OAS Discussion Continued

[Regas]

Thanks for the post Max, with the exception of limited scope, what you outline is pretty much exactly how we've structured the new OAS. We've tried to duplicate the old structure as best we could, adapting it to a single live server.

As for opening OAS to the other servers. I'd suggest we give the MS/OAS live structure some time and see how well it floats. Later if there's a strong call to open the other servers up to OAS or to rotate them, we can look at it. I think anything as grand as trying to open it to all servers would require support from the HDMs and a community vote, as we've been doing lately. When we pushed for this open OAS originally it was with the understanding that it would be only one live server. This allows DMs and players the ability to choose not to mix with the oas.

Also, one really big difference between oas and just dropping the pass altogether is we are capping player non-member advancement at a low level until they are accepted.

[Fionn]

[yadda yadda] trivial to fork the module to put a low level area (e.g. Rivermoot) on it's own server.

[The right way's hard]

Yea, I was meaning more copy it and delete everything you /didn't/ want duplicated Unless TSM is a bear to run already, that should be good enough. If anyone *cough*yeahright*cough* has the time, it can be cleaned up later to kill unused scripts et al.

[Xanthea]

Why would you even want to make it apply to all servers? Surely it's better to have them all on the same server for greater ease of new players with new characters playing together. Less of a level gap means a better time.

[Ronan]

I haven't dealt much with NWN2, but with NWN it would have been trivial to fork the module to put a low level area (e.g. Rivermoot) on it's own server. Can that be done without killing the tech staff or diluting the player pool further?

This has never been a trivial thing to do-- we've had people who believed it would be trivial, and then tried it and found out it wasn't, but I'm not aware of any instances of it actually being trivial to do. The trick comes in that it's certainly a fairly easy thing to export a bunch of areas and drop them into a new module (which is what people tend to be thinking about when claiming yonder assessment) and then you get to find all of the everything in those areas that was assuming they'd always be part of the server they were taken from, usually because it breaks.

Its relatively easy to do if you don't care about the module being a lot larger than it needs to be. i.e. you leave all the unused scripts, mobs, etc. in and just remove areas. Don't "export" anything as that only takes the dependencies the toolset can be bothered to find. The mod will be bloated and not much nicer to work with than the full version, but this isn't a problem from a player's standpoint because the extra stuff isn't downloaded to the game client like areas are.

If you do want to strip out unneeded scripts, mobs, etc., start with grep (and a lot of free time). All object-level module dependencies I can think of are stored as strings. On BG we use grep to find find stale resrefs inside objects or object blueprints (e.g. finding which areas spawn points live in, which stores items live in, which spawn points creatures live in, etc.). PITA though.

My test module is essentially BG stripped of everything except Beregost and it largely works fine. Most of our inter-area dependencies are static quests, which are typically few in number. Other servers might have other dependencies which might break things, which is why something like this would be a job for the server team, who are likely more familiar with their server's content than Tech would be.

Getting a host up and running is actually much harder than the above (MUCH harder than it was in NWN1, though it'd be easier in the case of as OAS as vault and database connections could be local). If a small OAS was desired, you could confine applicants to a portion of the server instead of giving them their own. I suppose this doesn't solve the "huge mod download" problem though.

Rivermoot would be pretty badass OAS. It has a large number of little quests, almost all of which are confined to its area, and a number of adventuring locals. Was it originally made in OAS-ship in mind?

Edit:

[yadda yadda] trivial to fork the module to put a low level area (e.g. Rivermoot) on it's own server.

[The right way's hard]

Yea, I was meaning more copy it and delete everything you /didn't/ want duplicated Unless TSM is a bear to run already, that should be good enough. If anyone *cough*yeahright*cough* has the time, it can be cleaned up later to kill unused scripts et al.

Yeah, this is pretty accurate

[Zelknolf]

Well, "provided that we're not worried about it being bigger than it needs to be" is a big "provided"-- TSM is a very large and inefficient server with a history of bugs resulting specifically from it being pruned and is characterized inflexibly-written systems. I'm not sure that I'd find BG to be a reasonable comparison-- its systems are more flexible, and its already a patchwork of several projects which don't seem to much care about the presence of the others.

