State of Ren'Py, 2014

[PyTom]

At the start of the new year, it's traditional to look back at where we've been, and forward to what's next. I hope you can bear with me for a little bit as I reflect on Ren'Py, the forums, and myself.

As Ren'Py enters its eleventh calendar year of development, I think it's going as strongly as ever. Indeed, it's blossoming as a proper open source project, with 20% of contributions coming from other Ren'Py developers. Contributions range from large features to translations to small fixes - every one makes Ren'Py better. I thank everyone who's contributed to Ren'Py in the last year.

I'm proud of my own level of contribution to Ren'Py. Although I didn't meet my goal of contributing to Ren'Py every day, I did manage to make contributions on over half the days - 825 total, according to github, including a 100+ day streak that's still going on.

I hope that we can meet the expectations of the Ren'Py community, which downloaded the SDK over 41,000 times. Over 130 visual novels and other games were released last year - potentially many more, as that only counts projects listed on games.renpy.org. Ren'Py was used to make all sorts of games - commercial Steam games, Kickstarter demos, multi-year epics, jam games, tiny experiments - and everything in between. I'm proud that we could be a small part of such a creative community.

I also think the forums have been running incredibly well for the past year. While we've had a few technical issues, those seem to have been sorted out as of late. We've been largely drama free, and we even have the kitchen spammer under control. People are friendly and helpful - the way people help each other in the Ren'Py section has been instrumental in giving me more time to program. Thanks to everyone who's been a part of this.

There were two releases of Ren'Py in 2013, giving us a new translation framework and massively improve Android support, as well as a host of other features. (Have you tried the shift+O console yet?) These raw numbers kind of underestimate the development velocity of Ren'Py, as 6.15 should probably count as two releases - there were dozens of improvements in the 3 months between the first and last 6.15 release, and much of the work for the next release is already done.

One of my goals going forward is to try to have smaller and more frequent releases. To that end, we came out with the Ren'Py nightly build site - http://nightly.renpy.org/ - where you can download the latest Ren'Py code, fully built, whenever you want. My hope is this will allow people to test and develop with the latest features, when they want them. At the very least, having Ren'Py built nightly will alert me to errors in the build process, making actual releases easier.

Another goal is to make it easier for people to contribute to the Ren'Py project. While we've had some phenomenal contributions by programmers, documentation writers, and translators, right now contributing to Ren'Py involves dealing with tools not everyone is familiar with. There's been a lot of progress on that front in 2013, and my hope is 2014 will bring it further along.

I think there's a lot of opportunities for people to contribute to Ren'Py, even if you're not that familiar with Python programming. At 800x600, the Tutorial and The Question game are kind of showing their age. The default Ren'Py theme never looked good. In both cases, the path to improvement will involve working with people who are far better at art, writing, music, and Ren'Py programming than I am.

(Of course, there are people who are way better Ren'Py programmers than I am. I mostly work in Python and Cython, and while I created Ren'Py, the way I use it is different than the way most people do.)

Beyond that, I plan to keep developing Ren'Py. Although I don't have firm plans beyond the next release, which will focus on improving the style system, my goals for 2014 are to modernize the foundations of Ren'Py, by updating the software it depends on, even if it means rewriting portions of Ren'Py.

I'm also hoping to do a few things - outside the usual programming - for the 10th anniversary of Ren'Py. Although I might have to pick some people's brains about this, maybe some sort of merchandise? I've also been pondering the idea of a contest involving educational VNs, so hopefully I'll be able to run that sometime this year.

Lastly, I have to say that the highlight of 2013 for me was going down to Dallas to see people from the forum. Being able to see fellow creators - people I've known online forever, but never met in real life - was phenomenal. I enjoyed myself more on that one trip than I did in any vacation I took in my adult life. I'd been to cons before, but being with such phenomenal people brought it to a whole new level of experience. (And then there was the time we wound up on the wrong side of the tracks - the bottom. ) I enjoyed it so much, I'm hoping to go to more cons this year - that is, if we can somehow organize LSF events, or even just informal hangouts, while we're there.

So here's to 2013, 2014, to us, and to the medium that unites us all.

[Omniknight]

I'm lovin' it.

[SundownKid]

I do feel like the default Ren'py theme could use an upgrade. At least something that is universally stylish and minimalist looking. Perhaps resembling the redesign the launcher received from before.

[Yomuchan]

11 years, eh? I'm proud of you. Few have survived this long.

Keep up the good work!

[Tempus]

I wasn't aware so many contributions came from other developers -- that's cool. Also, I volunteer to help with updated background art and/or music for the Tutorial and The Question.

