Well, the main takeaways that I had from the paper:

1) Piracy is always going to happen.  Some people just prefer to pirate.
2) DRM will never stop piracy. 
3) Even if your DRM is perfect (in terms of annoyance to the paying consumer), it still comes with a cost


The interesting thing that the paper only pays lip service to (and I hope they get into in further papers) is the network externalities created by piracy.  For those who are uninformed an example of a network externality in action can be seen in the old VHS vs. BetaMax (or for the younger crowd, BluRay vs HDDVD).  VHS was a worse format than BetaMax in most categories (technologically), but VHS had better adoption rates.  Once VHS got a little bit ahead, it made sense for film makers to release on VHS as VHS had more users.  This meant that users were more likely to buy VHS players as VHS had more movies available. ad infinitum

What I'm getting at is: I want to see the cold, hard analysis on how piracy helps games. e.g. spreads the word, has people try before they buy, etc.   

That is the problem with investing in China, building for iOS, or submitting to Steam. The rules are opaque and subject to change at any time. The gamekeepers can make decisions without justifying them to anyone.

That is one of the greatest things I've ever read.  It reminds me very much of "You fool! You fell victim to one of the classic blunders - The most famous of which is 'never get involved in a land war in Asia'"

While I don't share MattG's vitriol, I do agree with the sentiment.  I mean, a form letter that says "Sorry, we don't think you are a good fit for us," would take all of 5 seconds to send out, and would be worlds better than *silence*.

Yeah, I would say that no one should add Matt G until such a time as the other links show up on his site.

I'd also urge looking at the disclaimer on the site:

The fonts presented on this website are their authors' property, and are either freeware, shareware, demo versions or public domain. The licence mentioned above the download button is just an indication. Please look at the readme-files in the zips or check the indicated author's website for details, and contact him/her if in doubt. If no author/licence is indicated that's because we don't have information, that doesn't mean it's free.

In the future, I'd suggest trying to contact the author (it feels wrong to say that the person who creates a font is its author) in cases like this.  If that fails and there is no information about the license, then I don't really know what you should do, to be honest.  The safe option would be to only use a font for which you are sure of the provenance.

MSVC 2010 is not supported afaik. Panda should compile on it, but they provide the 2008 version of precompiled third-party libraries, so you have to recompile those (since the Microsoft compiler is not compatible between various versions). The easy way is to use 2008.

Tell me. If everything can be done more easly in python, why games are being made in C++.

Indeed, Panda's core is C++. He was suggesting using Python for the "gameplay" layer. Typically there is not a perceivable difference if you process your gameplay with Python or C++, because these are operations which you execute few times per frame. Obviously, when you have operations which are executed thousands of times per frame, you can not use Python, but this is not so common in the gameplay layer. Anyway, in this scenario, you could write only this specific code in a C++ library wrapped in Python, and access it from the remaining code, written in Python.

Thanks for stating better what I meant to say.

I would say that people still use C++ because of inertia.  In my experience game programming falls into an 80/20 type situation where 80% of the code takes up 20% of the run time.  The remaining 20% of code is typically the heavy duty lifting code, i.e. the graphics (and occasionally AI or physics).  My philosophy is that the small amount of heavy duty code should be written in an appropriately speedy language (usually C++) and the large amount of other code should be written in an appropriately easy language (hence Python). This is a relatively new idea in the world of software engineering (i.e. less than 20 years old), and people are resistant to change, hence why games are still made in C++.

Personally, I would recommend learning Python and trying Panda3d in Python, if for no other reason than to learn a new language.  You might find you like it.  If you don't, then you can curse my name for all eternity (it is Adam).

Read here.  I would recommend actually reading the paper for those interested in economic game theory.

While they don't say that DRM is always bad, they do definitively prove that the removal of DRM doesn't necessarily increase the amount of piracy.

Thoughts?

This is completely unrelated to your .dll woes, but is there a reason you are using the C++ version of Panda3d instead of the Python version?  Personally, the joys of programming in Python make it an easy decision for me, and I'm unsure as to why one would use the C++ version.

personally for radian values of π/2,π and 2π I just use memorized versions to a few digits (eg 1.5075, 3.1415)

ew, why wouldnt you do PI/2? thats less typing, no memorization required....

