<p>Recently I came across the fascinating story of "Haunts: The Manse Macabre" which is a Go written game which got successfully <a href="https://www.kickstarter.com/projects/2066438441/haunts-the-manse-macabre/" rel="nofollow">crowdfunded</a>. After some time of development, funding run out and the pretty complete(?) game got abandoned (and open sourced), citing the <a href="http://www.escapistmagazine.com/news/view/120314-Troubled-Haunts-Kickstarter-Goes-Open-Source" rel="nofollow">technical difficulties</a> especially with Go and its tooling. Later it became even more bizarre when even the original author claimed he <a href="https://www.kickstarter.com/projects/2066438441/haunts-the-manse-macabre/posts/420311" rel="nofollow">can't bring the code base to compile</a></p>
<p><a href="https://github.com/losinggeneration/haunts" rel="nofollow">Some enthusiast seems to work on the code base off and on for some years now but also with no successful build it seems... github.com/losinggeneration/haunts</a></p>
<p>So, while I'm not a Go programmer (but a developer) this is quite fascinating... what are the reasons for these excessive troubles? Someone can share insight here or has followed this project? </p>
<hr/>**评论：**<br/><br/>rauyran: <pre><p>I believe that the bulk of Haunts was written before Go 1.0. Also the OpenGL bindings at that time were notoriously brittle with dozens of half-supported packages. I assume those are the reasons why they could not compile the code.</p>
<p>I wrote most of a game in Go at around that time and it doesn't compile today because sorting out the obsolete OpenGL bindings is just too much hassle (<a href="https://github.com/iand/amberfell" rel="nofollow">https://github.com/iand/amberfell</a>). </p>
<p>Today the answer is simple, just use go-gl: <a href="https://github.com/go-gl" rel="nofollow">https://github.com/go-gl</a> </p>
<p>I'm using it for my own game engine and it's mature and well maintained.</p></pre>GoTheFuckToBed: <pre><p>A game engine in GO? Please tell us more.</p></pre>rauyran: <pre><p>Not much to talk about really. Right now it's great if you want to build a game consisting of rotating cubes. It's a vehicle for me to learn techniques like forward rendering and parallelization of game engine logic. It's not public but maybe it will be one day when I'm happy to expose it. </p></pre>neoasterisk: <pre><p>I know nothing about this project so what I am going to write here is merely a speculation.</p>
<p>First of all, this project was started almost 5 years ago. Personally I have trouble keeping my excitement about a project for more than 2 year. I am not surprised that the original developers left the project after 5 years, (though that is not an excuse for not delivering to their promises). </p>
<p>Also the choice of the language in combination with the time might have played some minor role. While the Go leaders take backwards compatibility very seriously, 5 years ago the landscape and tooling for developing games in Go was almost non existent. We didn't even have this new amazing low pause GC we have available nowadays. Add on top of these the higher chances of the developers being unfamiliar with the language or just inexperience with developing games in general and you have a recipe for destruction.</p>
<p>That said even if the game wasn't written in Go I think that the ratio of goals and funds was way too ambitious. To me it seems like a typical kickstarter case of overestimation and overreaching. Someone might think that $25k should be enough to develop an indie game but in reality it's a very low amount when support for 4 platforms (PC/Mac/Linux and Ipad) has been promised. In an ideal world where everything is written in Go, maybe building for different platforms would only require a simple cross compilation but since we are essentially dealing with C/C++ here, porting the game on different platforms requires tremendous effort and extra funds. </p>
<p>To understand what an impact extra promises and overestimation can cause to a kickstarter, one of the most popular kickstarter failures was <a href="https://www.kickstarter.com/projects/mightyno9/mighty-no-9" rel="nofollow">Mighty no 9</a> which despite reaching 4 whooping million dollars, till this day it still hasn't delivered the ports to the rest of the platforms it has promised, not to mention that the game was bad despite being developed by proven industry professionals.</p>
<p>There are hundreds of failed kickstarters like this. The fact that Go was chosen as a language, is pretty much irrelevant if you account the rest of the problems that the project seems to have had.</p></pre>lolomfgkthxbai: <pre><blockquote>
<p>Someone might think that $25k should be enough to develop an indie game but in reality it's a very low amount when support for 4 platforms (PC/Mac/Linux and Ipad) has been promised.</p>
</blockquote>
<p>That kind of lowballed kickstarters are a huge red flag. You can't pay even one dev with that for half a year in a country where engineers are cheap!</p></pre>gondur: <pre><blockquote>
<p>most popular kickstarter failures was Mighty no 9 which despite reaching 4 whooping million dollars, till this day it still hasn't delivered the ports to the rest of the platforms it has promised, not to mention that the game was bad despite being developed by proven industry professionals.</p>
</blockquote>
<p>Interesting. But unlike this one the developers of haunts had the honesty of handing the game to the community as open source, so that at least theoretical someone could take the game up and achieve what was promised. Therefore I'm surprised no one did this up to now. But when the game was developed against an immature < 1.0 go version it is understandable.</p></pre>epiris: <pre><p>Just want to note I like that they published the source too. Software kickstarter projects should be required to fully open source if they fail to deliver in my opinion. It's the only way to prove they made a best effort for delivery and didn't fraud & fluff the backers. </p></pre>gondur: <pre><p>I fully agree! The core reason why this was interesting for me... </p></pre>jerf: <pre><p>The average software project fails. The average game project on Kickstarter fails too, as I understand it. There isn't necessarily a lot of analysis to be done, and a lot of it would just be kicking some people while they're down, given that this is basically in public. I don't think there's much to learn here, and I'd actually politely suggest that we not try to dissect the project years later.</p></pre>gondur: <pre><blockquote>
<p>that we not try to dissect the project years later.</p>
</blockquote>
<p>hmm, is not the best case for learning "learn form the errors of others"? ;)</p>
<p>I believe this is exactly the case to dissect ;)</p></pre>jerf: <pre><p>Then go study it yourself. But we don't need to publicly dissect it any more than has already been done. I don't see any reason to believe there's anything special enough about this project to justify further public analysis; it failed for the usual boring reasons things fail, from what I can see. There's no reason to impolitely drag up old topics for that.</p></pre>gondur: <pre><blockquote>
<p>There's no reason to impolitely drag up old topics for that.</p>
</blockquote>
<p>Are you involved here? What's your problem? They open sourced this stuff to public, here is no "politness" issue at all.</p>
<p>Way around, my motivation was two fold: first learn something about reasons, second, in best case a solution could be found here.
would be win-win for everyone.</p>
<p>I have to say I'm quite surprised by your "better keep issues covered" approach & point of view.</p></pre>