[i1abnrk]

Yes, Shift-O \m/ and a w00t. It's pretty close to a full debugger. So good. Don't really need the linux prompt at all. An adapter console with renpy features. *back-sommersaults* it's so promising. The developer network is one of the best anywhere that I've seen as a lazy hacker. The only exception is perhaps freeciv but you have to dig to get the really good stuff there. But the good hacker stuff happens in general as well as the developer thread which is so rare. I might call it unique: ether chest in FF hard to find. Minetest is the third. Actually, getting an actual function called 'render' that pays the bills in a cross platform manner without c or java libs saves me rediculous time. Cheers for fostering a great community.

[LofnBard]

As a two month old user, I find it's not the new features that matter to me, it's the documentation. More and better features are nice, but the documentation is lagging so far behind it makes the second step into using Ren'Py extremely steep. You've done a truly awesome job with the first step tutorials and basic features. As the ad says, "so simple a caveman could do it." It helps that these features are fairly invariant across versions. But as soon as one tries to get the slightest bit fancier with a little Python or even Screens and Screen actions, there's a vast abyss of brief descriptions with no examples. Does a given function's parameter require quotes? No way to know without searching the forum. Searching the documentation just doesn't yield any examples in most cases.

It's true the forum is great, but its search function sucks. I was looking today for "ui.key". No examples of usage in the documentation. There is in fact an example in the wiki cookbook, but the documentation search doesn't look there. I only found it by going across pages of forum hits, where someone posted a link to the wiki cookbook page. Life would be infinitely easier if the forum search supported exact quotes, but it doesn't. Even a period isn't sufficient to make it one word, so "ui.key" gets split into searching for any page that has "ui" and "key". It doesn't even rank them by proximity as far as I can tell. It gets pretty maddening.

Now, letting the documentation lag behind has a few effects:
1) It favors experienced power users who want more features, and makes the Ren'Py club more prestigious and exclusive.
2) It greatly increases how much time veterans have to spend answering the same questions over and over on the forum, because folks can't find the answers in the documentation or by searching the forum, even when answers are already there.
2) It drives away new users, who give up in frustration when they get stumped. Repeatedly. It makes them feel stupid that they can't figure it out and keep having to beg for help.

Now I'm not a complete noob, I used to teach C++ programming in college. I can find how to do stuff in Python on the web just fine. But it took all my stubbornness, the support of my Pixelles mentor, the six weeks of "make your first game" group meetings, helpful people on the forum, and my childhood dream of making games to not give up and let frustration get the better of me.

I think Ren'Py is a fabulous tool for making great games. Yet I'm hesitant to recommend it to anyone who isn't already a programmer. That second step is a real doozy, and for the good this game engine and its future users, I really wish more effort went into documentation, especially in adding sample code for all the functions (or a link to a larger bit of code which uses that function among many others). All that being said, thank you for making Ren'Py available to all of us, and for your continued efforts to make it better.

[Asceai]

As a user of ren'py for a similar amount of time to LofnBard, I have to agree re: the documentation. I've gotten used to grepping the ren'py source to find stuff now, but it's not the best solution (and I worry every time I use anything that starts with an underscore that I'm using stuff that might not exist in future versions of ren'py)

Basically, the general thing I've found is that stuff I think will be quite complex turns out to be pretty simple and the stuff I think will be simple turns out to be a nightmare of many hours to solve. Which is more my inexperience than anything else, but better documentation sure would help. Complex screens (UI programming, which I generally think is a nightmare but ren'py makes really nice) are a cinch, but changing the metadata attached to an existing save slot? Bit harder!

Of course, I always like new features, and I'd sort of like to see live and image manipulator approaches to do 4-point deformation. This would instantly enable numerous effects and largely complete ren'py from a 2D graphical standpoint (the only other thing being live colouring) and, among other things, make a smooth-scrolling and -turning dungeon crawler feasible, which I bet people are itching to make. Smooth forward and backward motion is already possible with pre-deformed wall and floor textures and zoom, but doing that and having instant turns feels a bit off.

[PyTom]

As a two month old user, I find it's not the new features that matter to me, it's the documentation. More and better features are nice, but the documentation is lagging so far behind it makes the second step into using Ren'Py extremely steep.

I think the documentation is starting to catch up a bit. For example (and this hasn't made it into a release yet) renpy.exports is now fully documented.

The big problem is that there isn't a lot of intermediate level documentation. That's probably because writing the quickstart is easy - it's basically the talk I give when people ask me how to get started with the basics of Ren'Py. As you get into more advanced things, we start hitting problems - I've never taught anyone how to make custom screens, and I never learned how to make them. So it's kind of hard for me to write the documentation myself.

