[sldev] [AWG] Zero's office hour on Tuesday, 2007-10-30
dirk husemann
hud at zurich.ibm.com
Tue Oct 30 03:39:59 PDT 2007
Tao Takashi wrote:
>
>
> 2007/10/29, Mark Burhop <mark.burhop at gmail.com
> <mailto:mark.burhop at gmail.com>>:
>
>
>
> On Oct 26, 2007, at 1:42 PM, Rob Lanphier wrote:
>> The conversation about scope here is pretty timely. Zero,
>> Liana and I
>> had a great conversation yesterday about the topic, and
>> decided it would
>> be good to make sure we talk about this at Zero's office hour on
>> Tuesday, 1pm PDT:
>> https://wiki.secondlife.com/wiki/User:Zero_Linden
>>
>> Zero's going to also repeat the discussion at the Thursday,
>> 7:30am PDT
>> meeting as well.
>
>
> It sounds like AWG is backing off scalability and is focusing on
> the openness aspect - especially openness of protocols. Not that
> they aren't related, just that the focus has changed. Is this
> correct?
>
>
>
> I am just wondering at which point we actually start talking about the
> actual protocol ;-) Even the scope discussion might be IMHO more
> useful if there is something to look at. It might not be final or even
> just a sketch but it might serve as an example for the discussion as
> well as might be a starting poing for interested coders to get going.
completely agree... how about we start with the flows: what
happens/should happen when we want to log in? full graphical login, bot
login. and then take it from there. i've listed a couple of flows that i
think would be good to look at on
https://wiki.secondlife.com/wiki/Talk:AWG_initial_flows
>
> As for the scalability I think with opening it up the proposed way it
> should already become somewhat more scalable as not everything has to
> be in one place. Moreover the sim is not doing everything anymore.
> This design will also enable people to experiment with different ideas
> for solving the scalability problem.
scalability should come with an efficient architecture. we need to keep
it in mind when designing the system --- same as with security!
>
> As for the scope I think for me it's rather clear that in the end SL
> like it is today (as in features, object model and probably also big
> parts of data structures) needs to run to make sense for Linden Lab to
> do that project. I also think that we need to find some sort of
> balance between overengineering it and keeping it too much constraint
> to the SL use case.
agree. i'd like to start with the flows --- let's use that as a guide.
cheers,
dirk
--
dr dirk husemann, pervasive computing, ibm zurich research lab
--- hud at zurich.ibm.com --- +41 44 724 8573 --- SL: dr scofield
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.secondlife.com/pipermail/sldev/attachments/20071030/a845a4c9/attachment-0001.htm
More information about the SLDev
mailing list