December 3rd, 2004

Today's irritation is stupid game design.

I recently bought a copy of Rayman 3. Initially I wasn't going to bother with it, because I'd played the demo and wasn't exactly blown away by it. But I wanted Worms3D, and it turned out that at the MacExpo I could buy, for the same price as Worms 3D alone, a compilation pack with not only Worms 3D, but also Rayman 3, Ghost Master and the Britannica Children's Encyclopædia ("Children's", in the sense that when you look up R.E.M. you get the band instead of the sleep pattern). So I've now got Rayman 3, and I've found that it's every bit as annoying as the demo made me think it would be, if not worse.

The problem with the demo is, like most 3D platform games (particularly ones ported from consoles, it seems) was that the controls move relative to the camera, and the camera moves relative to a arbitrarily delayed version of the controls. Which sets up a feedback loop, and makes it really difficult to move with any accuracy in anything other than a straight line. The full version is just the same, except that sometimes the camera unexpectedly behaves differently, like showing the action from another angle, or remaining fixed and panning, or switching between all these different modes depending on which direction you face (which, if you're pressing any movement key, depends on where the camera is - another feedback loop).

I know I'm not the only person to think that Rayman 3 would actually be quite a fun and interesting game, if only the camerawork was a little bit more sensible. But the perspective moves so randomly, it just gets in the way of controlling the character and turns out to be incredibly annoying. This has been the death of so many 3rd person perspective 3D games, all the way back to the earliest (an example I remember particularly was called Galapagos, in which you guided a robot spider through an increasingly bizarre and dangerous collection of obstacles. You didn't control the spider directly, its movement was controlled by a simple AI. Rather, you manipulated the environment around it, with the hope of hinting and nudging it in the right direction. Unfortunately, the cameraman had slightly less intelligence than the spider, and much of the time it was completely impossible to see what was going on).

Update: Turns out Worms 3D has the same issue - you can only move relative to the camera position. Sometimes the camera won't go at a particular place because there's an object in the way - meaning it's impossible to move directly away from that object! That's not gameplay - that's working against deficiencies in the game design.

However, I digress. That's not what I'm most irritated by in Rayman 3. What I'm most irritated by is the way Rayman 3 continually cheats. It changes the rules with no warning. It stops you running through gaps which are clearly wide enough to run through. It doesn't let you climb on boxes and barrels if they're not the right box or barrel. But worse, is the fact that it's completely inconsistent, and as a player I have to constantly second-guess what the computer is about to not (not exactly a recipe for an immersive experience).

What I mean is that, for example, you usually finish a level by walking along a particular corridor, or climbing up onto a particular platform, or falling down a particular hole, but you don't know exactly which until it's too late and you've finished the level, which is a bit of a pain if they're other stuff you want to do beforehand such as collecting more gems or rescuing another of the Teensies who are locked in cages at various points. At one point, I had already seen a cage which I couldn't reach, but I was approaching a suspicious-looking tower. "A-ha!" I thought, "I'd better not go up there, otherwise the level will probably end and I won't be able to free this guy". So I wasted about ten minutes trying carious ways to reach him, before giving that up as a bad job, and just got on with the game and went up the tower. Whereupon I was given the tools to reach the cage and free its occupant. This wouldn't have been a problem if the game hadn't trained me into avoiding towers and other probably exits.

But even worse is when the behaviour of particular objects can change from one minute to the next. For example, in the very final act, the game sees you running around in a circular arena with an enemy who stands more-or-less in the middle. You finally manage to reduce his health to zero (this takes at least ten minutes in the best case, since you only get one chance to hit him in about thirty seconds), at which point he stabs himself with a sceptre of energy and grows to ten times his previous size. He stamps his foot on the floor, the ground shakes and you are thrown onto some rocks round the edge of the arena.

It's the same arena, but with one important difference: this time, if you step on the floor, you lose health and die. There's no warning, the floor looks exactly the same as it did thirty seconds previously. But it behaves like the pools of lava drawn previously. Couldn't they have given the player some kind of hint? They could have drawn fire on the floor, or something, or anything! Just don't present the player with one situation and then change it, especially in such a fashion that the player dies and, since there wasn't a save point and there's no quick-save function, has therefore just wasted the last quarter of an hour beating the enemy, only to die in a pointless and annoying way. Which is truly pointless and annoying.

