Discussion and poll: Donations

[indio]

I'll worry as much as conscience demands. And given you haven't as yet hosted a NWN2 module, I'm going to take my concerns about this plan as legitimate.

For the record, modules larger than 3GB, not 2 as I originally posted (physical size, not memory footprint) are now possible.

http://www.alandfaraway.org/phpbbforum/ ... hp?t=34395

With all due respect zic, I'd worry more about your actual NWN2 hosting experience if you're going to go down this road.

You think I'm arguing scalability tgheory when I'm actually arguing empirical data.

ALFA should only consider accepting donations if it has a well-researched plan for exactly what problems need to be addressed, and exactly how they can be solved.

[zicada]

I'll worry as much as conscience demands. And given you haven't as yet hosted a NWN2 module, I'm going to take my concerns about this plan as legitimate.

For the record, modules larger than 3GB, not 2 as I originally posted (physical size, not memory footprint) are now possible.

http://www.alandfaraway.org/phpbbforum/ ... hp?t=34395

With all due respect zic, I'd worry more about your actual NWN2 hosting experience if you're going to go down this road.

You think I'm arguing scalability tgheory when I'm actually arguing empirical data.

ALFA should only consider accepting donations if it has a well-researched plan for exactly what problems need to be addressed, and exactly how they can be solved.

Indio, i have no idea how the physical size of modules we are able to run have any relevance to this discussion. How is that a valid argument against anything that ive posted in this thread ? How is it even a valid argument in terms of hosting nwn2 servers in general ?

Give me the empirical data you are arguing please!

I would also love to know where you have your info that i have never hosted a nwn2 module.

And please, i would appreciate it if you stopped acting like you are speaking for anyone but yourself.

It appears you think that I have allready made up my mind, and is just waiting for a fat check to run off and buy an awesome box. You see, before we get anywhere close to that, we have to come up with an organized way of doing the whole donations thing. THAT is what i need your input on. It requires alot of changes and alot of work. Once we have that working, THEN we can figure out more specifically what we need, where to get it, where to put it, how to run it, and so fourth. Are you feeling me ?

[NickD]

With all due respect Indio, i would wory less about the technical aspects of this, and leave it to the members who have the most practical IRL experience and technical theory under their belts.

Many situations at many workplaces:
Tech Guy: We should get this new tech, it'll be much better!
IT Manager: OK, if you say it'll be better, I trust you.

3 months later

IT Manager: This doesn't deliver what you said it would!
Tech Guy: Well, it's new tech, you have to expect some glitches to start with.
IT Manager:

Moral of the story: Don't trust the tech guy, he only wants new and shiny tech to play with!

[zicada]

Lets, for the sake of this discussion say that we do somehow put whatever money we get from donations to excellent use and solve every problem for now in a way that everyone would be happy with.

Lets focus on the process of making it possible instead as intended.

[indio]

I'm just guessing you haven't hosted a NWN2 mod, as originally stated. Have you? I figured after I first guessed you'd have clarified. When you didn't, I guess I figured I was right.

The cold hard truth, is that for NWN2, we are most likely going to have to facilitate and accept donations to a greater extent than we do now with NWN1.

Sounds to me like you're speaking for other people there zic. I only had to look at one post to find an example. Double standards are probably not a great way to run Infra. That's me speaking, not all the others I'm usually speaking for.

Physical size of mods aren't terribly relevant, but your knowledge of NWN2 modules is. Sorry if you don't think so, but I do. I'm trying to get an idea of what you actually know, because as it stands, NickD's anecdote is ringing true.

Setting up Paypal is child's play and requires no work. What other work is involved? Do you require to change our thinking to start volunteering?

Finally Zic, when you condescend about your technical wizardry, expect some curry in return. It may wash with some people, but it doesn't wash with me.

[darrenhfx]

What appears obvious to me, other than the animosity, is that the current state of poll result suggests that Zic's ideas have some popular support for further investiagtion.

Indio your posts don't do much for having anyone take your position seriously, even if they are correct.

[zicada]

I'm just guessing you haven't hosted a NWN2 mod, as originally stated. Have you? I figured after I first guessed you'd have clarified. When you didn't, I guess I figured I was right.

The cold hard truth, is that for NWN2, we are most likely going to have to facilitate and accept donations to a greater extent than we do now with NWN1.

Sounds to me like you're speaking for other people there zic. I only had to look at one post to find an example. Double standards are probably not a great way to run Infra. That's me speaking, not all the others I'm usually speaking for.

Physical size of mods aren't terribly relevant, but your knowledge of NWN2 modules is. Sorry if you don't think so, but I do. I'm trying to get an idea of what you actually know, because as it stands, NickD's anecdote is ringing true.

Setting up Paypal is child's play and requires no work. What other work is involved? Do you require to change our thinking to start volunteering?

Finally Zic, when you condescend about your technical wizardry, expect some curry in return. It may wash with some people, but it doesn't wash with me.

All right, i dont honestly think i have to even reply to this, Indio. Your post really speaks for itself. Please refrain from posting in this thread again unless you have something of value to contribute.

[indio]

So that's a definite no on the hosting a NWN2 module.

[ç i p h é r]

Let's take a step back for a second. To begin with, NWN2 is not a multi-threaded game, which means it cannot benefit from multi core processors by splitting up the processing load between them. So, a single core in a quadcore setup should handle a single instance of NWN2server just as a single core in a dual core setup would. Provided there are no internal data bottlenecks, it seems to me we should at a minimum be able to run 3 modules on a quadcore server simultaneously, leaving the 4th core for everything else, including background processes and services like SQL.

