Once that finished, I was able to go back to the “add features” option and add all the tools, which filled my start menu with another half-dozen links.

So, with everything said and done, here’s a quick step-by-step:

1. Download VWD 2008
2. Install VWD with SQL Server
3. Download SQL Server 2008 with Advanced Services
4. Run and go to the Maintenance section, then select Edition Upgrade
5. After the upgrade is finished, go to the Installation section, then select New SQL Server stand-alone installation or add features to an existing installation

I’ve been listening to game journalists at places like 1Up gush about Little Big Planet for what feels like years. And no matter how great they say it is, I’ve always had the same reaction: “Huh?” Is that it? A platformer that lets you make and share your own levels - that’s supposed to be new, original, and paradigm-breaking?

Games have come with content editors for years. Doom was one of the first games with mod support, 15 years ago. The most popular multiplayer shooter in the world, Counter-Strike, started as a mod. A lot of success can come from mods, but you can’t rely on that as your #1 selling point.

So you say Little Big Planet makes it easier than ever? Doesn’t matter. Anyone with the design skill and commitment needed to make a fun game is willing to use a more difficult, but more powerful and flexible, editor than Little Big Planet could deliver.

But then I changed my mind. Via Tales of The Rampant Coyote, I found Fantastic Contraption. It’s one of those physics-based games where you build a machine and let it run. I usually don’t like these games, but Fantastic Contraption is a nice combination of simplicity (there are only 5 objects to build with) and freedom (any puzzle can be solved using a variety of approaches). Once I started playing, I couldn’t stop. I haven’t been so entranced by a game in years.

And once I finished, I immediately bought the game and spent hours browsing through the user-made content and solutions. (And I have to mention, it was the most fluid game-buying experience I’ve had. Clicking Buy opens a PayPal window, and once you hit submit, you immediately get all the benefits without even closing the original game window.)

But what really surprised me about it is that it opened my eyes about Little Big Planet. It’s not just about making it easier to make and share stuff, it’s about the fun of the experience of actually making it. Things like the Spore Creature Creator, Line Rider, and Armadillo Run never appealed to me, so I didn’t believe any similar game would. But if Fantastic Contraption can do it, maybe Little Big Planet can too.

And it’s not just the design phase. Fantastic Contraption also showed me the power of a good community, which I’ve never really cared about. You can save and share your designs with a single click, and there’s plenty of activity in the forum. This is supposed to be another area where Little Big Planet excels.

Still, the games are very different in their goals. As far as I know, the actual “play” portion of Little Big Planet doesn’t involve any design - it’s basically just a platformer. And the design phase doesn’t have any real goal or direction. In my mind this is still a huge hurdle.

I recently started playing Bioshock again. I liked the game when it game out, but I thought it didn’t live up to the hype. But now I’m more excited about it than I was when I started playing it the first time. The story, the atmosphere, and the gameplay are sucking me in. If I were to review the game after playing it the first time, I would have given it a Very Good. Now I feel like it deserves a Great - one of the best games in years.

Game reviewers usually don’t play games like the rest of us. I play games in my free time, which could be a few hours one day and a few minutes the next. They play games as a job, for hours every day. And they have deadlines, so there’s no time to let it all sink in, or revisit their favorite parts. For games like Bioshock, where the lasting impression is important, this doesn’t work very well.

But even worse is the end result of the review. Even if the reviewer writes a masterful article, detailing the highs and lows, it all comes down to the final score. Whether it’s out of 5 stars or 100 points, most people only care about the number.

I’ve seen two sites with better takes on the reviewing process. The (now defunct) Game Chair would review games sitting by sitting and wrap everything up at the end. This way the rating isn’t a single number, but a rolling accumulation. Maybe the end of the game is significantly better than the rest, or maybe the reviewer just isn’t having a good day. With this system, you get a better idea of what the reviewer’s experience was.

The other site is Kotaku, which does away with numbers completely and just gives a short description with pros and cons. This way the reader doesn’t have to guess what 7 out of 10 means. They can give their own interpretation. “The biggest pro is the engaging story, which doesn’t really concern me, and the controls are sometimes frusterating, so even though it sounds like the reviewer really liked it, I’ll pass”.

