[sldev] Two questions: Region geography and region handles

[Argent Stonecutter]

In my response to a private message offlist I realized I may have  
misled the gentle readers of this list as to my motivation for one  
comment in a recent message, so with no further ado:

I wrote:
>>> Alas... there's no way to get region size from LSL. This means   
>>> regions larger than 256x256 will break so much in-game content I   
>>> don't think it's practical in the foreseeable future.

Thomas Grimshaw responded thus:
>> I hate this philosophy of "don't do that, it'll break user content."
>

> I have an ambivalent attitude towards this philosophy. One reason  
> Windows remains popular may well be that Microsoft has been over  
> the long run less likely to deliberately break working software  
> than their competition. On the other hand when Microsoft makes a  
> system call return an inconsistent value because a DOS program  
> misinterpreted it as a file descriptor back in the '80s but it  
> happened that the error value in 1990 happened to be a valid file  
> descriptor... well, I think that may be going to far.
>
> Regardless, it *is* a Linden policy, which is what I meant by "I  
> don't think it's practical".

The path forward, I believe, would be for LL to create something like  
llGetRegionInfo() (as in <URL: http://jira.secondlife.com/browse/ 
SVC-1554 >) and state that assumptions about the meanings of specific  
region coordinates should not be relied upon as of $DATE.