HolonT is a new Holonforth system that creates Tcl applications. HolonTForth uses Tcl as its native language. With HolonT you have the Tcl universum available in Forth.
Why Tcl? Tcl offers everything I could possibly want. Strings, lists, regular expressions, fine graphic tool kits (Tk and others), web tools, a great database (Metakit), OO in many flavors, and more. Tcl is available on all platforms, and Tcl applications run everywhere. Moreover there is this brilliant Tclkit, which contains the whole Tcl system in one executable file. Tclkit embodies a virtual file system and in it the commonly used Tcl packages, including Tk and Metakit. It is included with the HolonT packages.
So why don't I just use Tcl? I did. Four years ago I switched to Tcl because it offers the features that I couldn't find in Forth. I turned to Tcl in order to develop the new Holon (which I presented here last year), and have now used the new Holon to build HolonT.
So why Forth again? Tcl has, for me, one fundamental problem, its prefix notation. Prefix notation creates an additional layer of complexity that I simply can't adapt to. I still have problems with the rules of argument substitutions, with braces and brackets and taking care that the command arguments really form correct lists. Tcl is described by 12 rules. If you convert Tcl to postfix notation as in HolonTForth, 6 of the rules simply vanish. Braces, brackets, substitutions ond argument lists evaporate. Back to the roots of simplicity!
Note: This is a first version, still experimental. It provides the compiler mechanism and Forth words for the commonly used parts of the Tcl kernel as well as a couple of Tk words that I have used so far in the sample chess and private programs. Also note: this system is not competing in the speed race.
> HolonT is a new Holonforth system that creates Tcl applications. > HolonTForth uses Tcl as its native language. With HolonT you have the > Tcl universum available in Forth.
Wolf Wejgaard wrote: > HolonT is a new Holonforth system that creates Tcl applications. > HolonTForth uses Tcl as its native language. With HolonT you have the > Tcl universum available in Forth.
> Why Tcl? Tcl offers everything I could possibly want. Strings, lists, > regular expressions, fine graphic tool kits (Tk and others), web > tools, a great database (Metakit), OO in many flavors, and more. Tcl > is available on all platforms, and Tcl applications run everywhere. > Moreover there is this brilliant Tclkit, which contains the whole Tcl > system in one executable file. Tclkit embodies a virtual file system > and in it the commonly used Tcl packages, including Tk and Metakit. It > is included with the HolonT packages.
> So why don't I just use Tcl? I did. Four years ago I switched to Tcl > because it offers the features that I couldn't find in Forth. I turned > to Tcl in order to develop the new Holon (which I presented here last > year), and have now used the new Holon to build HolonT.
> So why Forth again? Tcl has, for me, one fundamental problem, its > prefix notation. Prefix notation creates an additional layer of > complexity that I simply can't adapt to. I still have problems with > the rules of argument substitutions, with braces and brackets and > taking care that the command arguments really form correct lists. Tcl > is described by 12 rules. If you convert Tcl to postfix notation as in > HolonTForth, 6 of the rules simply vanish. Braces, brackets, > substitutions ond argument lists evaporate. Back to the roots of > simplicity!
> Note: This is a first version, still experimental. It provides the > compiler mechanism and Forth words for the commonly used parts of the > Tcl kernel as well as a couple of Tk words that I have used so far in > the sample chess and private programs. Also note: this system is not > competing in the speed race.
How about the original HolonForth will it be open source one day, I started migrating it to the MSP430, had the assembler working but without the source that was the end of the progress.
I really like they way it organizes the source code, makes it very easy to work on a project.
On 13 Mai, 02:20, Don Seglio <don.seg...@sbcglobal.net> wrote:
> How about the original HolonForth will it be open source one day, I > started migrating it to the MSP430, had the assembler working but > without the source that was the end of the progress.
> I really like they way it organizes the source code, makes it very easy > to work on a project.
For a while I have offered my cooperation for porting Holon to new targets (see the topica mailing list). The assembler is only the first step in the involved process. You need a monitor in the target, must create the Forth kernel words, adapt the umbilical communication, tune the debugger to the target uP, and more. There have been several attempts but only one project actually made it to the end (HolonR2k). It was considerably more work than anticipated for both parties.
The process particularly showed that the old source of Holon needs a lot of explanation, it contains the what and the how but not the why. I don't think it is a good idea to distribute it.
Instead of commenting source of an old closed world, I will rather give you all the information you want about Holonforth, based on the new open source systems. Just ask, there exists a google.group holonforth ready for discussions.