I do something like a combination of these two methods for myself. I keep a diary-like game log of what I play each day. It’s not for mass consumption, so it’s full of broken sentences and incomplete thoughts. But if I made it presentable and condensed it to just a few days, my Crysis “review” would look something like this.

Day 1
I’m excited for this game. Farcry was good, and Crysis looks like more of the same, with better everything. Especially the graphics, of course.
After a couple levels I’m getting used to the nanosuit, not using much except Armor and Stealth. And I feel like the guns a either underpowered or too inaccurate. Unless you hit an enemy in the head when they’re not looking (because if they are looking it doesn’t do as much damage), it takes a full clip to put them down.

Day 2
After a few more levels, everything has gotten way too repetitive. It’s just Stealth for 5 seconds, wait, and repeat. If you’re spotted, the enemies are quick to forget about you. And the nanosuit isn’t panning out. It’s still just Amor and Stealth, or Speed to save time going through an area you’ve cleared.

Day 3
Despite the repetition, it’s pretty fun after coming to accept the gameplay, which is mostly learning how to deal with the AI. At first I thought the AI was terrible, but it’s actually pretty good. Not “good” as in smart, but “good” in that it feels like there’s actually something making decisions. And after playing for a while you can learn their tendancies and figure out how to exploit them.

Day 4
All of a sudden, I feel like I’m playing a completely different game. After running freely around a jungle for hours, I’m now doing escort, rails, and defend missions. There’s no more playing against the AI (and hence, against the developers, something I’ve always loved). It’s just playing against arbitrary, soulless entities.

Day 5
I can’t continue because of a memory leak. I can’t say I mind, though, because I haven’t enjoyed the game for a while.

Conclusion
Just like Farcry, it didn’t quite live up to my expectations, but I had some fun. The story never grabbed me, and I felt like I could give up at any time. But I stuck around because I really enjoy exploiting the AI, and because the graphics are simply the best you’ll find today.

I think if someone would take a system like this, and give it to a reviewer who can write well and make better points than I can, it would be a good model for reviews. It doesn’t directly fix the problem I would have reviewing Bioshock, but that can be fixed. Since there’s no final score, or a preset timeline, you can always go back and add to your review. So if I were to fire up Crysis again, and have a different experience than during the first time through, I could update it.

Of course, the major review sites will never do something like this. The industry is too focused on the numbers.

Games released in the months prior to The Legend of Zelda: Ocarina of Time:

• Half-Life
• Grim Fandango
• Fallout 2
• Metal Gear Solid
• Unreal
• Starcraft
• Tom Clancy’s Rainbow 6

Games released in the months prior to Grand Theft Auto 4:

• Generic Shooter Part 3
• Generic Brawler Part 3
• Other Generic Shooter Part 4
• Super Mario Galaxy

1998 reads like a Top Games of All Time list¹. 2008 hadn’t had anything memorable until GTA4. Sure, Smash Brothers Brawl was huge for fans of the series, but it’s nothing new like MGS or Rainbow 6. You have to go back to last year for Unreal 3 and Call of Duty 4, and they’re nothing compared to Half-Life, which is still on almost everyone’s top 10 list².

I’m sure GTA4 is great. Not as fresh as the previous versions, but maybe more fun. But I believe its place as the highest ranking game of all time according to GameRankings.com is inflated due to the complete lack of anything else.

And here’s something else I thought was interesting. The user rating for GTA4 is around 8.0. The user ratings for the other top games on GameRankings are must higher, mostly around 9.0. I’m not sure what that means.

1. Not my list, not by a long shot [↩]
2. Except mine [↩]

I recently switched to a color theme with a dark background in Visual Studio 2005. The choice was almost purely logical: there’s more room for color variations. A light background will have dark text and a dark background will have light text, and the differences between light colors is more noticeable than the difference between dark colors (eg. compare #5f0000 and #00005f to #ff0000 and #0000ff). Therefore, the different elements of the syntax will have more variety and be easier to understand.

