Version 0.10 Last updated 2002-07-17-03:20 -0700 (pdt)
I just had a nasty crash while updating the oMiser sketch. This prompts me to establish more reliable practices and backup procedures (N020700). In addition, I want to start keeping a history of the stages of the sketch. It may be of interest to others who follow me down this path. It is of special interest to me to capture my changes in thinking as I articulate what Miser is and what there is to learn from it, and that I am learning from setting that all down.
-- Dennis E. Hamilton
Seattle, Washington
2002 July 3
Here are preserved the latest copies of each identified version of the sketch. (That is, 0.10 is the last updating of version 0.10, etc.)
Some of the earlier versions were edited slightly to connect them into this history structure. From then on, al that we do is rename the archived snapshot, separating it from the evolving, on-going sketch.
My brain runneth over. There are notes and thinking out loud and items like that in the body of sketch versions themselves. Other notes move out to here and are dealt with or not as time permits. The sketch backlog is interesting to me, and it might be helpful to others in seeing where I am trying to get with this.
2002-07-10-11:25: Technical Things. There are some organizational things to provide. I need to create the oMiser directory, and an oFrugal directory for the definition and construction of those artifacts. I can register on SourceForge later, after I have done the work for DMware open-source and its license. I need to know how to make sidebars with text flow around them, just as for images on pages. I would like to have that all working in version 0.04. And I need to make a Miser Project logo and stop reusing the Orcmid's Lair one as a placeholder for one. I can change it, as time goes on, I can even change it as different stages of the Miser Project are taken on. But I need a logo now. So I am declaring today a Miser cleanup day while I rest and prepare to do some major hacking and hewing in the jungles of version 0.04. Remember to have 0.04 start using the images already here in the Note area, so that I don't have to redo those all of the time. Move the Visio versions of them over here too, for preservation along with the PNG forms.
2002-07-10-11:17: There needs to be a better way to indicate shifts in languaging in the sketch. Perhaps with sidebars (find out how to do that in HTML) and with indications of the domain that is being spoken in. At some point we will have the egg face the chicken (and vice versa), but lets keep it as clear as possible so that conversation can be grounded. I think it will be "turtles all the way down" but the bigger question is, what is it "all the way up?"
2002-07-10-11:14: It looks like we are building up to the following levels of Miser:
- oMiser, the fundamental mechanism of "pure" Obs, as being sketched. This, with oFrugal can be built and confirmed and tested. Even though it is not very appealing from the standpoint of practical application, it is important to provide a reference implementation and also build some working library code. More than that, it raises the issue of maintaining "purity" and what does and doesn't work, in that it is damaging to the theory (that is, creates an invalid interpretation). This will come up when I propose to make an ob-Close individual that takes an Ob and makes it an individual where its oApInt is the application of the Ob that it encloses. This is fun to describe, but it has problems with regard to identity. The question is how to have it and not have the problem with identity. Or show that is no problem. I think it forces incompleteness, and I want to see what can be done about that.
- sMiser, the symbolic version of Miser that deals with naming and "free variables" at the next level, yet works when we consider that there is only one "distributed" ob›. sFrugal is much more interesting to write, we can have namespaces, so we can remember things and have distribution work, and we can begin to demonstrate all sorts of things about Miser done in Miser. Lambda-definability comes up here, with lambda a program in Ob.
- cMiser, the imperative, computational Miser that has all of the odd stuff like assignment, exceptions, and whatnot. This is where we (begin) to deal with types and the typical. We'll have lots of demonstrations already, so now we can really look at what types and classes and so on might be in terms of computation, logic, and language.
- lMiser is for logical Miser, and it might be before or after cMiser or evolve along with it. That is the computational logic of Miser.
- iMiser, for interactive Miser, where we step outside of the effective computable and look at how do we address Miser computers as situated in the world and interacting in the world. No amount of inspection of individual Misers will reveal anything about that (i.e., in cMiser). Where is the theory of the system that interacting Misers (and people) form, and what is its manifestations?
- tMiser, maybe a better designation than lMiser, but about Miser having theories or in some sense manifesting theories and logic.
- Later, when it is time to develop a language that is not purely formal, such as Frugal, we will initiate eo and eoware.
- 0.10 2002-07-03 Create Initial Boilerplate and make placeholder for more (orcmid)
- I had a crash after doing great work updating the sketch. This led me to put in a history of the sketches as a safeguard against too much loss in the future and as a way to preserve the refinement of thinking and the pieces that were evaluated, changed, etc. I then captured the last-backed-up 0.01 here.
created 2002-07-03-14:10 -0700 (pdt) by orcmid
$$Author: Orcmid $
$$Date: 04-01-06 13:03 $
$$Revision: 11 $