Java Games Factory (Donation Drive)

General Discussions
#1

We need $500-$800 to secure a cheap deal for JGF’s future hosting. If you don’t know what JGF (the Java Games Factory) is, follow the link in my sig below; it’s a volunteer-run site promoting java games and supporting java games developers.

Over the last 4 months, the price of 1200Gb/month hosting has risen by around 40-60% (c.f. the cheapest package from servermatrix.com), and appears to keep rising every month or two. The nearest alternative packages that could support the Java Games Factory cost $100 per month for half as much bandwidth (the costs of known packages have gone up in the last 4 months, and the amount of b/w on each has gone down).

Although the next version of JGF is not yet ready, most of the current site can easily be converted across by the end of this month (although the new fancy features could take another month or two). The only chance we have to ensure realistic prices is to sign a contract now :(, and I don’t have the spare cash to do it myself. For a list of the immediate benefits to the community, c.f. my next post.

So. Inspired by the success of LWJGL, I’ve setup a paypal account and a “donate” link and I’m asking for donations to sign up a hosting account. Because the paypal account is brand new, there is no charge levied on the first $1500 of donations (which is way way above what we need), but … it cannot receive credit card payments. If people can only pay by credit card, I’ll upgrade the account, although that means we lose 3.4% of every donation to PayPal :(.

If we cannot raise money through donations, I have prepared a business plan for JGF that I might be able to get some investors for. That’s a lose/lose situation though: if they invest, I lose control, and JGF has to become a lot more profit-driven (no-one will invest if they don’t see evidence of healthy profits to come); if they don’t invest, I’ll have wasted several more months and delayed the next version of JGF by at least that much time.

If JGF ever generates a profit, after paying games developers royalties on their games, I’ll gradually use that money to repay people who’ve donated cash previously (unless you really don’t want it back). But you should definitely assume that isn’t going to happen!

#2

If we raise enough money, the following will happen immediately:

  • register a domain (!)
  • setup a dedicated server with servermatrix.com (1200Gb/month)
  • provide free hosting for ALL games that are on JGF (with 1200Gb, we certainly ought to be able to handle the traffic for all those games)
  • mirror the current work-in-progress of JGF version 3 to the server, on a different port (e.g. 8000) so that you can see it in-development

Over the following weeks:

  • complete enough of version3 so that it can run a complete site just like version 2
  • …and then copy all the data from version 2 to the miror of version 3, so that the new server / new version can start being actually used
  • add more of the version 3 features, like forums, user-accounts, admin status for some of the more active volunteers, etc

By christmas:

  • “deprecate” version 2, eventually replacing all pages on grexengine.com with “permanent redirect” to the new domain
  • (gradually the search engines will notice the migration)
  • add the rest of the core new features, according to feedback from developers, players, and admins.
#3

PS - the paypal “donate” link is at the top of the “About” page on the Java Games Factory site!

#4

Read this over and thought it might seem too negative. Apologies if this is the case, don’t not donate because of it:

I like JGF. Honest I do. And I’d have no qualm about donating some mooooola… I also know this is a long shot… but…

Did the Sun guys say that this was also something they were going to do!?

I’m all for a community managing and directing itself but it’d be really nice to see some sign of life from the guys over there at JGO Towers about supporting a gaming community?

More scary of course would be of course donating the cash, getting JGF up and running and then getting replaced by a new (probably less capable site) with a Sun badge on the front.

Kev

#5

Do you have an idea where your bandwidth is going to? 1200 GB a month is some extremely large amounts. Even all my sites with Xj3d.org, j3d.org and the company websites, all hosting high usage CVS servers do not even use a quarter of that amount. Are you doing implicit compression on the web pages as they’re being served or any of the downloads?

#6

Yes, and nothing happened.

Then it re-appeared in a “community board” minutes from 5 months ago, and I got in touch with the person responsible - Gregory Pierce. And nothing happened.

That’s not an indictment of the GTG - presumably it simply means they have more pressing issues.

I wouldn’t be doing this if I thought it might be a waste of time - there’s so much else I could be doing with my time, but I spend a lot of it on JGF because I’m convinced nothing else will fulfil JGF’s roles otherwise.

ChrisM et all have explicitly stated that they do not see this as part of their role; what you are describing is similar to what I had expected them to do originally, and you may remember the huge arguments that blew up when my expectations collided with their actual plans; it took a while for me to understand that I’d had the wrong idea all along about their intentions.