[Ronan]

If someone wants a Rivermoot-OAS, they should try it and see what happens. My money is on most things working just fine. The nature of NWN discourages the use of arbitrary area-specific objects being used to store critical global state, which is where problems would arise. Don't delete those few areas which do store important global state though, such as the morgue or start area

[Regas]

I certainly don't object to having a stand alone OAS server. Up til now though we've not been able to muster the resources and or interest to launch and maintain one. Getting an OAS mod up and then maintaining it is a bit more than just editing down a live server. If we want to pursue this idea while we're running the OAS on MS and maybe work up to a stand alone OAS at some point then I'd say great. At a minimum we'll need commitments from folks to build, host and maintain an OAS. We'll need a buy in from Heegz, as that's not what he's just signed up for.

Also, Rivermoot is one of the most finished areas in all of alfa (who the heck made that area anyway- major props); and, I think we owe it to Curm to allow him to weigh in on if he really wants a copy of part of his mod posted as the OAS (I certainly wouldn't assume that we should expect him to be up for that)- If I was him I certainly wouldn't give the are up altogether.

The one big advantage to OAS on a live server is it creates a lot less overhead than maintaining a stand alone OAS.

[Ronan]

I was just trying to answer Fionn's question with my BG experiences, thats all. I've no idea what you guys want to do or have the manpower for. Edit: we could possibly run an OAS off BG's host, if anyone wanted to.

Fionn, as you may have gathered, its a lot harder to deploy a NWN2 ALFA server compared to our NWN1 modules.

[Regas]

I was just trying to answer Fionn's question with my BG experiences, thats all. I've no idea what you guys want to do or have the manpower for.

Fionn, as you may have gathered, its a lot harder to deploy a NWN2 ALFA server compared to our NWN1 modules.

Well, the OAS team would be glad to have you and Fionn on board if you deiced you want to build us something

[Ronan]

Edited the above. I would not mind trying an OAS out on BGs host if you guys would prefer that over the MS solution.

[Regas]

whachtya got in mind?

and only if it doesn't cut into Wednesday nights


(note I'd still want to run the oas on MS for the time being to see how that runs, but a stand alone OAS has advantages)

[Veilan]

thanks for your advice Maxcell.

Some points:
DMA does not control the admit process and never intended to. We are just providing the tool by which PA can bring back the OAS
Ideally anyone could be an OAS DM, however because OAS is overlaying MS, we have to abide by the don't play where you DM rule and the rule that only DMs may DM on a live server. That's why I'm accepting any DM onto MS who wants to do the OAS if the PA/Heegz wants them.
I do think that we should have more area for play than the original OAS, if only to attract players. We're not selecting out of a pool of applicants anymore, it's more like trying to attract existing good RPers who are also collegial people (the two do nto go hand in hand).

Thanks for the clarification, Castano.

With responsibility so clearly remaining with the PA's office, I guess using an existing server really is the best option - we have trouble achieving concentration as is, and I think both old and new ALFAns would have more fun being able to directly interact.

[oldgrayrogue]

If a relatively simple "dungeon" area and another simple "forest" area were added to the 'Waterdeep Docks" area of MS I'm guessing that could serve as your OAS. I still think that having a server like TSM be the OAS is a better choice though. As Castano notes, we have a limited pool of potential applicants, so every attempt should be made to impress them with as much content as possible, both DMd and static, when they decide to log on.

[CloudDancing]

Ugh. So why is everyone wasting so much time on this when what we have works great.

*It keeps our dms from getting spread out.
*It brings interesting iconic characters played by members into the potential members views
*It effectively markets Alfa for what it is
*There are ten times more chances for someone to log on and meet a player or a dm to play with
*It is a complete server with inspiring views and vantages
*The dms feel comfortable on MS and can use it effectively

All this energy should NOT be used to create some new server. It should be used to build a series of starting statics in Corwell and Lowhill which teach about Alfa's core principles.

I hate to see you guys wasting SO much energy when Zelk and Foam and the tech team spent SO much time converting MS. It is like you are spitting in the face of their work, the DMA's efforts to open Alfa to a wider audience, and to all the people involved with the OAS team currently and saying, "Sorry you are not good enough (for me personally.)

We are doing fine. Build some static content for MS. Please, i've been asking for two years. Dm on MS. Meet some new people. It works just fine.

[Swift]

Dozens of things in ALFA have been changed despite their predecessors working perfectly well. Should people just not talk about these things? What we have works, but there are ways it could be better which is the direction this discussion has turned.

There are also downsides about using a full, live server compared to a purpose built OAS. We discussed them at length back in the day and are discussing them again now. There is nothing wrong with that.

I would argue that NOT having those discussions about what can be improved and what the details are is a disservice, even if very little ends up changing.