(if you don't have PI defined somewhere, well go do it man)

Yeah, I mean, 3.1415?  The rounding is even bad.  Why not 3.1416? 
Personally, I use boost's constants when in C++ world.  The fact that they are accurate out to 100 decimal (in most cases) is a bit overkill, but hey why not.

e-mail sent

Interesting!

And of course, being able to draw attention probably has a lot to do with the success of temporary pay what you want sales, etc.

Another question - I've seen some people (The Gratuitous Space Battles guy, for instance) have had some success with "Pick a Price" sales, where they offer several discrete suggested prices and you can pick any of them.

Got any opinion on those? Although superficially similar, I wonder if there's any psychological difference in how customers respond to it. I guess from a technical perspective it might be easier than pay-what-you-want as you could just create multiple "products" with your e-payment provider of the same game at different prices.

Well, as Proun's data shows, people essentially discretize the prices themselves.  People like to have nice, round numbers, and I would guess that the salient price points for consumers of indie games are probably:
0 --  1 -- 2 -- 5 -- 10 -- 15 -- 20 -- 30
And the 30 is probably stretching it.

Responding to Joost (and thanks for contacting him Klaim!)

I didn't realize that the torrent went through his site.  I assumed that when he broke out paying 0 and torrenting (effectively the same thing for the consumer, although different bandwidth usage, etc.) that he meant that paying 0 was from his site and torrenting was elsewhere, my bad.

I don't really understand this portion "Even if only 1% of those would be converted to paying if the game were not available for free, that would still be a 50% increase in the amount of people who paid for the game!" 
It seems there are a couple of assumptions rolled up in that they I think are specious at best. Does he know that he didn't convert 1% of those people already? I.e. is each person only counted once (with the highest amount they paid) or are all downloads counted. I'm guessing it is the latter.  Given that most of the comments that I have seen (over at indiegames.com and MindEraser's comment) seem to indicate that people treated the free download as a demo.  Does he have any statistics on the number of people who downloaded it for free, and then kicked a couple of bucks back?  Or the number of people who dove straight in and just gave him money. 
If all of his sales were converts who downloaded it for free then he got a conversion rate of about 1.8%.
If we assume that all of the people who paid above $2 were people who dove straight in, and all people who paid $2 or less were people who played, liked what they saw, wanted another level, and paid in, then the conversion rate is ~0.8%. 
Given the limited data with no cross-correlation information, I'm going to guess that he got somewhere on the order of a 1% conversion. 

tl;dr  I think it is incorrect to assume that all of the people who paid would have paid had there been no "demo".

Maybe his data actually says otherwise,  but from my limited armchair point of view, I'd say that he did pretty well with the pay what you will. 

If I were in his place here is what I would have done (knowing what we know now):
*Have a demo (Which he essentially did, i.e. a lesser free version)
*Have a greater differentiation between the demo and the paid version (What is the incentive to pay, except for the heartwarming feeling of supporting the developer?  One level doesn't seem enough)
*Release the game through Steam (If it isn't that hard for him, why wouldn't he do it?  It seems like a bad business decision to not go through the number one digital distribution system)
*Go for a fixed $10 price point.  At first.  The market seems to indicate that $10 is about what people will pay for this type of game (i.e. an abstract, surreal racer [I'm going to say that AaaaaAAaaaAAAaaAAAAaAAAAA!!! is the closest analogue for Proun]). When sales start to flag, then go for the limited time pay what you want model.  I think a major factor in the success of the HIB's is their seeming scarcity.  They only last for a short time, they only pop up a few times a year.  It makes people think, "ooh I have to get in on this now"

Well, I think there are a whole bunch of faulty assumptions in his post.  He thinks he would have made 5 to 10 times as much with a fixed pricing model?  Where is the data that supports that?  Here's my quick arm-chair analysis:

(I'm going to round up to the nearest 100 from his chart, to be generous)
He had ~1800 people pay $2, ~1800 pay $5, and ~600 pay $10. Assuming that all of the people who paid would pay any reasonable price, he has 4200 paying customers. I'm going to assume that $10 is the maximum reasonable price (given that it is the highest price that still had any significant number of sales) then he would have made $42,000 instead of the $20,000 that he made.  And this is in magical christmas land where increasing the price has no effect on the number of customers. 

