[MUD-Dev] Re: Modular MUD [Was:Finer points of Telnet programming ...]

[Caliban Tiresias Darklock]

On 10:00 PM 8/25/98 -0400, I personally witnessed Adam J. Thornton jumping
up to say:
>
>I'm not smart enough to write real programs without screwing up.  And I
>like it when my mistakes don't make me have to reboot, and I just get
>"Segmentation Violation [core dumped]", and then I can debug using the core
>file and my executable, and realize that I passed in the value stored at
>some place rather than its address, or something like that.

I don't generally have to reboot either. Visual C++ is pretty smart at
catching my bonehead mistakes before I ever run the program. Which is
another thing that bugs me about Unix; the development tools are very good
at doing what you tell them, but they don't generally tell YOU a damn
thing. I don't mind having to sleuth around to find stuff (Windows
certainly requires it enough), but I'd rather not have to do it ALL THE
TIME. Unix requires me to think of everything myself, and never jogs my
memory. Some people work better that way. I'm not one of them. I would
certainly be annoyed if Visual C++ popped up a dialog for every error and
expected me to fix it immediately, but I like it to sit down there and just
tell me what it finds. Ideally it shouldn't find anything, but at least I
can tell where I did something strange on purpose... and that's usually a
good place for a comment about why I did it.

>And with a modular, extensible framework, where you can't necessarily
>control the quality of the code being run, well, I'd be really wary of
>running it in an environment that didn't rigorously enforce process

>isolation. 

I used to co-own Telmaron, where an awful lot of would-be server hacks
learned what *not* to do to your MUD. A MUD is quite capable of bringing
the machine to its knees even WITH process isolation. Bad code is still
bad, no matter how well it's sectioned off. There are errors Windows
handles with ease, and errors Windows chokes and dies on. Unix may not
choke and die on most errors, but it can certainly become functionally
useless on some of them. Unix just suffers from code in *different* places
than Windows does. 

I'm sure there's some way of comparing those places and declaring one O/S
"better", but I'm of the opinion that operating systems are like languages
-- you pick the best one for the job based on what the job is, where it's
being done, and who's doing it. I certainly wouldn't tell a crack VB
programmer to work in C++ if VB would do the job, regardless of how much I
hate VB (a lot). And if I can write an application in five minutes using
Perl or half an hour using C++, chances are I'm going to use Perl. 

>Are there really that many people out there who *want* to run MUDs who
>don't know how to?  

Yeah, they just don't know it yet. Are there really that many people out
there who want to be on the internet? ;)

>The question is, you're going
>to want to support things you can't even think of now.  Is JavaScript
>extensible enough to do that?

The arrogant answer: Of course it is. 

The truthful answer: I hope so. :)

---
=+[ caliban at darklock.com ]=+=+=+=+=+=+=+=+=[ http://www.darklock.com/ ]+=
"It must be remembered that there is nothing more difficult to plan, more 
doubtful of success, nor more dangerous to manage than the creation of a 
new system. For the initiator has the enmity of all who would profit by 
the preservation of the old institution, and merely lukewarm defenders in 
those who would gain by the new one."              -- Niccolo Machiavelli
=+=+[ FREE KEVIN * http://www.kevinmitnick.com/ * IT COULD BE YOU ]+=+=+=