Hello Fare, Thanks for the response. I do not have mailing list or public repository for this topic. I still read different papers. I try to formulate better the problem in order to be solved easily. Since nobody answered me so long, I thought more about the form of my writings. The second attempt to define the problem:
"I interpret the writings from the institute (vpri.org) in the following way. What was possible before from the many people and big organizations, today should be possible from a single person. So, let us test this assumption. What is the most simple and expressive way to program a computing device (FPGA) in order to compute a symbolic expression as input from a keyboard and to output the result in a display (or in a console)? The man should be able to change the system down to the metal. In fact it is possible to compute the symbolic expression in the simple way (but not expressive). Just ones should take the main road using a modern computer, Linux OS and Maru language. But if ones want to deviate from the main road, there are tons of programming languages, external libraries, APIs and the operation system that he has to know just for a small change in the hardware or the software." Fare, The Linear Lisp by Henri Baker is one alternative route for symbolic computing in simple and expressive way. From his paper I understand that there is no arithmetic operations in Linear Lisp, just operations on symbols. He proposes the usage of his system for symbolic algebra. Maybe this is your field of interest. I am still evaluating different implementation languages. Maru is one of them. I believe that it can make *instruction level simulation* of the Linear Lisp, since more of the instructions are present in Maru itself (car, cdr, push, pop and etc). But for the *cycle accurate simulation* Maru has to have notion of time or to use in some way the object oriented features (send messages to objects). It is area in which I do not have much knowledge. I also personally spend most of time thinking before produce any code. For example, it took me one year to find the solution of a problem. The time to create the solution as simple expression in Maru took me 2 weeks including the time to learn the basics of Maru. Regards, Iliya 2013/10/10 Faré <[email protected]> > Dear Ilya, > > I've been wanting to experiment with a Linear Lisp in Maru for a long time. > Do you have a mailing-list to discuss and/or public repository to work > with? > > Regarding practical programming with Linear Types, see Jesse Tov's thesis > at > http://www.eecs.harvard.edu/~tov/pubs/dissertation/ > Jesse also recommends reading about Nikhil Swamy's work on Fine and Fstar. > http://research.microsoft.com/en-us/um/people/nswamy/papers/ > And if you are into foundational theories, > you could give a look at Computability Logic. > > PS: Sorry for a reply long in the making — I'm just out of > a long procrastination period and ran out of excuses > (having nailed the coffin on ASDF, a quick-build extension for it, > and the Lisp-Interface-Library using that). > > —♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• > http://fare.tunes.org > If you argue and rankle and contradict, you may achieve a victory > sometimes; > but it will be an empty victory because you will never get your opponent's > good will. — Benjamin Franklin > > > On Wed, Sep 18, 2013 at 4:46 AM, Iliya Georgiev <[email protected]> > wrote: > > I would like to make a simulator of a Lisp Machine. If it is successful, > > then I will program a FPGA with Linear Lisp and use Lisp-like > interpreter to > > execute programs over it. My main motivation is to be able to understand > and > > change the system as I want. The system should be as simple as possible. > > The history of the ideas in computer science is used as a method of the > > research. The ideas of different LISP Machine theoretic models were > found. > > They differentiate in the type of register or type of structure and the > > instructions. The Linear Lisp model seems encouraging for the purposes. > It > > uses finite state machine with pointer register, data structure with > unity > > reference counts (tree?) and omits the usage of garbage collector. [1] > Also > > the interpreter for Linear Lisp is already created. So efforts in this > > direction will be saved. Maru language is proposed for the > implementation of > > the simulator, because it is not bounded to its own structure. > > > > After this lengthy introduction I have two questions: > > > > 1. Is it possible to make "data structure with unity reference counts" in > > Maru? This is crucial for such a simulator. > > > > 2. Do you know of example that describes real finite state machine with > > pointer register? Most papers in this topic seems to be just theory. [2] > > > > > > > > References: > > 1. Publication on Linear Lisp theoretic model - > > http://home.pipeline.com/~hbaker1/LinearLisp.html > > 2. Pointer machine - http://en.wikipedia.org/wiki/Pointer_machine > _______________________________________________ > fonc mailing list > [email protected] > http://vpri.org/mailman/listinfo/fonc >
_______________________________________________ fonc mailing list [email protected] http://vpri.org/mailman/listinfo/fonc
