Exporting stuff, permissions, and licenses (was Re: [sldev] realXtend Global inventory tests successful)

ordinal.malaprop at fastmail.fm ordinal.malaprop at fastmail.fm
Sun Aug 24 13:39:25 PDT 2008


On 24 Aug 2008, at 21:21, Lawson English wrote:

> ordinal.malaprop at fastmail.fm wrote:
>>
>> On 24 Aug 2008, at 20:56, Lawson English wrote:
>>
>>> ordinal.malaprop at fastmail.fm wrote:
>>>>
>>>> On 24 Aug 2008, at 19:46, Lawson English wrote:
>>>>
>>>>> I know it might have felt like this was a technocracy where we  
>>>>> started, but do recognize there was a tremendous amount of  
>>>>> technical framework that had to get laid down before we could  
>>>>> even get to the point of actually having a rational discussion  
>>>>> about what the options were, and we are now approximately at  
>>>>> that point.
>>>>
>>>> The thing is of course that requirements _always_ come before  
>>>> technical specs. Or should do.
>>>>
>>> Sure. But in this case, the problem space wasn't even remotely  
>>> well-defined, so trying to determine the "requirements"
>>> was like putting the horse before the cart when you didn't even  
>>> know what a cart looked like, or the horse, for that matter.
>>
>> I think that we did all know what the cart and horse looked like.  
>> Well, I don't want to stretch the metaphor too far - we knew what  
>> the desired outcome was, or at least, we were quite capable of  
>> coming to what the desired outcome was without reference to  
>> anything technical, because the desired outcome was not a technical  
>> one, rather concerned with how and when assets were distributed.
>>
>> Technical frameworks are what one develops _after_ the desired  
>> outcome has been decided on; technology is the slave of desire.  
>> Perhaps it might turn out that the desired outcome is technically  
>> impossible, in which case the specifications need to be revised.  
>> But they always need to come first.
>>
> At best, we knew what *A* cart and *A* horse looked like, but the  
> OGP is supposed to accommodate a number of different carts and  
> horses, if you want to continue with this, and we weren't even sure  
> if the current cart and horse would work in the larger metaverse.  
> We're still not sure, but at last we've discussed discussed more  
> fully the terrain that our favorite cart and horse might need to  
> face...
>

Well, why not, it is entertaining enough.

What I see is a lot of cart-makers and horse-breeders discussing what  
sort of carts and horses they think are appropriate, and then, once  
they've decided on that, presenting their models and breeds to the  
wider cart/horse-using community. Who may well then say "hold on a  
second, that cart won't haul turnips at all, and I want to move  
turnips from one place to another". (Or, in this instance, perhaps  
"that cart will haul any sort of turnip, and I only want my turnips  
going to certain people".)

This is the wrong way around. Functional specifications (merchant  
turnip-moving desires) should always, always, under every circumstance  
ever no matter what, be developed before technical ones (cart + horse  
turnip-moving solutions) and the latter should come from the former.  
Unless one is developing both functional and technical specs as the  
same person, which is still relatively rare, it just can't work  
otherwise.


More information about the SLDev mailing list