<div dir="ltr"><div>Yeah, I would want to get the arguments from the stack, but everything else is unnecessary to do things like FloatArrayPlugin.</div><div> </div><div>BTW, did anyone use Lowcode in Pharo? I didn't know it, I just saw references to it in the opensmalltalk-vm source tree and saw a presentation about it.. it looks like writing primitives using Cog JIT RTL opcodes from within the image, and you can call functions from external libraries too... </div></div> <div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Mon, Jun 1, 2020 at 1:14 AM David T. Lewis <<a href="mailto:lewis@mail.msen.com">lewis@mail.msen.com</a>> wrote: </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">The interpreter proxy provides the interface to the Smalltalk call stack, along with a number of other functions. Without that interface, you cannot take arguments off the stack or return values back to the image. It also provides the tools you need to (for example) create a new instance of String that you can return back to the image. Dave On Sun, May 31, 2020 at 12:14:36PM +0700, Luciano Notarfrancesco wrote: > About the interpreterProxy, it looks like there???s a subset of slang that???s > good enough for a lot of applications and might not really need > interpreterProxy. I don???t remember all that it is used for, but thinking > about MiscPrimitivePlugin and FloatArrayPlugin, all they do is receive > arrays (of floats, bytes, it would be nice to also have float64, int32 and > int64) and read/write stuff from/to them. They never deal with the object > memory, they don???t instantiate objects, they don???t even access instance > variables. This subset is all I need for my use case, and I???m guessing is > also all that would be needed for a vector graphics plugin and many other > applications. > > On Sun, 31 May 2020 at 4:25 AM, Philip Bernhart via Cuis-dev < > <a href="mailto:cuis-dev@lists.cuis.st" target="_blank">cuis-dev@lists.cuis.st</a>> wrote: > > > Hi David, > > > > "David T. Lewis via Cuis-dev" <<a href="mailto:cuis-dev@lists.cuis.st" target="_blank">cuis-dev@lists.cuis.st</a>> writes: > > > > > Don't believe everything you read on vm-dev ;-) > > > If you are interested in writing your own primitives for just about > > > any reason at all, then named primitives are exactly what you want. > > > > Ok, if you say so. But why should I believe you? :-) > > > > > Regarding stand-along slang generation, as a point of comparison you > > > may want to take a look at the interactive "slang browsing" in the > > > traditional VMMaker. This is set up to attempt C code translation for > > > any instance-side method in any class (not just the VM classes). > > > > I don't know what you want to say with that. If just as a feature to > > compare code between old and "new", then it's interesting and cool. > > > > Still... there is the problem of how to sustainable co-evolve against > > the same VM interface. For example which implementation of the > > interpreterProxy should be serialized? I would need some answers > > around that whole massive topic, before starting to implement > > a Cuis Slang PluginMaker. > > > > Anyway the feature inspired me to implement that for Cuis Slang: > > > > > > <a href="https://github.com/Phantasus/Cuis-Smalltalk-Slang/blob/master/Journal.md#30th-may-2020-jpb" rel="noreferrer" target="_blank">https://github.com/Phantasus/Cuis-Smalltalk-Slang/blob/master/Journal.md#30th-may-2020-jpb</a> > > > > > > Thanks, > > Philip > > -- > > Cuis-dev mailing list > > <a href="mailto:Cuis-dev@lists.cuis.st" target="_blank">Cuis-dev@lists.cuis.st</a> > > <a href="https://lists.cuis.st/mailman/listinfo/cuis-dev" rel="noreferrer" target="_blank">https://lists.cuis.st/mailman/listinfo/cuis-dev</a> > > </blockquote></div>