That’s my theory at least. We’ll see how it pays off.

I haven’t settled on a particular theme yet. Scott Hanselman has given some suggestions. Jeff Atwood has written about it, of course. So far I’m using Dark Visual Studio from Brad Wilson. I prefer its blues over some other themes’ browns.

Initially, I had a hard time seeing and understanding my code. I was looking for green comments, but they were purple. But I realized that the problem wasn’t the colors, it was the code itself. Even though I had written it the day before, and was looking at it just minutes ago, it become gibberish with a new theme.

The problem is that, when you’re writing code, you get used to seeing the shapes and structures of the text, more than the code itself. That method with a couple long green lines at the top and a too-deep if-nest at the bottom? Yeah I know what that does. But when the green code is purple and the blue statements are orange, it’s brand new.

I immediately refactored a large portion of the application.

It’s good to be familiar with your IDE, including the syntax highlighter, but I think it’s also good to step out of your element every now-and-then. If you’re working with the same code and the same colors for weeks, you can lose sight of the deficiencies. Change helps bring them back.

There’s one thing left though. This was just for VS2005. I also use VS2003 at work, and I sometimes use VS2008. Then there’s Notepad++, for anything non-.Net, which has dozens of themes itself. I really with there were a standard settings format that could be used by all major editors. Better yet would be if they all use the exact same file, so a single file in your user directory controls the highlighting in all your editors.

Software development is like making a table. It can be as fun, frusterating, time-consuming, rewarding, and difficult as you want.

Someone tells you the dimensions, the shape, the color they want, and acts like that’s enough to get a definitive estimate. A square, 4′x4′ table can be made in under an hour. Just nail a few boards together. Or it can take days (maybe weekx, I don’t know, I’ve never made a table), if you want to make it sturdy and robust, with intricate carvings, maybe a hidden drawer, and whatever other frills you can think of.

Of course, if you go to a table maker, you probably have a better idea what you want. I don’t think that’s the case with software development. I was asked to give some estimates today, to add some functionality to an existing app, and I’m just hoping I can get them right within an order of magnitude. Getting the functionality working, in the sense that the app can do what they want, if they ask it just right, any never want to change anything, would take only a few dozen hours. Getting it working right, so it won’t crash, and it errors gracefully, and it can be changed without a complete rewrite, and it has tests to verify that it works, might take a few hundred hours.

What type of table does your client, customer, or boss want? I think it’s usually safe to assume they don’t want the one-hour table. And the fancy table might be too much - they’re more concerned with the functionality. But that still leaves a lot of room to work with. If you try asking what they’re looking for, they’ll say they want a fancy table for a one-hour table price.

The trick is to give the estimate they’re looking for. When someone says, “give me an estimate”, what they mean is, “I’m thinking of a number”.

I’ve updated my JavaScript raycaster. It doesn’t look much better than before, but it works better. I’ve also included a preview screenshot for the next version.

By now we’ve all had a chance to try out IE8, and we’ve all found quirks, bugs, and pages that don’t quite work right. PPK has a summary of CSS issues. John Resig talks about JavaScript. Erik Arvidsson gives his impressions. I’m sure there’s plenty more out there.

I’ve found a few interesting things myself. Navigation on sites like Sourceforge and Codeplex doesn’t look right. MSN TV listings don’t work. But the biggest issues I’ve seen are with my raycaster.

There are basically 3 causes of the problem: opacity, overflow, and zIndex. IE has never done opacity the same way as Firefox or other browsers; it uses filter: alpha(opacity=50) instead of opacity: 0.5. IE8 doesn’t seem to have any opacity support at all.

More noticable is overflow. I was using large spritemaps for the enemies, which were clipped to show a single sprite¹. So the Sarge sprites are all in a single image file, and different parts of the image are shown for different individual sprites. In IE8, the images wouldn’t clip and the entire spritemap would be shown.