However, I think there is now a bigger issue. Now that JGF is well established (despite being a dull site with little to do we’ve now had over 3000 unique visitors [2] for 3 months in a row!), and proves that we can do this without having to bother Sun, there is a clear choice:

Do we want an independent JGF that is run by the community, or a clone that is run by Sun and has to kow-tow to corporate policy and politics?

Now that I’ve seen it’s viable, I massively prefer a site run by independent members of the community, with no requirement to obey Sun’s policies, working towards their visions of what they believe the community needs. JGF features and policy have been driven pretty much entirely by the actual issues faced by java games developers; no matter how hard ChrisM et al try to understand us, they are NOT out on the front-line actually doing this stuff.

Do we want a site in the hands of people like Cas (AlienFlux, Super Elvis), myself (Survivor, Crystals), and all the other volunteers who have given something to JGF precisely because they are working on games of their own and it helps them as well as others (people like: elias, swpalmer, kevglass, cfmdobbie, middy, weston, and all the others from these boards who’ve sent in stuff for the tech page, or written an article, or suggested games, etc)?

Or do we want a site owned by Sun, that has to adhere to all the onerous requirements imposed by their marketing, their corporate strategy, their branding, etc - yet is probably only a part-time activity for someone else whose main job is to actually make money for the company? No matter how good their intentions, unless McNealy dedicates a budget to “Games publishing manager”, they will be hampered by additional constraints yet simultaneously have limits on how much time they’re allowed to spend on it.

The only way JGF is going to get replaced is if the developers and game-players “vote with their feet” and stop visiting; as I mentioned to Cas last week at ECTS, JGF is a tool: it exists because we (mainly the developers) need something to be fulfilling it’s role (we need somewhere to promote our own games, we need somewhere to point disbelieving C++ programmers at to prove that java programmers write great games, we need categorized information on what API’s are available to us to program games in java, etc).

EDIT:
[1] that’s 3k unique users each month, 3 months in a row - with quite a lot of users coming more than one visit, so number of visits is much higher.

NB: there are things that a Sun site would be hard pressed to do. Individual developers couldn’t do this at all: for instance, we can get advertising contracts negotiated based on the total traffic for the domain and get excellent rates, which are then passed on to the developers whose games are on the site. If they tried it just for their own site, most developers would find they couldn’t generate enough traffic to even get a contract, and if they could it would probably be at very poor rates.

#7

[quote]Do you have an idea where your bandwidth is going to? 1200 GB a month is some extremely large amounts. Even all my sites with Xj3d.org, j3d.org and the company websites, all hosting high usage CVS servers do not even use a quarter of that amount. Are you doing implicit compression on the web pages as they’re being served or any of the downloads?
[/quote]
We’re currently chugging along at > 2.5Gb month just for static webpages and thumbnail images (i.e. easily cacheable - most ISP’s have large automatic caches for this kind of thing, in addition to any of the user’s local caches) . Off the top of my head, the b/w is around 30% to 50% HTML, the rest images.

NB: to be absolutely clear here, that is with NO downloads hosted on the JGF server.

When you look at the typical download size for a JGF game (approx. 1Mb to 3Mb) and the typical download size of 3rd party libraries (approx. 0.5Mb to 4Mb per game) then it’s clear that b/w will go up by a large amount when we host things locally. We do need to host locally because of the large amount of people who either don’t have good webspace or whose webspace has too low b/w limits on it. Also for all the poor people who’s webspace makes webstarting a pain :(. Hosting is an issue for something like 20% of games that are submitted - but to date we’ve been unable to help.

The new site is also going to generate a lot more traffic per person (the presence of forums etc massively increases the number of page views per person per time period compared to a fairly empty static site), and much less of it will be cacheable by ISP and local-machine caches. All of which adds up.

1200Gb is a good relatively future-proof figure: my predictions suggest that we could be soaking up 100Gb/month within 6-9 months without doing any download hosting, so e.g. a 500Gb/month figure would have me worried. I’m hoping that we can keep it significantly lower than that once the server is in my hands (and I can put on e.g. jardiff and any and all other compression I can think of…at the moment, I don’t have direct access to tweak the webserver).

