[bitc-dev] Retrospective Thoughts on BitC

[wren ng thornton]

On 3/23/12 11:33 PM, Rick Richardson wrote:
> This should be published in a PDF, not in a mailing list.
>
> While I can understand your frustration, I hope that you don't feel
> that your effort
> was wasted. Because I, along with many others, learned a tremendous amount
> simply by sitting on the sidelines of this project.
>
> Thank you.

I agree it'd be good to present the retrospective to a wider audience. 
These sorts of things are too often lost to the sands of time, but it's 
the sort of insider knowledge you really hope for when mining old 
projects for new ideas. John Regehr had a post about this issue 
recently: <http://blog.regehr.org/archives/667>

One of the big things about BitC, IMO, is that it focuses/ed on the 
theory of systems languages, which is a deeply underexplored area within 
PL research. There's a good deal of systems research, and a good deal of 
theory, but very little of their intersection. Which is why we're still 
stuck with languages like C and C++ as the dominant choice for systems 
code, despite the fact that they're, at best, languages we love to hate.

But maybe we can learn enough to make a better go of it next time. (For 
example, I don't entirely follow why you now think subtyping is requisite.)

-- 
Live well,
~wren