Discussion about Game Server Costs

[theG57]

Hello Fer, my 2 cents. From running servers from my own house to paying for a server from ofp to arma1. So I understand what you mean and feel in the bank account. I think this should go to pay to play. Yes those bad words. Keep it cheap ($1usd per person per week) first timers give them a freebie. So they can see what it's like and get hooked :-0 Then pay for it by week or month. Change the password every week. This should help your wallet out and I don't think most would complain about $1 a week...

G out.

[LiddleFeesh]

Hi Fer,

Thank you for hosting the folk sessions for us. I haven't experienced any lag or network issues with the server and thus far the network/server experience has been pretty good, even when filled to near capacity with all those MARSOC folk, too.

However I did get a little 'sticker shock' looking at the jestservers price tag for the ArmA server. Especially for a server that is only in use for a couple of hours at the weekend. It did give me a couple of ideas for how you could make the costs a little more bearable and help keep Folk sustainable. Two suggestions that are not mutually exclusive:

• The current server costs $59 per month which means that you have to find $14.75 per session in order to fund it. This can be achieved by changing the ethos of Folk to a "pay to play" dynamic in which to play a session it costs a dollar, paid in advance and those who have paid receive the server password by return mail. Indeed, all that would need to happen is that the server gets passworded after each session ends and whenever a player donates $1 they receive the password by PayPal receipt. Even if seven players come, it would cover half your costs which is a much better place to be than not. New guests can join for a free session if they wish. So it's 50 cents per hour for quality, organised game-play.
• Change server. Since the server doesn't operate like a MARSOC or TZW server (i.e. in daily use) and is just up for short bursts, the typical 'game server' model doesn't seem the best way to pay for it. I would recommend, if at all possible, looking to host the ArmA II server on an Amazon EC2 (or indeed Microsoft Azure) instance and just pay for the hours that we need it. That would give us a more powerful server for less money, it just needs switching on at the start of the session and off again at the end. Pros: Much cheaper, at least as powerful and with a capable network as current server if not more so. Cons: Not always left running (would be more expensive if left running).

And yes, I had been talking about this with G over Ventrillo shortly before posting this message (as I see he's just posted one also!).

Mike.

[fer]

Thanks for the suggestions, G and Feesh. Since Need to Know threads are more about reference, I thought I'd split your posts out into a new thread here in general discussion (hope you won't mind). Couple of thoughts in response:

• Just to put this into context, since I made my last post to about server, a few people have donated enough to cover our costs for December and most of January, so we're in pretty good health for the moment. Thank you to everyone who donated!

• There isn't a lot of precedent for blanket pay-to-play in the ArmAverse: groups like MARSOC, Shack Tactical, Tactical Gamer and United Operations - all of which offer bigger sessions and extra features like Yoma (for addons) - are all free to play, or voluntary funding only. I think we'd find it difficult to explain why folk's proposition is so much better than those groups as to justify blanket pay-to-play. Tigershark isn't that handsome.

• It's also worth remembering that there are different ways to contribute: making missions, helping to organise joint sessions, or just turning up to every session and helping the playercount stay healthy - TBH, many of those are more important than money; I'd much rather that people came to sessions, and anything that might discourage that would concern me.

• Curious to know more about the idea of using Amazon or Azure - although having our server available mid-week is very useful for testing missions (because developing MP content locally often fails to identify bugs associated with dedicated servers). Something which I discuss with other contacts in the community pretty regularly is the idea of server sharing - and we already save significant costs by using MARSOC's TS3 server. Alternatively, perhaps we could start making use of the server more often than just once a week?

[Tigershark]

Whilst I fully appreciate the experience G brings to this and the sound arguments from Feesh. I am having trouble with a pay to play model. Even though I know it makes sense.

[LiddleFeesh]

• There isn't a lot of precedent for blanket pay-to-play in the ArmAverse: groups like MARSOC, Shack Tactical, Tactical Gamer and United Operations - all of which offer bigger sessions and extra features like Yoma (for addons) - are all free to play, or voluntary funding only. I think we'd find it difficult to explain why folk's proposition is so much better than those groups as to justify blanket pay-to-play. Tigershark isn't that handsome.

