Solar Soyuz Zaibatsu: Sandbox

Venimus, Vidimus, Dolavimus

Adapt, Adopt, Improvise

Guerrilla UNIX & Plan 9 Development

The street finds its own uses for things.

William Gibson

WARNING:

This is the online sandbox for SSZ. What this means is that this is where we develop and test tools, toys, and applications. It should not be considered stable or fit for use. The revision levels can range from alpha all the way to gold.

Some Unix/Linux/Plan 9 Related Resources

Other Interesting Resources

SPL Programming Language

Twine - A Semantic Web Social Site

4-21-2008

I'd heard about Twine over a year ago and sent off a request to participate. Recently I was notified I'd been accepted as a beta tester. While the site is still in very early development it is coming along quite nicely.

If you'd like to learn more then google Twine, also visit Wikipedia and search on Semantic Web. If you'd like to sign up for testing/early access visit Twine.

7-7-08

I've been using Twine for a couple of months now and in general it's a good front-end but I do not believe it will be a flagship Semantic Web product. At this point it lacks several key features like e-mail on/off-site, no real-time chat, PIM functionality, GUI is very inflexible with regard to building individual items or sharing items in bigger structures or relationships. One of the most irritating things is the amount of duplication of effort. Multiple twines on the same topic without a way to transitively join them. Also, it does not catch duplicate submissions of if sites, which means you get to see the same thing several times in different twines.

My biggest dissapointment to date however is that I have to input all the resources and such, and the system is very poor in cross-linking users to build conceptual communities. It really needs a mechanism for a user to crawl a site for new changes and automatically import them into Twine. It should further have some sort of natural language analysis to do automated topic searches into Google and (ideally user defined) other sites. It should also auto-cluster users who have similar tags and submitted resources.

The really minimal support for HTML is rather annoying if you want to create your own structures, and especially if you want to utilize off-site resources.

My bottom line is Twine is a good start but it has a long way to go. It should take as a primary goal with regard to the user experience the mandate that a user should never have to go off-Twine. There is supposed to be a major update in July, 2008. While I have stopped using it on a day to day basis I do intend to stop back around the end of July or early August and see how it's changed.

Igor - An experiment in using Eliza-like AI as an Agent Front-end

My thought here is to have a box on a webpage at the top. A user enters queries or requests in quasi-natural language and the agent then processes it and returns the item below the agent query box. The basic architecture is to build the query formatter in Javascript and off-load that to the users machine for resources. An agent sitting in cgi-bin would then take that request and process it. A response formatter would then build the response page and send it back to the users browser. I see there being different agents for different sorts of requests, so the code in cgi-bin may be nothing more than an agent selector.

Javascript will be used for client side processing, Perl & Python for server side processing. There is a Python library specificaly for Natural Language Processing that we'll be using.

This javascript version of ELIZA was originally written by Michal Wallace and significantly enhanced by George Dunlop.

My name is Igor, how can I help?
Input:

This is where the results of any actions Igor takes would normaly appear.

Scientific Open Office Edition (SOOe)

This is a project to combine Open Office, Python, SAGE, Tex, STIX Fonts, and other Open Source packages into a idea engine, using XML and RDF as intermediary file formats.

The motivation is that I keep a notebook of ideas, resources, and projects to the tune of several hundred pages a year. However, due to many limitations I am not able to follow up on much of it. I have been using Scientific Notebook for about 10 years now. This is a word processor that has Tex support and Maple built-in. This allows me to create documents of a very technical nature and use the Maple math engine to generate results and graphics in-line. It also has support for the web being able to both connect to websites and to export its documents in notebooks. The goal is to build an even more flexible tool using Open Source resources thus allowing almost anyone from jr. high upward to get value out of it.

Icon

My conception is of an obviously juvenile pig with a pink bow between her ears. Above her head is an illuminated light bulb. I'm not an artist so if anyone would like to take a shot at a freely redistributable image, please do so.

Business Plan

My initial goal with regard to budget is to obtain operating funds for the following (ideal):

Work on SOOe full time for 2 years to create the initial product and community building
Income of $3,000/mo for 24 months for living expenses, this is after taxes and deductions
$25,000 for the hardware, software, network access, and incidentals related to the development environment and project homesite
$10,000 for travel expenses as needed to work with component development groups and build the user community (ie attend conferences, user groups, etc.) per year

My initial project time line

Year 1

30 days to build homesite and development environments
30 days to familiarize myself with OO, PyUNO, and SAGE architecture and interfaces. Develop basic 'hello world' demos
60 days build first Python based interface between OO and SAGE and get SAGE to take the input, process it, and get that data back to OO for display to the user
90 days to refine the initial UI and glue code to create an integrated product
90 days to steamline the build and installation process and develop packaging
60 days to contact Linux vendors for inclusion in distributions, as well as other groups with the intent to develop quality release processes

Year 2:

First 6 months devoted to building community
Second 6 months developing the next release for the beginning of Year 3

NOTE: This is Plan "A" and with my current circumstances is not possible within my own resources. I am currently looking at alternate budgets, development timelines, and goal sets that would fall within my budget and time constraints.

Component Status

Open Office

It's there and working quite nicely. I need to get the development package and then get some basic hello world apps built to verify my understanding of extending the menu systems, connecting extenions, data conversion, and import/export to Sage. This looks to be my next major effort over the immediate future.

RVBurke "Simple python calc add-in example"

PyUNO

This is done and already integrated into Open Office. From what I've seen of it the install and startup process needs to be cleaned up and scripted.

STIX Fonts

I installed the beta fonts and they were quite nice, had some minor problems like decenders and such. The beta program is closed and the first gold release of the fonts is slated for July, 2008.

SAGE

In general SAGE is done. It's major shortcoming is the way the package installs, it is not friendly to system wide installs such as /usr/local/bin. It's config/startup script assumes it's in the build dir and the related directory tree is immediately available. I've experimented with a couple of approaches using links and such, but nothing that is really robust enough. I suspect I'll need to rewrite the startup script so that it will work as expected. It's graphical interface requires the user to open a browser to a very high numbered port. I need to spend some time looking at ways to integrate into a web server for easier redirect and hooking/wedging.

Javascript URL Manager

URL Manager

A test feed from Daily Rotation

Permission to use components of this website for non-commercial use is granted.