Zugg Software :: View topic - CMUD is Double-precision and 64-bit Integers!

Posted: Fri Apr 14, 2006 1:43 am

This will be happy news for a lot of people:

CMUD supports Double-precision floating point (compared to Single-precision in zMUD). And more exciting is that CMUD supports 64-bit Integer math (compared to 32-bit in zMUD).

This means that the maximum number in CMUD is 9,223,372,036,854,775,808 which should be large enough to handle all of those billions of experience points that you have.

My guess is that this will be a killer feature for a lot of people and encourage conversion to CMUD.

GURU Joined: 23 Dec 2000 Posts: 4834 Location: USA

should be, but it's definitely one of those all or nothing options. Simutronics (at least in Dragonrealms) doesn't even HAVE large numbers that bumped up against the old single-precision limit. Was very annoying to bump into this limit, though, when ZMud was operating in glorified calculator mode.

Magician Joined: 29 Nov 2005 Posts: 482

rofl mattlofton. Now that was a funny statement.

Posted: Mon Apr 17, 2006 11:42 pm

Okay, now the gauntlet has been thrown down.

Can anybody suggest a creative use for a 64 bit integer while playing a MUD???

Posted: Tue Apr 18, 2006 1:23 am

Take a look at some of the high-precision math scripts in the Finished zMUD Scripts forum. This has been a common request for MUDs that do not have a 32-bit "cap" on experience.

With 32-bit math, the largest integer is 2,147,483,648. That's 2 billion. There are some MUDs that have more than 2 billion exp. Just displaying the experience would be no problem in zMUD because zMUD just treats it as a string. But when you start doing things like calculating the amount of experience needed till the next level and stuff like that, then you need larger numbers.

That's the most common use I have seen and the main reason people have requested it (at least in the email I've received). I also didn't use to think it was needed, which is why zMUD was only 32-bit. But I got enough requests for larger numbers over the years and it was time to add it to CMUD. Also, the way CMUD is designed, making this change was very easy.

GURU Joined: 23 Dec 2000 Posts: 4834 Location: USA

Posted: Tue Apr 18, 2006 4:14 am

That's actually a difference between Single precision and Double precision for the floating-point math. I don't think it has anything to do with the Integer size (I'm assuming you really meant 1.0/32000 as a floating point value). Or yes, if you were trying to get around the floating point issue and handle it all as Integers where you'd multiply by 32000, then do some calculations, then divide by 32000 again later, then yes the larger integers will help with that too.

But the change from Single to Double precision will have a big impact on anyone who was trying to use floating point. Single precision is only accurate for a small number of places after the decimal (7-8 digits). So you'd get a lot of rounding errors and stuff like that. Double precision not only allows larger floating point values, but also provides more accuracy to smaller values (15-16 digits).

GURU Joined: 23 Dec 2000 Posts: 4834 Location: USA

err, yeah, that's what I meant.