Why haven't Inform been fully open-sourced?

I can see only benefits of doing that: Faster development because more people could take a stab at improving Inform, improving the Inform parser (I mean cross compiler into Inform 6 here) using the advancements in the natural language processing field (A thing a single dev may not be able to fully realize, despite all the work that has been put so far into Inform) so it can understand more ambiguous sentences and phrases, faster development (the last version seems to be from 2015).

Plus quality of life stuff such as

  • Ability in the Inform IDE to separate the second half into its own window which would help in multi-monitor setups with non-uniform resolutions so you can have docs on one screen and the code on the other.
  • Ability to separate sources into separate files that are then joined into the main file using an instruction similar to C++'s #include, which would help organizing the code e.g. on a chapter-by-chapter basis.

If that sounds like a feature wishlist, it’s because it kinda is. But many of these (e.g. the IDE stuff or the include thing, which would probably use a phrase like "Put a file here as part of the source) I could try and do on my own if Inform would be fully open-sourced.

2 Likes

To make that happen, the copyright holders of Inform 7 will need to actively grant license to recipients under some free-software license. (They could choose the Apache License 2.0, for consistency with Inform 6, as an obvious choice.)

You may want to lend your support to the same request made at the Inform 7 suggestions site: inform7.uservoice.com/forums/57 … -on-github

1 Like

The IDEs are open source.

As for the compiler, please be aware that Graham doesn’t read this forum, so making an argument here is low-impact. The link above is the official suggestion line.

Is Graham the only copyright holder in the compiler, standard library, etc.? My understanding is that we need a discussion among all the current copyright holders in the work, to grant license under free-software terms.

Graham, Emily, and Zarf probably share copyright, at least over parts of it.

Make sure you read Graham’s recent presentation for an update on his plans: emshort.com/ifmu/inform.html

He does want to open source it eventually, but he hasn’t thought it was ready previously. Many of us would disagree, but it’s not our call to make.

Yes, the I7 compiler and the Standard Rules are entirely written by Graham.

Not me. I don’t think Emily either, although she wrote a bunch of the manual examples.

The I6 parser and template code is more mixed, but that’s because its history goes back to Inform 6, which is open source. (Perl license, by the way, not Apache.) Sorting that out will not be a big issue.

This passage is directly relevant to the question:

1 Like

Presumably the “twisty little passages” are being fixed with this cleanup. The article doesn’t mention though, whether the announced 2018 release will be open sourced.

Also note that these features already exist:

You can open another project in a separate window and browse the documentation in that one. Source code can be separated to different files by using project-specific extensions (see ch. 27.3. in the manual.)

Even if these aren’t exactly what you’re looking for you don’t need compiler or core Inform support to make them, only the IDE code which is already open source as mentioned.

That’s nothing stopping Inform from being open-sourced, really. If anything, the more volunteers would be able to clean up the code base much faster than a lone programmer, with many other things to do, things that pay, ever could. And “huge Perl scripts which were just mazes of twisty little passages” is a tautology, I think. Nothing wrong with Perl as a language, but it’s very good at producing incomprehensible code, that’s kinda Perl’s thing.

It is if Graham thinks it is.

If you disagree, tell him.