Aleric Inglewood wrote: > It seems to me we have consensus on the need to add a new capability for this, > except from Linden Lab? To quote myself from 2 Dec:: > > I regret it that we don't have a good communication link with ALL LL > developers, > only with the handful that is working on snowglobe, or maybe I'm just missing > an overview. Is there anyone here who could say: Ok, lets change the > server-client protocol? > > That question was never answered... Merov? > > There's been talk forever and a day within Groupies/AWG/OGP/MMOX/etc about defining a protocol for a client-initiated request for a new service. My suggestion is that people pretend that some OGP-ish mechanism is in place (or will be) and come to a rough consensus about what it will look like and present it formally on the OGP list while simply going ahead and implementing it, with or without LL's stamp of approval. As long as everyone understands that the protocol for actually GETTING the url might change in the future when everyone (somehow) comes to agreement about The One Best Way to do these things, a given client, GPL-licensed or otherwise, having the ability of to change the URL on the fly won't break everything internally and shouldn't cause huge problems down the line.
Remember: the IETF counts working code higher than hand-waving, so if you're tired of LL's handwaving, get together with the realXtend, libomv, OpenSim, (and whomever I may have missed) and Just Do It. As they might say in the IETF: hum something up and get it working, THEN debate which is the best way (or ways) to do the thing. Doesn't mean you shouldn't argue while you're doing it, just that you're not going to get anywhere proposing something entirely new on THIS mailing list and then waiting for LL's blessing before starting. Lawson (ducks back into learning smalltalk and implementing Yet Another Proof of Concept) > On Fri, Dec 4, 2009 at 11:10 PM, Lisa Denia <[email protected]> wrote: > >> Carlo Wood wrote: >> >>> Yes, but in order to allow opensim to ever support this, >>> snowglobe should ASAP implement a capability to get the URL. >>> >>> Then, since opensim will report it doesn't have that capability, >>> we can also fall back to the old way of displaying map tiles. >>> >> Sounds like a very good idea to me. >> >> Dahlia wrote: >> >>> that may require the capability to also exist on the SL grids. >>> >> Of course. I don't know much about Linden Lab's internal server software >> releasy cycles, but is that a problem? >> Snowglobe is still something "experimental" as far as I know. Since >> introducing that capability doesn't break anything for Snowglobe users(the >> map >> could load a bit slower maybe), it can safely be added at the next regular >> server update. Snowglobe could even force HTTP map tiles when it detects that >> it's connected to Linden Lab's grid(maybe a bit dirty :)). >> _______________________________________________ >> Policies and (un)subscribe information available here: >> http://wiki.secondlife.com/wiki/SLDev >> Please read the policies before posting to keep unmoderated posting >> privileges >> >> > _______________________________________________ > Policies and (un)subscribe information available here: > http://wiki.secondlife.com/wiki/SLDev > Please read the policies before posting to keep unmoderated posting privileges > > _______________________________________________ Policies and (un)subscribe information available here: http://wiki.secondlife.com/wiki/SLDev Please read the policies before posting to keep unmoderated posting privileges
