[Date Prev][Date Next][Thread Prev][Thread Next][Author Index][Date Index][Thread Index]
various gc proposals
- To: <tribble>
- Subject: various gc proposals
- From: Mark S. Miller <mark>
- Date: Mon, 20 Nov 89 19:24:22 PST
- Cc: <xtech>
- In-reply-to: <Eric>,25 PST <8911202202.AA25429@xanadu>
Date: Mon, 20 Nov 89 14:02:25 PST
From: tribble (Eric Dean Tribble)
1) Can Nat fix some of the unimplemented features? such that those
changes can then be publicized? we could even pay some for this. I
realize that it looks vaguely ridiculous, but it might be the easiest
thing to do.
I'm going to be persuing this. I think it's our best option.
2) Can we restrict the times when a garbage collection is allowed?
3) Can we generalize on 2)? We could define some way for users to
specify when garbage collection is allowed (or disallowed). ...
I hope that 2 and 3 let us move smoothly to compilers that support
more features of the language.
I would be very scared to go down this route. It seems *extremely*
complex to me. I haven't thought too hard about it, but it seems to
remove much of the motivation for having a garbage collector. It
would probably also be very dangerous and hard to check.
I think we're stuck either fixing the language in a way that can get
highly ported (like getting it into the 2.1 release), or doing without
&& and || in our code.