[e-lang] E-on-CL progress, and request for information
kpreid at attglobal.net
Sun Jun 26 07:53:33 EDT 2005
On Jun 25, 2005, at 22:16, Mark Miller wrote:
> Kevin Reid wrote:
>> Not having looked at how boot-comm works in E-on-Java: would a
>> reasonable implementation approach be hooking up a deSubgraphKit
>> builder and recognizer across a more-trivial comm system that can
>> pass their messages?
> This may eventually be a good way to do a boot-comm system, but it
> won't help you talk to the one we've got. The one we've got doesn't
> serialize, but instead imposes some draconian restrictions.
I did mean for E-on-CL as a standalone E system. In the general case, I
do not expect to be running it in a JVM.
>> Communications happen by pointer manipulation, not serialization, so
>> the only PassByConstruction arguments that may be passed or returned
>> as results are those that are DeepPassByCopy.
> That's not quite true. It also makes a special case for ConstLists:
> you can pass a ConstList, given that each of the things in the
> ConstList are either PassByProxy or DeepPassByCopy.
Do I correctly understand that the boot-comm system *is* what all
intra-process inter-vat communication happens on?
Kevin Reid <http://homepage.mac.com/kpreid/>
More information about the e-lang