[DGD] Pros and cons of distributing mudlibs via statedumps

[Steve Wooster]

DISCLAIMER: I'm making my mudlib for fun, without real expectations of 
finishing it, so it's unnecessary to point out that it's better to start 
with the kernel-lib.

If I ever get far enough to distribute my mudlib, I'm thinking of using 
a statedump instead of zip-files with source-code... the reasons are 
that the file-system is stored internally and my idea currently has some 
chicken-and-egg problems which using a state-dump would solve easier 
than a start-up script.

I was wondering if there were any disadvantages to using statedumps to 
distribute mudlibs... It'd be possible to import/export file to the 
outside file-system, so working with the source-code wouldn't be a 
problem. I'm a little worried about things like maximum array sizes and 
such... Is it ever possible to decrease the maximum array size from that 
used by a a statedump? ...or would I need to make sure that the 
distributed state-dump had minimal sizes so people people wouldn't be 
stuck with ones larger than what they wanted?

Are there other things I've missed?

-Steve Wooster