The game came out in August and I spent the next month or so fixing bugs. That basically meant reducing the amount of distorted, rotated, variably-colored, or very-tightly-fit text. This sounds laughable considering how much text is in the game, but there were plenty of cases where I decided to replace some planned text with an icon or to remove it entirely.Īnything is possible with Photoshop, but I kept the general parameters of what a localized version of the game would need in my mind as much as possible. Limited the use of text as much as possible. These identifiers are helpfully validated at compile time if you mistype something, you’ll get an error message right away. Haxe has a nice feature where you can reference a specially-formatted xml file from a class definition to have it pull text from the data.
PAPERS PLEASE PASSPORT GENERATOR CODE
Some displayed text was split out from the code and moved into the data. These were a few small things I squeezed into the initial release: But I’ve worked on localized games before and I had some basic understanding about the things I’d need to make Papers, Please localizable. So for the initial release I’d resolved to ignore localization. This makes rapid prototyping really easy but sets you up in a slightly bad way for later production.
The build tools recompile the assets and code at the same time so the common incentive to keep data out of the code in order to improve iteration didn’t really apply. One tricky aspect of using Haxe/OpenFL to make the game is that updating a data file takes just as long as updating the source code. Haxe/OpenFL somewhat discourages separation of code and data. Waiting until the game was completely done made everything easier. There would’ve been a lot of wasted work if localization-friendly assets had to be generated each time I changed something early on. When creating things for the first time, the ability to iterate quickly in Photoshop on a single image/document was great.
PAPERS PLEASE PASSPORT GENERATOR FULL
Papers, Please is full of documents with precise layouts and an obscene number of different, tiny, pixel fonts. Limiting to a single language is much easier during development. In the end, that process took 4 months so in retrospect I’m glad I didn’t try to fit that into the initial release. Along with setting up the game to accept localizations, I’d also need to coordinate with others to do the actual translations. I work alone and among all the stuff I had to do to finish the game, “localize it” was near the bottom. But it would’ve been a lot of work without knowing for sure that it was worth the signifcant time, effort, and money investment. There were some requests for localized versions of the beta, and I had a general idea that I may want to eventually localize it. Skip to Part 2 for a less technical topics. Most of the information in Part 1 is pretty technical. Undeterred and unwilling to move again to wordpress or something, I’ve broken this post into two parts.
It came as no surprise that tumblr also can’t handle a single post of this size. Instead of breaking it up into multiple forum posts, I created this blog for hosting. It’s too big for the TIGForums devlog where it belongs. This massive post will go over my process for localizing the game.īefore writing this post I knew it would be long, but not this long.
Even though it would’ve been great to support other languages besides English, I decided to push that work until after the game’s initial release. Papers, Please was released on August 8th, 2013 without multilingual support.
0 kommentar(er)