April 8th, 2004

Today's irritation is Intego.

That's maybe a little harsh, but sometimes I think anti-virus companies are almost as much of a problem as the viruses they seek to cure. Not that I approve in any way of the electronic vandalism of viruses, trojans and such, but it's not entirely clear that the anti-virus companies are exactly squeaky clean either. Take for example the (many) companies whose product sends "You are infected by $VIRUS, detected by $PRODUCT_NAME" to the From address in the virus-laden email - even though they know that the virus forges such addresses. The person who actually is infected doesn't get the mail and the person who did get the message probably wasn't infected; they just get scared about it and read a company name who might be able to solve their problems. The only benefit of such messages is to the antivirus companies who might consider it free advertising (the rest of us might call it spam).

But I digress. Intego have done something different - it looks like someone noticed a potential security flaw to do with the way MacOS X presents files and file types to the user. He asked around on a Mac programming group to make sure he wasn't being paranoid, people there confirmed it was possible and one even made a test case (totally benign - it runs code but does nothing else). Here's a link to that thread on google groups.

Intego caught wind of this, and immediately issued a press release describing how the sky is falling, no-one can trust anything any more, claiming credit for the discovery, and by the way have you noticed we sell a product which will prevent infection? Buy it now!

From the press release: "While the first versions of this Trojan horse that Intego has isolated are benign, this technique opens the door to more serious risks. This Trojan horse has the potential to do any of the following:

• Delete all of a user's personal files
• Send an e-mail message containing a copy of itself to other users
• Infect other MP3, JPEG, GIF or QuickTime files

Due to the use of this technique, users can no longer safely double-click MP3 files in Mac OS X. This same technique could be used with JPEG and GIF files, though no such cases of infected graphic files have yet been seen."

At the moment, the Trojan horse has the potential to do none of those things, because all they've discovered is a test case (if reading a request for comments thread on a programming newsgroup really counts as discovering anything). Users can safely double click on anything if they know the source of it. There are currently no other infected files.

An infected file will have an Application icon (instead of an MP3 icon) unless the writer pasted the mp3 icon onto it - in which case he needs to find a way to maintain the resource fork. Packing it in a .sit file will do this, but this would be an unusual way of distributing an mp3 file and hopefully would tip off a potential victim that something was awry.

Intego paint the situation much worse than it currently really is, because they want ordinary users to be frightened of getting a virus. And that's because people who are frightened of getting viruses buy anti-virus packages. This is called spreading FUD (fear, uncertainty, doubt) - and the fact that they can profit from that is currently a greater threat than the symptoms of this so-called trojan.

April 4th, 2004

Today's irritation is browser incompatibilities introduced by non-conformant browsers.

Here's where most people would complain about minority browsers which don't behave the same way as standard platforms like Internet Explorer on Windows. But no, in this case I'm talking about IE and specifically its broken support for .png images and css position: fixed.

First the .PNGs. Spot the difference?

The first is a standard IMG tag with a PNG image in it. The problem is that I use an alpha channel to get the nice round edges, and IE manhandles that in a particularly bad way: it composites using the alpha channel onto a grey background and then puts the whole thing - grey background and all - onto the page. Other current-generation browsers can handle this properly without a problem. What's most baffling is that the Mac version of Internet Explorer - essentially unchanged for about five years - also handles it beautifully. Do they not share code between platforms in Redmond, or what?

The other is a massive hack which displays properly in IE >5.5 on Windows, and hopefully also works on other browsers without modification. It relies on image filters which are only implemented by Internet Explorer 5.5, the filter in the IMG tag prevents the (wrong) image from being displayed in IE, and the one in the DIV tag actually does display it properly. In any browser other than IE, both filter features are ignored, so the IMG tag displays the image file as usual.

This is of course incredibly ugly, I'd prefer to use straight html IMG tags but that would impair the page for probably at least 80% of people who might look at it. At least it doesn't use evil browser-identification code, which would have to be either something client-side - which would require ECMAScript (nee JavaScript) to be enabled in the browser - or something server-side which serves up different pages to IE users - which behaves badly with proxies and caches, in particular that people using different platforms who share a web cache may end up getting the wrong version.