Are bigger sessions better? Whilst I enjoyed playing with MARSOC, it's probably unlikely that I would join them for a game. The same goes for Shack Tactical, Tactical Gamer and United Operations and any other squad that requires either: a) practice, b) structured attendance, c) ranks, d) calling people 'sir', e) to dress in combat gear prior to play, f) takes itself too seriously, g) uses the phrases 'fireteam', 'oscar mike', 'tangos' too often.

Fun? Perhaps, in a scientific curiosity kind of way. It was hugely fun to join with MARSOC in the combined sessions, but I probably wouldn't join them as a squad member. If anything, I enjoyed playing >against< them as an OPFOR (there I go with the lingo) rather than with them. And if we can get some missions made up that take us fish out of the proverbial barrel, I'd like to see how well our loosely organised teams fare against a structured squad. I'm digressing...

I wouldn't say these teams offer more to the average non-military person who wants to read books to their children and then, shoot people in the face while having fun with old friends and new ones. I like the premise of Folk a lot and the non-committal aspect is a crucial part of it.

• It's also worth remembering that there are different ways to contribute: making missions, helping to organise joint sessions, or just turning up to every session and helping the playercount stay healthy - TBH, many of those are more important than money; I'd much rather that people came to sessions, and anything that might discourage that would concern me.

The sessions are great and it's in my calendar each week

[

• Curious to know more about the idea of using Amazon or Azure - although having our server available mid-week is very useful for testing missions (because developing MP content locally often fails to identify bugs associated with dedicated servers). Something which I discuss with other contacts in the community pretty regularly is the idea of server sharing - and we already save significant costs by using MARSOC's TS3 server. Alternatively, perhaps we could start making use of the server more often than just once a week?

If you hosted using a pay-per-use server it will be available whenever you wanted it, including mid week. You just have to turn it on first and remember to switch it off when you're done.

And server-sharing (or sub-letting) seems a good idea. I was recently playing on Amarak's (a fellow TZW member) server and it was lag free (although being US based my ping was higher than to the FOLK server). His server is mostly empty (I think). It might be worth considering approaching Amarak and offering him a donation for the use of his server on Sunday evening (and the odd MP test mid-week).

And just throwing it out there, but what are the bandwidth requirements for hosting an ArmA server? I've heard 5-7 players per 1mbit server upload being touted as the recommended limit. That would mean even my 1.5mb upload could serve up a session for an average FOLK get together without the need for a paid server.

[fer]

Are bigger sessions better? Whilst I enjoyed playing with MARSOC, it's probably unlikely that I would join them for a game. The same goes for Shack Tactical, Tactical Gamer and United Operations and any other squad that requires either: a) practice, b) structured attendance, c) ranks, d) calling people 'sir', e) to dress in combat gear prior to play, f) takes itself too seriously, g) uses the phrases 'fireteam', 'oscar mike', 'tangos' too often.

In fairness to those particular communities, they are all of them far more relaxed than some of the less well-known, but more hardcore milsim groups. I'd actually encourage you to check one or two of them out if you can spare the time. Bigger is not by itself better, but the opportunities open to sessions with high numbers of good quality players are really significant. At the most basic level, knowing you have 60 decent people who will respect the spirit of an adversarial mission's design (rather than seek exploits) is worth literally dozens of mission-maker hours. And, ultimately, for any community seeking to create complex experiences (rather than play Domination), that is quite the prize. I'm not saying you have to be big to produce good experiences at all - and would hope Folk is proof of that - but neither would I want to suggest that being bijou guarantees quality.

Back to topic, though, I agree with you that we should think about server sharing (btw, Amarak is a guest here already, and I hope he'll have time to play in a session sometime). That said, something that I really appreciate about our current set-up is that all the 'heavy-lifting' server admin tasks (like downloading patches) are done for us, or one-click operations. When we switched over to Combined Operations this week, it really was as simple as 3 lines of Skype chat with Jest. That was really cool!

[LiddleFeesh]

All true, Fer...