If he's thinking that he would have converted a significant number of the people who paid nothing, then he's an idiot.  Even if he hadn't had his game up for free, four times as many people torrented it than downloaded it for free from his site.  The most logical conclusion in my mind is that the people who downloaded it for free would have just torrented it, had he not had it up on his site. 

I have the opposite take away that he had.  I think the game was a mild success, and that the pay what you want was a moderate success.  The standard word online is that the game didn't have anywhere near enough content, and people who downloaded it for free and played the 3 levels didn't really care to pay to get the extra 4th level. 

I think you have the right idea CowBoyDan.  If he had an always free version (i.e. demo) with the 3 levels, and a pay version with at least 10 levels (if not more, I think 20 would be the ideal start), I think he would have been much more successful.
I agree that removing the free version is probably a bad move.  I think he's coming at it from the wrong angle.  Instead of thinking, "I need to get money for this version" he should be thinking, "How can I add value to this free version, so that people will want to pay for it."

Umm...why don't you just use FlxButton?  It handles the collision checks and what not.  All you need to do is supply a function that it calls when the button is pressed.

You don't understand Flash, do you?  You can certainly run a Flash game without internet access.  Anyone with the Flash player can run a .swf, and you can always make a Flash Projector from a .swf (which bundles that version of the Flash player with the .swf).

As for the smartphone thing, ok, I'll give you that.

As per AIR being stand alone, it seems that AIR 3 will solve that.  The captive run-time option bundles your app with AIR, so users won't have to download AIR.  Seems to be the AIR equivalent of making a Flash Projector.  Now, I'm not familiar enough with AIR to know why you would want to make an AIR application instead of just a Flash application, but if you need the things that AIR brings to the table (I guess better file i/o options and the potential to use native API's might be worth it to some people) then it would seem that AIR will be a decent option.

It's really just a matter of discipline and resolve. I keep at it because I have faith in what I'm doing, and the eventual creations that will result from it. And above all, I love the process itself. The creative aspects of what I'm accomplishing are deeply satisfying to me on a personal level. Even if I never produce a hit game, I will never look back on this as time that was wasted.

This, this, a thousand times this.  I mean, sure I want to succeed and leave my job to do this full time, I imagine most of us do.  But even if I'm never able to do this, I have a lot more fun and am a lot more satisfied with my life making video games as opposed to not making games.

Why not?

Circle Cat Games
http://www.circlecatgames.com/

You've both been added.

I tested both and wound up using Flixel.  As Hangedman said flixel is more of an all inclusive package than Flashpunk is.  This has the same benefits and drawbacks that these things always have, i.e. it makes doing what it wants to do really easy, and makes doing what it doesn't want to do very hard.
I'd suggest taking Flixel for a spin and seeing how happy you are with it.  If it does ~95% of what you want it to do, then go for it.  Otherwise, trying to mod Flixel to do what you want it to do is probably more hassle than it's worth and I would go with Flashpunk. 

Also, search the forums before starting the n-th thread on a topic.

Not sure if this is the place for this, particularly since it is more of a digital collage than anything else, but here is my love letter to the MBTA (The public transportation system in the Boston metro area) and Super Mario World. 

Well, the only relevant taxes (in the U.S.) are sales taxes and income taxes.

Sales tax -- Only necessary if you have nexus with the person you are selling to, i.e. if you are in the same state as the person you are selling to.
Income tax --Keep track of all of your sales, as you should report this income to the IRS.

I'm no accountant, but that's pretty much the extent of it (to my limited knowledge).

I second what increpare said.  I tend to think of gcc as the gold standard of the compiler world, but I guess you could go with vc++.  I think most people tend to vote along the lines of whether they came from a linux background or a windows background.