[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: State of the VM interface
From: |
Tom Tromey |
Subject: |
Re: State of the VM interface |
Date: |
16 Dec 2005 14:50:26 -0700 |
User-agent: |
Gnus/5.09 (Gnus v5.9.0) Emacs/21.3.50 |
>>>>> "Roman" == Roman Kennke <address@hidden> writes:
Roman> I investigated the state of our VM interface by looking at the stuff in
Roman> native/jni and here are the classes that seem to directly access native
Roman> code without utilizing a VM interface:
I think there are already PRs for most of these.
I would suggest filing PRs for the remaining ones and making a
"native code out of core" meta-PR linking to them all.
Roman> The remaining stuff is for the midi and iconv providers, I think these
Roman> are optional anyway and don't belong in a VM interface (or do they?).
No, they don't belong in a VM interface. A provider is already behind
an interface wall. IMO the appropriate approach here is to simply
write multiple providers.
Tom