#8

Seems like you are extremely highly overestmating the amount of data you’re going to be sending. For reference, j3d.org has been steady at between 11K and 15K unique visitors a month for the last couple of years. Xj3D.org has about 30 downloads a day of the largest installer (about 20MB a hit) and double that in spikes for the dev releases (typically once or twice a month). On top of that we have the CVS servers for xj3d and j3d.org, both of which I don’t have logs for but I know we have 40-50 users actively tracking CVS changes on a daily basis. For my sites, they’re hosted on a business class (1MB both ways) DSL line and it costs me a total of US$180 a month.

Another site run by a friend of mine (mcnews.com.au) is only generating about 600GB a month of traffic and he hosts one of the most popular motorcycle sites on the web, particularly for up to the second racing news and features - about 100K unique visitors a month. There’s about 3K forum members there too.

If you get anywhere near that 1200GB/month capacity, you’re going to be hosting one of the biggest websites on the net! I think you’re waaaay over what you really need to budget for right now - particularly if you go for the DIY approach with DSL etc.

#9

[quote]Seems like you are extremely highly overestmating the amount of data you’re going to be sending. For reference, j3d.org has been steady at between 11K and 15K unique visitors a month for the last couple of years. Xj3D.org has about 30
[/quote]
Different traffic patterns; bear in mind that the majority of our visitors come to download at least one game (and I believe more like 3). Where you have 30 downloads from 500 visitors a day, we’d expect more like 500-1500 downloads a day from the same number of visitors.

EDIT: xj3d appears to be in a niche that is much smaller than JGF’s niche (NB: JGF focusses on gamers as much as on developers, exposing it to a much larger audience). We’re getting 20%-30% of your hits after only a few months with no advertising (no press releases etc); once v3 is complete, and the site is credible in the wider world of gaming (has it’s own domain, for a start!), we will be doing advertising and active promotion (so far almost all promotion has been word-of-mouth).

…which is MORE than twice as much as the deal that gets us 1200Gb a month! If you think I’m proposing wastefully spending lots of money for b/w we don’t need then you’re way off the mark - I spent a long time (months) researching the alternatives; this is an incredibly good price and fits JGF’s traffic patterns extremely well, and on top of that it’s cheap compared to the best deals you can get elsewhere. I really don’t see any reason why it is NOT a good idea to secure this deal!

#10

[quote] but … it cannot receive credit card payments. If people can only pay by credit card,
[/quote]
If not credit card what? Personal check? Money order? Thats a huge pita. I think sacrificing the 3.4% would work itself out in convenience.

#11

[quote] We need $500-$800 to secure a cheap deal for JGF’s future hosting
[/quote]
Whats that in real ;D money? about £280-£450?

Sorry I may have missed it as I read the thread in a hurry but what is that for 6 months? or just to set it up?

also ditto on what options do we have to pay other than CC? (btw I’m quite prepared to help out, don’t know how much I’ll be able to spare though)

You mentioned forums, will they just be kept to playing the (complete?) games? I’m abit concerned that they will distract from this site otherwise. Maybe you could link to here as developer forums?

just my 2p
Dan.

edit: just went to the paypal site though the link and found it down for maintence :slight_smile:

#12

Size matters though :slight_smile: Those smaller numbers of visitors tie up the pipe for a lot longer.

The numbers I’m giving for reference are totals for a single server on that pipe that handles all three sites. Rough coverage would be fairly similar when you combine them all together (xj3d.org, in particular, has been directly linked from the front page of slashdot twice in the last 6 months).

[quote]We’re getting 20%-30% of your hits after only a few months with no advertising (no press releases etc);
[/quote]
At about the same time, each of our sites have been about 50% of their current rates. What you tend to find with these sites is that they ramp up very quickly, peak, and then drop back a bit (about 30% off the peak) then become stable. The numbers that you’ll have at the end of 6 months is what you’re likely to keep long term without a very large, active marketing budget.

[quote]…which is MORE than twice as much as the deal that gets us 1200Gb a month!
[/quote]
Ok. I visited the servermatrix.com front page and the best deal I saw there on the front page was $198 a month for a an average box that you had a reasonable level of control over (ie could install any software you wanted, separate root priveleges etc, not a shared, subleased box).

I’m giving you reference information to base your numbers from, from a set of sites that has similar coverage and style of audience, and has been established for a long period of time, as well as at least one site that is far bigger than what JGF is ever likely to become with a similar style of audience.