I made a CSS overflow test to pinpoint exactly what the problem is. I was using absolutely positioned imgs in a relatively positioned div. In Firefox and IE7, the children of relatively positioned elements are clipped, but not in IE8.

To fix it, I’m using a staticly positioned div for the container, and positioning all the content (walls, sprites, and sky) relatively.

The third problem is with z-index, for which I made a CSS z-index test. In IE7, there was no interleaving of children of siblings. So if 2 adjacent nodes, A and B, each have children with different z-indexes, there is no way for A’s children to be layered between B’s. Either all of A’s children are in front of B’s children, or they are all behind B’s children. And it all depends on A and B’s z-indexes, not their children’s. In Firefox and IE8 there can be interleaving if the z-indexes of the parents aren’t set. Also in IE8, there can be interleaving if the z-indexes of the parent are set to 0.

For more about IE8’s CSS handling, there’s the Windows Internet Explorer Testing Center and CSS Improvements in Internet Explorer 8.

1. I did this so that only a few images needed to be downloaded. Instead of one for every frame and angle for each enemy, there is just a single image for each enemy type [↩]

With the recent release of Internet Explorer 8 Beta 1, everyone’s talking about web standards again (or rather, they’re still talking about them, but more loudly now). Is IE8 standards-compliant? Will Firefox 3 be standards-compliant? What does it really mean to be standards-compliant?

One of the more popular ways of evaluating standards-compliance is with the Acid 2 Test, which IE now passes. Or does it?

Although we said that IE8 Beta 1 passes the ACID2 test, some of you may be seeing results like the image above; we thought we should explain what’s going on. IE8 passes the official ACID2 test hosted on http://www.webstandards.org/files/acid2/test.html. (Note, this seems to be a popular destination at the moment. You may have trouble reaching the site.)There are also a number of copies of this test around the net. One popular copy that I’ve seen of late is http://acid2.acidtests.org/

IE8 fails the copies of ACID2 due to the cross domain security checks IE performs for ActiveX controls.

IE Blog

What I don’t get about all this is why anyone cares whether an arbitrary set of HTML and CSS renders the way an arbitrary group of people say it should. The Web Standards Project and the W3C do not set standards. The browser makers do. More specifically, the popular browsers do. So right now, Microsoft sets the standards and by definition, anything that works in both IE6 and IE7 is standards-compliant. That’s what a standard is: “an object that is regarded as the usual or most common size or form of its kind”. If it’s not implemented by most browsers, it can’t be a standard. And right now, IE6 and IE7 make up “most browsers”:

Designing a web page around “standards” that don’t work in IE your site will be broken for about 3/4 of your viewers. When writing and testing your HTML and CSS, you should do it in this order:

1. Make it work in the most popular brower.
2. Make it work in other browsers, to the point where the extra effort isn’t worth it.
3. Make it adhere to the suggestions of the Web Standards Project.

So some sites, IE6 and IE7 aren’t the most popular browsers. For a lot of “tech” sites, Firefox is just as popular. For Apple-focused sites, Safari probably is. If that’s the case, then go ahead and ignore IE. You may have trouble attracting new readers though.

As for IE8, there’s one site I’ve found that works in IE7, Firefox 2, and Opera, but not IE8: my raycaster.

I was really excited about Final Fantasy XII, but so far it’s been a let down. It seems like every few minutes you run into something that just isn’t right - whether it’s gambits that don’t work like you want or just having to stop for 2 seconds and sheath your weapons when fleeing¹. I was planning on writing a big post on all the little details that are broken, but instead I’m just going to suggest a few changes that would make the game much better.

Cut back on the randomness

In all the previous Final Fantasy games I’ve played, spells like Stone, Sleep, and Death only work a fraction of the time. If you cast the spell, the entire battle depends on the random number generator. That’s devastating randomness, and I hate it.

FFXII cuts back a little - by making spells like Stone and Death start a countdown timer, so they’re not instant kills - but it’s not enough. Spells like Sleep and Slow still only work every once-in-a-while, so the battles as still controlled by the random number generator. Whenever a spell results in a “miss”, it’s an instant Fail for the game designer. Spells should either hit (with some degree of power) or be ineffective (if the enemy is immune). There is no miss.

