[Date Prev][Date Next][Thread Prev][Thread Next][Author Index][Date Index][Thread Index]
various gc proposals
- To: <xtech>
- Subject: various gc proposals
- From: Eric Dean Tribble <tribble>
- Date: Mon, 20 Nov 89 14:02:25 PST
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.
2) Can we restrict the times when a garbage collection is allowed?
This could work for the backend because all transactions are bounded.
We might be able to make it work for the frontend as well. To first
approximation, we could garbage collect after every event. We might
even be able to guarantee that the intervening computation takes a
bounded amount of space.
3) Can we generalize on 2)? We could define some way for users to
specify when garbage collection is allowed (or disallowed). This can
only be answered by those who've sweated over making the programs
garbage collectable. ?
I hope that 2 and 3 let us move smoothly to compilers that support
more features of the language.