Will F # ever be open?

Question

In early 2009, the issue of whether Microsoft will release a source for the F # compiler under the MS-PL / other license was discussed. A https://stackoverflow.com/a/4148141/how-to-setup-statement-in-javascript/232837#232837

A lot has happened since then. We saw the official version of F # with the .NET Framework 4.0 (and Visual Studio 2010), and as far as I know, it is still completely closed. Is Microsoft simply silent on this issue or has explicitly stated that they no longer intend to open the source code of the compiler? Perhaps things are already in the process. Basically, any news / thoughts?

Answer 1

(As others pointed out, the source was always available, but not yet under the MS-PL "approved open source" license - it currently has a more restrictive license.)

If I were completely speculative, I could assume that there are a number of things that could “delay” the open source release of the F # compiler, including the following:

• The compiler code needs a little clean up. The source has always been publicly available, but not many have looked at it without an open source license. If you open it, people will watch, and in this case it would be nice if the code complied with at least some basic style rules, for example, using RecommendedDotNETNamingConventions , and not old_legacy_ones . In a sense, the F # open source compiler will be one of the “canonical F # applications”, so it would be important for the code to be of high quality with respect to basic things such as the F # coding conventions (which evolved over time in the past 5+ years since compiler code was developed).
• Current code is hard to build on any platform. An open source release will require at least reasonable documents on how to build a compiler (still non-trivial today!) And ideally build scripts for major platforms (e.g. Windows / linux / etc).
• Even if IronPython / IronRuby ("open" teams) "just downstream" from the F # team to MS, making things "open" still requires a lot of buy-in / extract from the manual, and re-registering a new manual. if management changes before you get the first open source release out of the door.
• (all the usual "overhead" for managing an open source project)
• All of the above applies to the workforce, and the labor spent on these things is human power not spent on other things, for example, work on the next version of F #. Thus, in practice, it may be more appropriate for several people who do F # to bite off at work above in their free time, instead of devoting, say, a whole month, to focus on the open source release. So it can slow down. (As other of them relate tangentially, pointing to links to job postings, some of the workforce can hypothetically fill up with interns at MSR.)

I emphasize that all this is a completely hypothetical assumption, since for a long time there was no official word to anyone.

Answer 2

As Robert's comment on your question indicates, source code is already available as part of each installation, although it does not come with an open source license. Besides reading between the lines, I think that things like this blog post from Don Syme still indicate open source release as a priority for the team.

Answer 3

Is that a question? I’m not sure, it’s rather a request for any news related to the current situation. Adding “considerations” to the query is confusing, what are the considerations? The MS C # compiler is a closed source, but the C # specification has ECMA. The F # specification was not open to the wider community, which is an important part that I feel.

Answer 4

The decision is left with Microsoft, I don’t think anyone here can answer that. However, even if it is a closed source, we are likely to still have all the advantages of this framework, as Microsoft is fully committed to improving and providing more functionality within the framework. I think that even if something is not open or not, but the creator supports it or not, this is the biggest problem I have. We have the tones of open source projects, but they become garbage code after they are not supported and never improved.