Separate to that, I do believe you are going to waste your money until you have actually justified your need for that capabilities. As you stated earlier in the thread - you’re barely using 1/50th of the allowed traffic right now (2.5 GB a month). These deals are around all the time, there’s no absolute imperitive to upgrade right now when the need has not yet been justified. JGF is a good site, I just don’t want you getting horrible burned on a rather rushed decision based on “deal of the month” situation, which this has a very distinct smell of.

#13

OK. If you read all of every thread on this forum, you’d see that I have been watching the server hosting market for the last 24 months (in fact, I’ve been keeping tabs on it for the last 8 years, ever since I ran a consultantcy where we had to source high-end hosting cost-effectively for intranets)…but there’s no reason why you would read them all :slight_smile: so I should probably have mentioned this explicitly.

According to my experience, and the many hours (probably a total of around 1 month solid) I’ve spent over the last 12 months looking at this market, these are my conclusions. Maybe I missed something major - perhaps there’s a trustworthy hosting company that offers TB a month for under $50 all the time - but until someone actually points that company out to me I’m assuming I’m pretty accurate.

Around 18 months ago, a series of cheap deals appeared in the dedicated-server market. For instance, £30 (approx $45 at the time) per month for a complete dedicated top-of-the-range AMD server. As it happens, one of Grex’s licensees went with just such a deal. There was nothing else that came close to the level of reliability and cost at the time. 6 months later, they started ramping up the price. The last time I checked, the exact same deal now costs you £50 a month - and this is for a b/w limit of 10Gb/month - and it’s still going up.

The same has happened with servermatrix.

These deals come along every six months or so, at different companies, and the “super cheap” part of the deal is random each time. There aren’t many companies who can safely (legally, etc) guarantee such deals on b/w, and the other similar deals seem to rest more on the simple fact that you have a dedicated server as their main USP.

Here we have an opportunity to grab the most important feature for JGF that is liable to be the biggest cost in the future - b/w - in one of these rare deals. Maybe the next deal by someone else in 6 months time will also have huge b/w; maybe not. Our processing requirements are minimal, ditto webspace - so e.g. deals offering 500Gb diskspace would be worthless.

These deals tend to come along only once at any particular company (or no more than once every few years). This is a completely different market to the “flavour of the month” deals which almost every company changes every 2-4 months (one month it will be “double RAM in the server” the next it will be “an extra 15% b/w for free”, etc) which are on a much lower scale and mostly aimed at the mass-market non-dedicated servers.

#14

Today it’s a little under £400; recent trends have been downwards for a long time - which either means it’ll soon bottom-out, or that it might get even cheaper ;). Either way, GBP are an excellent way to pay for US services right now…

This is for 6 months. There is a vast setup fee (30% of the cost for the first 6 months!) which means it’s not worth buying unless you know you can afford to pay for a minimum of something like 5 months; during the 6’th month, it becomes cheaper than all the other options at the same stage (which each have monthly costs well over $100).

Um. Good question. I can’t seem to find a list of what you can and can’t pay (paypal’s docs on their site are really crap - it’s 90% marketing BS on every single help topic >:().

I think I’ll just upgrade the account anyway and accept the 3.4% because otherwise it seems it’s too much hassle.

Personally, I want forums mainly so that each game has it’s own unique private (i.e. controlled by the game-authors) forums. We’re doing some funky CMS stuff so that it won’t be:

  • Go to a forums page which has hundreds of categories on, including one for each game

instead it will be more like:

  • On the download/details/screenshot page for each game there will be an embedded mini-forum just for that game.

  • …and we will experiment with embedding other forums elsewhere on the site. For instance, the mandatory topics for any site: “suggestions”, “bugs”. Also we’ll probably embed a mini-forum for each news item (as is common practice on CMSed sites these days).

  • and probably some semi-private forums e.g. for the various site-admins to discuss things quietly.

The intent is to complement JGO…Competing with these forums is the absolute last thing we would try to do! The point of JGF is to “fill the gaps”, not to “fight the incumbents”. Alexa guesstimates there are around 300k hits per month to jgo.org. Obviously, the number of unique users is a lot less than that, but I’m sure there’s at least an order of magnitude more visitors than JGF gets; it would be stupid to “compete” with jgo forums.

#15

This is absurd.

Sun is a hardware and networking company. They ought to sponsor the server and bandwidth. After all, the JGF is promoting their technology.