Zic, do you have a cost comparison between dual core and quadcore servers? If running 3 modules on 1 quadcore COSTS LESS THAN running 3 modules on 3 dual cores, then we're still ahead of the game w/o factoring in any of the other benefits, which from a technical point of view are quite significant. And unless a single NWN2instance utilizes 100% of CPU processing power, we should still have some headroom left over to scale with, just how much likely depending on the amount of server activity we experience.

To be perfectly honest, this is the sort of technical configuration a project like ALFA should be examining. Coincidentally, we happen to have 3 modules now in beta with a 4th looming on the horizon, which means we have a good opportunity here to test things before LIVE, if the economics of a quadcore setup are not unfavorable.

[zicada]

So that's a definite no on the hosting a NWN2 module.

Not that i understand how its important, since its very much the exact same thing as hosting any other game from a technical point of view.

But yeah i have hosted a few modules for testing. On various machines. Better yet, i understand the problems associated with nwn2 and hosting through reading discussions and posts by the developers. Again i am not sure how relevant my personal experience hosting nwn2 modules is to this discussion.

Feel free to start another thread on that subject, or to answer any other questions you might have. Leave this thread alone though please.

[zicada]

Let's take a step back for a second. To begin with, NWN2 is not a multi-threaded game, which means it cannot benefit from multi core processors by splitting up the processing load between them. So, a single core in a quadcore setup should handle a single instance of NWN2server just as a single core in a dual core setup would. Provided there are no internal data bottlenecks, it seems to me we should at a minimum be able to run 3 modules on a quadcore server simultaneously, leaving the 4th core for everything else, including background processes and services like SQL.

Zic, do you have a cost comparison between dual core and quadcore servers? If running 3 modules on 1 quadcore COSTS LESS THAN running 3 modules on 3 dual cores, then we're still ahead of the game w/o factoring in any of the other benefits, which from a technical point of view are quite significant. And unless a single NWN2instance utilizes 100% of CPU processing power, we should still have some headroom left over to scale with, just how much likely depending on the amount of server activity we experience.

To be perfectly honest, this is the sort of technical configuration a project like ALFA should be examining. Coincidentally, we happen to have 3 modules now in beta with a 4th looming on the horizon, which means we have a good opportunity here to test things before LIVE, if the economics of a quadcore setup are not unfavorable.

Since the machines in question usually come with 2 cpu sockets and 8 or 16 memory sockets, we can start small and upgrade when needed.
Getting a machine with a single quadcore processor and 8GB ram, 2x 73gb scsi(or sata) raid1, would be about $5000. A Dual Quadcore with 16GB ram would be about $7000.

This would give us the best bang for the buck + upgrade posibility.

If we decide to with with 2 or 3 core2duo machines with 4 or 5gb ram each, we will have to spend 2 or 3 times as much in hosting each month.

What i've been thinking all along, is adding targets to the donations drive. Where we list a typical target machine and sum.

[MShady]

Let's put things into perspective...

Zicada hosted the entire ALFA VAULT for years. He is also a highly trained Information Technology professional, who has been very generous and we he helped keep the community together by hosting the vault when no one else could/would.

How anyone can turn around and say he is unqualified to host, or organize hosting of NWN2 modules, when we are not even live, is ridiculous. Or turn it into an ethics issue. This is probably the most innovative idea in a long time and merits serious discussion, particularly amongst DMs, and maybe it should be moved there at this point.

As someone else just said in a post decrying Hialmar leaving, the previous Infra Admin, this kind of drama is not helpful. Lets just discuss this on the merits...

Mike

[indio]

Sounds like it's ok to talk about technical stuff now.

So, a dual quad, 8 cores, is $7000.

My question is this....how will we ever get a chance to see if such a computer can host more than 5 servers without buying one? 5 players on each, database running, etc etc...

See, theoretically it can. 8 cores, single threaded nwn2server.exe app, lots of ram, no problem. But that's precisely where theory and I diverge. How will we ever be able to justify spending $7000 on one computer when it may only do half of that well without testing it first?

Can anyone here access a dual quadcore? If it can only do half of what we want in practice, and you'll note right there zic I'm in favour of the idea, we've got ourself a $7000 computer doing what a $1000 computer could.

[NickD]

I have a question. What are your plans for the money? Are you going to wait until you get $7,000 to get the most value for money, or do you think you'd run the 3 month drive and buy whatever you can with the money then upgrade as more money comes in, potentially wasting money on upgrades, but getting immediate results?

I'm not favouring either way. I'm just wondering what your thoughts on the matter are.

[zicada]

Well, to be quite honest, all of those questions are what this thread is meant to discuss

It isnt meant as a thread where i say "THIS IS WHAT I WANT!" and everyone getting their chance to saying "OH NO, THAT SUCKS!" That isnt a discussion at all is it ?

I've allready contacted the developers to get their point of view in regards to scalability ofcourse. All i have said so far is that theoretically it is possible. And if it is, then it makes a hell of a lot of sense because it solves so many underlying issues,- many of wich the entire base of alfa members neither have to, or should have to deal with understanding.

All we really want to do, is ask everyone who is a member of alfa for a one time donation of around $25 on average to make things work much better for everyone. And this thread, is about figuring out how to do that in the best possible way.

Ive spent 3 pages now trying to get that message across in various ways. The next post i see in this thread better include ideas or valid points in regards to organizing donations, documenting, figuring out who should own/operate/administer to make this transition as smooth as possible. I am not educated enough in matters like that personally, so i can only talk from logic, and not experience.

If nobody is able to post anything that can help answer those questions, or step up to write something outlining guidelines in regards to this, i am closing this thread and moving it elsewhere, since i allready know where the majority stands in terms of donations as an idea.

Thanks for contributing.