I hate SanDisk Cruzer Thumb drives. Right now, I am running a temporary internet cafe for a conference of oceanographers and meteorologists, and so they need to upload their presentations. Therefore, they plug in their thumb drives, and I’ve seen the performance, reliability, and design of different models and brands. After plugging in literally 200 different thumb drives, I really really hate the SanDisk Cruzer drives.

The software is slow and buggy. When you plugin a SanDisk Cruzer drive, it loads in software it calls U3, to enable connection with the thumb drive. This software is slow, full of eye-candy, and offers nothing new. And the most grievous point: Why must a thumb drive require custom software? I can understand drivers, but custom software? Please. No one cares, and no one wants the extra software. I cannot imagine a single extra use that the software could provide, beyond encryption or security. Even that is better and faster done by external products, like TrueCrypt, or GPG.

When I say slow, I mean slow. Every other brand, even old old usb drives are done loading in Windows XP in less than 30 seconds on four year old machines. The U3 software means I have to wait well over two minutes, with the drive appearing and disappearing not once, not twice, but thrice. When cheap no-name brands of thumb drives can be done on an old machine in less than 30 seconds, your software is slow.

Just don’t buy SanDisk thumb drives. Maybe they might get the hint, and stop forcing the abysmal U3 software on their consumers.

I was reading this very long and informative rant/lecture on the shortcomings of current rpg design. Except its dated, because it mentions Ultima Online! The saddest part of the rant however, is the fact that everything stated there still applies to the latest and greatest of rpgs, mmorpgs, etc.

I just had one idea to put out there, for the general improvement of mmorpgs as a whole. One problem with the addictiveness of say, WoW, is the unlimited play model currently in effect. The player pays ~$15 , and they could theoretically play 720 hours(One whole month, with no sleep, downtime, eating, etc).

The downsides of this model are readily apparent, in that being offered unlimited anything, is that we like to get the best bang for our buck. If you pay five bucks for a buffet, you’ll have several helpings, just to take advantage of the deal. It is a natural human tendency to do so, however, it has deleterious effects when applied to new players of an MMORPG.

It creates a pressure, both subconscious and cultural pressure, of playing as many hours as possible to take advantage of the theoretically unlimited play. This can, and does lead to addiction with the game, which in turn can destroy lives, just as any other addiction can do so.

So to combat this, the solution in retrospect is obvious: pay as you play. The idea is to offer new and casual players the opportunity to buy “hours” of game time. Instead of buying unlimited game time, the player instead buys 60 hours of game time, which can be used over several days, weeks, or months.

The benefits are simple and varied. The player can now relax, in that they have bought 60 hours of game time, and can use at any time, with no pressure to play every hour of every day. For the hard working 30-somethings that just want to relax and raid a dungeon with their friends, this is quite literally perfect. As well, it is an immense boon to not have to worry about regular credit card charges, since you’ve been charged once, and will not be charged again until you want more game time.

This is a good thing, for the players and the game designers. It produces a non-addictive game, and allows new players to literally try risk free, with no credit card numbers needed.

The more difficult question, however, is, Do you allow for players to purchase unlimited play alongside the pay as you play, or only do pay as you play?

I’ve been having some issues(again) with bittorrent programs. I only have these issues on Linux it seems(Ubuntu specifically). However, its the way they blow up that illustrates one of the key sources of bugs in code: error handling.

In the first place, there is Deluge, primarily a python bittorrent client. A very nice one… but it fails in a pretty mundane way. Right now, it will not add nor remove any more torrents. Rather frustrating when I wanted to download the Open Disc. I finally found what had happened. It appears that when I upgraded from Ubuntu 7.06 to Ubuntu 7.01, that deluge’s behaviour breaks. It was quite silent on any errors in fact, because they were not handled. And here’s the kicker, it will still have errors even after uninstalling, and reinstalling deluge.

The reason for this behaviour is that something the developers did not think could happen did, and therefore, unpredictable behaviour happened. Because they did not have anything in place in case there was an error they did not envision, the program is literally unusable. So, there goes my favourite torrent program.

I turn then, to Azureus… and it crashes too! This one was likely out of the reach of the developers of Azureus… as I believe it is an issue with Java itself. However, I’m still annoyed at the lack of good bittorrent programs for Linux. Well, good and stable programs.

The lesson though, is communicate with your users what happened. No matter the mistake, the bug, whatever, the users should still see something telling them something went wrong, instead of seeing the program start and behave wrongly. That is bad behavior.