The author of this article should have written another article at the same time that was language agnostic.
Memcache as an accelerator for web apps is fantasticalicious, but it's not the best solution.
As a loooooong time memcache user, I've found that Redis is much more reliable when it comes to removing items on schedule, and has a bunch of cool new features as well.
I thought your first post in this thread was missing the point a bit (i.e., these platforms exist for social integration, so rolling your own doesn't really address the bigger goal of these systems. But this one? This is great stuff. Thanks for taking the time to do this.
My point was simply to spur the industry on to build an open source implementation that anyone can use and run, a la PhoneGap, Apache or PostgreSQL.
I wasn't missing the point. OpenFeint was built on the backs of indie developers, and they sold out and didn't protect the indie developer. Okay, so selling out isn't so bad, but why then make the developers do all the work to get compatibility when THEY were the ones that made you rich?
Just admit it's a bad implementation and commit to fix it.
Except with Twitter and Instagram, I need to authenticate to get the information about WHERE an action occurred, so at least there's a trail.
With your implementation, I can scan all my competitors users with public profiles (at the very least) and see who's buying what.
If this is a transparency play, then say that and make a stink about it. But when Mike gets the crap kicked out of him, also step up and cover his medical bills.
Edit: Okay.. so the location IS truncated, kudos. Why is this an unauthenticated API?
HeyZap didn't build the games that power them, the developers did. HeyZap, in releasing information such as installs, etc., to unauthenticated API calls is saying "This is HEYZAP's data".
It's not. It's the developers data too.
Without that developers game, you'd have NO data, and each player is a customer of the developer once they buy the game. They're not just HeyZap's users, they are paid customers of that developer.
Treat your developers businesses with respect and limit information to the developer and let them decide if they want to release it.
Hi, just to clarify, we don't have install data or anything from developers, only public information of what's on the app store, and what users share on Heyzap. It is the user's right to share this information if they would like to, in the same way that I can leave a review of an app without the developer's permission.
I'm sure the guys at HeyZap are good people with the best of intentions, but the issue is that they can't predict the future.
They can't tell you that their servers will be compromised, and all your users will be open game. They can't promise you that they won't bring in a CEO that just wants to get the company sold, and doesn't care about the developers or their customers.
They can't tell you that when they pivot, all your games will be busted. Not because they're hiding this gem from you.
It's simply because they don't know it.
When you roll your own, you know it's going to survive because YOU are responsible for it.
No BS politics are going to force you down an ridiculous migration path you don't already know about, and since YOU are the CEO, you can sell out and not worry about screwing companies who are dependent on your product or service.
Even IF you choose to use a third party, you owe it to yourself to write your own and mirror that data on to yours so that when, not IF, they go down or out, you can flip a switch and you're back in business within seconds, not days.
The deal with GREE stinks of potential currency manipulation, IMHO, and I wouldn't put my livelihood in their hands for one minute. Again, not because they're bad people - I don't know them from Adam, but because they've already shown that they're willing to allow the customer, we developers, to do ALL the work to satisfy THEIR migration to their API instead of requiring OpenFeint to develop a transparent proxy layer and make it turnkey.
Perfect example of why you should roll your own platform libraries, or at least try to use an open source engine that allows you to control the servers.
All those users just lost faith in yet another service, and all these guys are doing is saying "Hey, we failed you as a SAAS vendor, but we made $100 million, so lube up and try GREE because they need to [overcharge you voraciously to] recoup their investment.
You'd think the geniuses in the game industry could put their collective [self acclaimed] intellects together to come up with a platform agnostic approach to these common issues, and open source it for the betterment of the customer.
Oh, yeah... the... customer. Oops.
>>>>>>>>>>> UPDATE <<<<<<<<<<
OOH.. and it gets better... From GREE's agreement, paraphrased:
ARTICLE 5: Virtual Currency; Revenue Sharing
We've created our own currencies, not unlike BitCoin, called J-Coins and G-Coins. We reserve the right to change the ratios to capture more revenue, but right now you should just trust us that we're being honest.
Customers will pay US, not you. We'll tell you how much they spent of that money with you, and then pay you what we think is fair, in our sole discretion.
Oh yeah, to further confuse the customer, you can't call YOUR currency anything like ours, and we get to decide what you call it, and if you don't change it, we don't have to pay you.
I don't think rolling your own, or hosting your own is a feasible option for indie mobile and social game developers in the current climate short development cycles and small budgets.
The whole POINT of OpenFeint was basically: Hey, indie game developers! We know that building and running a game server is really hard and expensive. Why don't you let us run the servers for you, and you just use our (fairly simple and easy-to-integrate) SDK to make some calls to set high scores and unlock achievements?
Exactly, and they sold out and screwed the indies.
If you think OpenFeint had issues, just try getting your concerns addressed now that (a) GREE is in Japan, (b) time shifted 13 or so hours ahead of the US, and (c) works in YEN and uses at least two levels of currency conversion to compute your revenues.
1. Use a hosted service, and IF they end up shutting down either migrate to another similar service or roll your own.
2. Roll your own from the start.
I'm having a hard time seeing why option 2 is better. Just make sure the services you use give you a clean migration path.
A lot of game developers don't have experience building web services, so it makes sense to outsource the things that aren't their core competency. Do we really want every game developer rolling their own potentially unreliable, unscalable, insecure web services?
I came to the same conclusion for adding usage logging and promo codes to my app. Use Parse for now but try not to use too much of their API and to keep it to a couple of particular lasses that I access it through so that if I need to move it is just a day changing the app and one or two setting up and securing a server with a database.
Not a game but a lists app focussed on reusable lists for things like shopping and packing with a couple of neat features. Http://iTunes.com/apps/fastlists. Free with in-apps if you like it.
I haven't decided how to do shared lists yet, iCloud, Parse, custom server or something else.
Thank you Evan. Ontop of that this game was also mentioned in two http://www.buildwindows.com/ (build conference) presentations..... With the same URL. Please do more thorough research before making accusations. That being said, I do appreciate you trying to keep these forums clean however =).
The base domain is http://signalrshootr.cloudapp.net/. We simply registered a better domain name for it..... You can still reach it from the original domain. You'll see it's the same server.
He works at Microsoft, his real email address is in the registry info, and "login information" is all through oauth- he would have no access to your password or whatever.
You're making a pretty serious accusation with basically no evidence.
Thanks, yes, we're already working on that and have some domains reserved. So, the overall appearance is good? Would you join it?
Our team is preparing to launch a new trading platform offering small design service. Its goail is to operate as a medium between the buyer and the designer. If you are a student who wants to earn some extra money or a small business who can't afford to hire a professional this would be the place for you. What do you think?
Memcache as an accelerator for web apps is fantasticalicious, but it's not the best solution.
As a loooooong time memcache user, I've found that Redis is much more reliable when it comes to removing items on schedule, and has a bunch of cool new features as well.