This is easy to fix. Instead of making Slow “hit-or-miss”, just make it effective to a certain degree. Sometimes it’ll slow the enemy down by 10%; sometimes by 40%. Same with Sleep - instead of “awake” and “asleep”, let enemies be “sleepy”. And as they get more “sleepy” they may slow down, miss with their attacks more often², and block fewer attacks.

Fix the Gambits

I love the idea of Gambits, but they don’t quite work as implemented in FFXII. I think there are 2 main problems, the first being that they’re too specific, with too many choices. Instead of “Cast fire on enemies that are weak against fire; cast thunder on enemies that are weak against thunder; etc.”, there should be something like, “Cast whatever will do the most damage to this enemy”. Of course this brings up the question of how the character should know which spells are the most effective. The solution is simply to include it in the beastiary. If you find a spell that’s especially effective, the character should take a note of it and use it again next time.

The other problem is that there’s no way to easily change gambits. There’s a nice interface, but it requires a lot of clicks. And until you spend the time to reconfigure everything, you keep using the same settings. It would be nice to be able to set up different classes of gambits - like Battle gambits, and Safe-zone gambits, and maybe even Aggressive or Careful gambits. So once you enter a battle, you switch to battle gambits, which might place emphasis on attacking until characters are especially low on HP. At the end of a battle, you’re back to the safe gambits, which involve more preparation for the next battle.

Combining all this, and taking it further, you could get rid of the list of instructions (”if A do B; else if C do D; else …”), and turn Gambits into equippables. So a single gambit that you find or buy would consist of an entire battle strategy for a single character, and you just assign each character a gambit (or a couple - one for battles and one for between battles). And different gambits would be tailored for different classes - tank gambits, mage gambits, healer gambits.

About the closest thing to gambits I’ve ever seen successfully implemented was in Ogre Battle. In Ogre Battle, you had no direct control over your characters, you just gave them simple instructions like “attack leader” or “attack weakest”, and the AI took over. This particular system wouldn’t work in a Final Fantasy game, because the battles are too different, but I think it’s something to consider. The best solution might be somewhere between FFXII and Ogre Battle.

Don’t make my characters stop

I don’t know where the idea of stopping to cast spells, or stopping to sheath a weapon, came from, but it needs to be sent back. In FFXII, whenever a non-leader character casts a spell, he has to stop for the entire duration of charging the spell, casting, and deciding what to do next (the leader can move for everything except the actual spell casting). This accomplishes nothing but annoying the player.

Fix the bosses

Maybe the worst part of the game, and the hardest to fix, is the bosses. I don’t mind a long boss battle, as long as they’re not so boring and common as in FFXII. I just spent 15 minutes fighting one boss, casting the same spell again and again, and the very next battle is another 15 minute boss. I used to enjoy the Final Fantasy bosses, but not anymore³.

Every boss feels the same. First learn its attacks and equip something to nullify the status effects it throws at you. Brace yourself when it gets down to 50% HP, because it uses a super-powerful attack. Again at 25%. And for the final 10% or so it just attacks relentlessly.

Mix it up a little. Throw in one that switches its weakness every few minutes. Or one with multiple stages. I think it was FF3 (that’s FF6 to Japan) that perfected the multi-stage bosses.

Overall, I’d say the Final Fantasy XII just feels unpolished, which is especially surprising for a game of that name. I was hoping it would be good enough to make me play the games I skipped in the series. I still do want to play them, but for a different reason: I want to see the full progression from 6 to 12. Each game in the series adds new gameplay elements, and I want to see what I’ve missed.

1. Seriously - if try to flee you have to stop and sheath while the monsters catch up to you - fleeing actually slows you down [↩]
2. I can tolerate blocks and misses with regular attacks because they rarely have such a drastic effect on the overall battle [↩]
3. I can’t remember exactly what it was about FFVII that made be stop playing the series, but it could have been the bosses [↩]