[Oberon] Slim binaries / OMI / Juice implementation details availability
skulski at pas.rochester.edu
Wed Aug 29 21:07:32 CEST 2018
From: Oberon [oberon-bounces at lists.inf.ethz.ch] on behalf of Erik Terpstra [erterpstra at gmail.com]
Sent: Wednesday, August 29, 2018 2:46 PM
To: ETH Oberon and related systems
Subject: Re: [Oberon] Slim binaries / OMI / Juice implementation details availability
I am not particularly interested in its portability or compressibility aspects, but the fact that slim binaries preserve the control flow structure of the original program, this enables (among other things) interesting solutions for aggressive (re-)optimization based on run-time profiling data.
I also like its security aspects (like not being able to construct slim binaries that violate the scoping rules).
On Wed, 29 Aug 2018 at 20:24 Jörg <joerg.straube at iaeth.ch<mailto:joerg.straube at iaeth.ch>> wrote:
I don‘t have access to OMI / SlimBinary source.
On the other side, I wonder whether this concept is really needed anymore. Seeing that the RISC-5 CPU is running emulated on Windows, Mac and different FPGAs one could say the RISC-5 „assembler code“ can be seen as kind of SlimBinary :-)
Like the Java classes can run in any JVM. The goal of SlimBinary to distribute executables that can run anywhere is somewhat reached, isn‘t it?
> Am 29.08.2018 um 18:58 schrieb Erik Terpstra <erterpstra at gmail.com<mailto:erterpstra at gmail.com>>:
> Are there still implementation details around Slim binaries / OMI / Juice available in the form of:
> - source code of a compiler that targets it
> - source code of a slim binary loader & code generator
> - example files (i.e. actual slim binaries to inspect in a hex editor/viewer)
> If I understand correctly the (now deprecated) Mac Oberon & Win Oberon systems implemented it, but I can't find actual implementation details, source code or example [slim]binaries.
> Any help would be appreciated.
More information about the Oberon