#16

[quote]This is absurd.
Sun is a hardware and networking company. They ought to sponsor the server and bandwidth. After all, the JGF is promoting their technology.
[/quote]
GP was interested in trying to get them to do this too (no luck, last I heard). Sun used to do this in the dotcom days with lots of cheap Solaris deals IIRC, but before and since they don’t seem to do it much. I know people doing corporate server dev who’ve got hold of tonnes of free kit from HP, Compaq, IBM, etc - but rarely from Sun. Maybe they just don’t know the right people?

Shrug. It’s not much money, and none of us here have a vested interest in promoting Sun’s server business (we only care about Java, right?), but we do have a vested interest in maintaining independence. It might get tricky if Sun could cut off the air supply at any point, so AFAICS if anything it’s their loss more than ours…

Unless, of course, we can’t raise enough money. In which case, I’d go begging to that part of Sun :). Or perhaps IBM might do it, seeing as they’ve been making strategic investments in server-side games dev for the last 4 years.

Either way, I rate the chances of a services + kit division donating/loaning for this as next to zero; especially given that we won’t be willing to use java.net or any of their other existing products.

PS: I did spend a couple of months trying to persuade some ISP’s to give us free server + bandwidth in return for promoting their services…I had a couple of leads, but in the end they each lost interest.

#17

Ok, so I see alot of bantering back and forth. Is this roughly the situation -

  • There is a pretty good deal right now at servermatrix
  • blah3 would like people to donate to get this deal
  • Not everyone agrees JGF needs the quality of services servermatrix is offering

What is the absolute worst that will happen if enough money isn’t donated? - No more JGF, reduced bandwidth to it?

If blah doesn’t get enough money, what will be done with it?

Does jgf need any money in general, ie outside this offer?

Dr. A>

#18

…and no-one has proposed anything cheaper. It’s not as if I’m saying “I could spend $5 but I want to spend $500”. It’s more like “all the dedicated server plans cost around the same amount, and are needed to run java, and spending it now will not only save money but also guarantee certain things like being able to host all games locally for free, long-term”.

To be honest, I’d not considered the worst-case until you asked.

We cannot stay on grexengine.com for much longer. JGF is now using the vast majority of grexengine.com’s bandwidth, and is dangerously close to getting Grex Games fined for going over-allowance. I suppose the absolute worst that could happen is that I can’t get any money from anywhere, the b/w peaks early one month, and I have to delete all the images from the site. I might even have to remove the whole site too, just to err on the safe side (404’s don’t sap much bandwidth). I’m certainly not in a position to sit around getting my employer fined each month, and paying the fines myself; JGF gets free space from Grex Games as a charitable act that I persuaded the other directors to do, not as a right.

So, actually, I have to move JGF somewhere else ASAP anyway, although it’s not yet “urgent”. I’d been assuming that if I couldn’t get enough donations, I’d know within a couple of weeks, and that would give me a couple of weeks to seek private investment.

With private investors, JGF will have to dump the low-end games (and, probably, some other stuff too) and be profit-motivated rather than community-motivated.

I’m not much use here financially; by the time I personally next get a wadge of money, it will probably be too late to make much difference.

If you mean “if 99% of the amount I was aiming for” or something like that, then… if we only get $200, I will simply login to paypal and hit “refund” on each item. $200 rents a server for 0.00001 seconds.

If we get $460 or more, I’ll sit down and think carefully about it. That might be enough to keep going until I had spare cash to pay for it myself - but I’d want to be sure before I committed to it.

Anything in-between, I don’t know. Either refund it all, or throw it open to public debate.

No (apart from $10 for a domain name ;)).

#19

Correction: due to employment law, JGF may have to pay everyone working on it if/when it goes commercial, and investors will probably insist on that just to cover their backs (otherwise there is a chance of volunteer works later filing lawsuits to claim large percentages of the company. It has happened, it will continue to happen). That means several extra side-effects if I have to raise the money from investors:

  • no volunteers allowed
  • less direct community involvement in decisions
  • higher running costs

If I understand the situation correctly, in practice we don’t need to worry about this law unless people are working for more than 25 hours per week, every week, for months on end… but investors tend to be skittish on such things, leaving no choice :(.

#20

Does anyone have any alternate hosting solutions to present? ???

blah3 - I will donate some money to your cause. For one, you’re one of the few who would have hired me in the java certification thread. :wink: Another thing is that your advice to myself and others is certainly worth some support.

I won’t promise how much. I get paid later this week, so I’ll drop you a line when I do it.

Regards,
Dr. A>