There's also the issue of my motivation. Very little of the compensation I receive for making Ren'Py is monetary in nature. Most of what I get out of it is being able to work on an interesting problem, followed by getting to play the games people make. So it's tough to motivate myself to work on documentation, rather than the many far more interesting problems that remain. I've been trying to do it - but it does take a long time.

Of course, Ren'Py is open source. If people have guides, tutorials, etc, that you think should be included with the docs, send them to me and I'll try to edit them in.

[Obscura]

An idea: if there's anyone excited about teaching Renpy basics, advanced methods, and programming concepts to users, perhaps you can set up a website with lessons and examples and ask for donations to fund you from lesson to lesson.

I don't think this would be a huge money maker given the size of its user base, but I can foresee it being a fun project or a way to make a little extra cash on the side, especially if you're a programmer or aspiring programmer.

[LofnBard]

The big problem is that there isn't a lot of intermediate level documentation. That's probably because writing the quickstart is easy - it's basically the talk I give when people ask me how to get started with the basics of Ren'Py. As you get into more advanced things, we start hitting problems - I've never taught anyone how to make custom screens, and I never learned how to make them. So it's kind of hard for me to write the documentation myself.

There's also the issue of my motivation. Very little of the compensation I receive for making Ren'Py is monetary in nature. Most of what I get out of it is being able to work on an interesting problem, followed by getting to play the games people make. So it's tough to motivate myself to work on documentation, rather than the many far more interesting problems that remain. I've been trying to do it - but it does take a long time.

Of course, Ren'Py is open source. If people have guides, tutorials, etc, that you think should be included with the docs, send them to me and I'll try to edit them in.

I hear you, and understand why it wouldn't be much fun for you, so let's think about how other groups solve this.

These days they typically make a wiki so fans can fill it with the relevant information. We already have folks who like to answer questions and solve coding needs, but their efforts get lost in a difficult to search forum. Some things make it into the cookbook section of the forum, but for the most part they're complete solutions, not tutorials, and posts aren't the ideal format for that.

When I started I pretty much ignored the Wiki because it said it was out of date, and I didn't want to learn deprecated functions. There's no easy way for a newbie to sort out what's current and what's ancient in there. I also feel your pain in dealing with spammers, I've had the problem myself, so it shouldn't be completely open. Yet we do have trustworthy people who've proven themselves in the forum. It's just that their work is being rewarded by the appreciation of a very small number of folks.

If I bust my ass to make a tutorial and have one person thank me, that's nice. If pageviews show that thousands of people consulted my wiki sample code page on the 'usage of the Key user interface statement' (which has NO examples whatsoever), well then, I feel like a bloody hero! Put a link to it from the documentation, and we're making solid progress.

Or, make a full function index in the wiki that clearly shows which functions have no examples. Let people see what needs doing, and let them apply for authorization to make those pages. Check the person out in the forum, see if you'd trust them, and away we go. Once you have a dozen or so authors, ask them who wants to take over the authorization process so it's off your hands. Divide and conquer, or rather, Divide, Delegate and Conquer.

Have a Thank You page in the Wiki which counts how many articles an author has added, and a way for people to find all the pages of that particular author. This way authors get compensated by prestige, fans of their tutorials, as well as helpfulness (shown by pageviews on each article). Everyone wants appreciation, just give them a way to get some.

[saguaro]

If there was a list of things needing examples (or needing multiple examples) I would be happy to help chip away at it if I could. As it is, I'm not sure where to start, I've been working on some SL cookbook recipes with the ideas that people who don't want to learn will use it as-is and people who do want to learn will use it as an example/tutorial of how to make a thing. But I don't think the cookbook forum gets too many eyeballs.

[nyaatrap]

To be honest, I think Ren'py is one of the most well-documented VN engine officially. You can get enough document by searching ren'py folder in SDK. You'll go crazy if you start leaning more famous Japanese VN engines without outer documentations, which are made by users. I think It's user's task to make basic/intermediate tutorials, not a creator.
BTW, I agree the forum search engine is crap. I just use google instead.

[PyTom]

http://www.renpy.org/wiki/Special:RecentChanges

Shows the problem with the wiki in a nutshell. Back when it was open to the world, the amount of spam we had was immense, to the point where dealing with it was becoming a burden on the development of Ren'Py itself. I don't know how to fix it, but I'm skeptical that a return to the wiki form is part of the solution.

I want to come up with ways of getting other people involved with documentation - but I can't take on another large moderation burden, especially one as unrewarding as a wiki.

[jack_norton]

With my own wiki I setup so that I manually approve each person. Usually there are two types of person: those who are committed and WILL contribute, and those that don't. So instead of making open to anyone, you need to give access only to the first group