As for position:fixed, I haven't found a solution for that yet. On Netscape-derived browsers such as Camino, or KHTML browsers such as Konqueror and Safari, or Opera, or even Internet Explorer on MacOS; the black bars at the top and left of this page stay in place if you scroll the text in the content area (in fact, if your browser can do alpha channels on repeated PNGs, the text will go under the drop shadow next to the black bars). It should end up looking something like this:

Internet Explorer on Windows, on the other hand, just seems to not support the position:fixed css property (or repeated png backgrounds, although you had probably guessed that by now). There might be a hideous IE-specific hack to fix this one too, but why should I go to the trouble of implementing that - can't IE just support the standards? Given that the Mac version of Microsoft's product doesn't have these issues, it would seem that they as a company don't have a problem understanding the specifications, nor a difficulty in implementing them. They must have taken an active decision to publish a broken browser!

Could it be that Microsoft are unsatisfied with their overwhelming market share and want to push for a 100% monopoly? Think about it - obviously a web designer wants to avoid having to code the layout twice, so many will code only for the one browser used by the majority of their target audience, and at the moment that means coding for IE on Windows. Microsoft seem to be attempting to contrive that a page hacked to look good in IE must be so hideously deformed that none of the other browsers can possibly display it (and a lot of web designers seem to have been persuaded that this is a problem with the other browsers), so the stubborn users who don't run IE yet will end up finding more and more broken pages until they finally give up in disgust and actually use it.

March 18th, 2004

Update:Play.com's site was updated on March 19th 2004

Today's irritation is play.com

It's a shame - they've previously been a good example of internet shopping done right. but I've had a truly appalling experience of customer service from them, and at the moment I don't feel like buying from them again. (Maybe it's something about companies with play in their name? MacPlay are still irritating)

There's a double CD out, "Anthology" by Sky. But the track listing that Play have listed on their site is not the same as the listing of the CD I actually received from them. In particular, the track "The Animals" (which I've been trying to find on CD for some time) is listed as starting CD 2, when in fact it doesn't appear at all.

I haven't tried to return the CD, because I'm quite interested in having the other tracks, and anyway I'd like the record company to get good sales so maybe they'll consider reprinting the rest of Sky's albums. But I did email play.com on 9th of March informing them of the error. I got back a canned response addressing a different situation:

Many of the CD's we supply are imported from North America. Whilst the majority of them will not differ from the European releases occasionally they may feature different track listings of alternative features. This is also why the release dates are sometimes different to European releases.

We would ask you to consider this information before ordering as we are not able to offer refunds on unwanted items that have been opened or played.

We hope that this information answers your query.

No, sorry, but that information does not answer my query (and if you'd actually read the email I sent you in the first place you'd have known that). The listing is on your site, you sold me the CD, and they are not the same. Any variation between different territories releases doesn't enter into the equation.

I replied (in somewhat less harsh terms than those) but got no reply back. I sent another mail, and got an autoreply on the 11th ("Your enquiry is being dealt with and we aim to reply to you within 24 hours") but nothing else since. I filled in their "Seen a mistake on this page? Tell us about it!" form but the site still has the wrong track listing on it.

I don't mind the mistake. Mistakes happen. It may not even be their fault (the record company's site also has the same wrong track listing - I emailed them on the 9th and the 14th but have had nothing back. That's appalling customer service too, of course, but since their business probably isn't geared up for dealing with end users it's easier to forgive them).

Play.com, on the other hand, exists to sell things to customers. Customers are their life-blood. They should talk to customers. And when customers have a problem with their product, it should be a vital part of their business to try to sort it out. And if play.com don't want to even say "thanks for the information, we'll update the site" then I daren't imagine what hassle I'd have in dealing with them if I actually needed to get money back. Maybe I should avoid putting myself into that position by buying from somebody else.

Previous Irritations

I have previously been irritated by the Carbon event model in MacOS X Jaguar, bad game design decisions, MacPlay, and parking at Cambridge railway station.

Am I wrong?

If you have any comments or questions about the articles here, send feedback