OSDV: Overview for the Developer Community
From OSDVwiki
Here’s the whole scoop on the OSDV, the mission, the who, what, why, the organization, the work product – basically a summary of everything you need to know to have enough context for work in the development community.
Contents |
[edit] Mission
The mission of the OSDV is to dramatically increase public trust in computing technology used to conduct elections in the U.S., thereby increasing public confidence in election processes and results.
[edit] Who We Are
The OSDV is a meritocratic group of technology and policy experts in the computer technology sector including members of academic, commercial, and public administration organizations. The OSDV project is conducted by a volunteer community that is supported by the OSDV Foundation’s funds and staff, including a team of experienced technology and policy folks who serve as architects and project leads.
More: You can learn more OSDV people in the wiki’s bio section, and about the OSDV developer community via the blog or just start reading at the wiki's Main Page.
[edit] What We Do
We do mainly two things: technology development work, and outreach activities to foster adoption of the technology results.
The developer community’s activities have the basic objective of building and demonstration systems that show how electronic election equipment can be built to be trusted, can be publicly vetted, state-certified, and used in voting precincts with the simplicity and accountability that’s needed for public confidence.
More: You can learn more about high-level rationale and goals from the FAQon OSDV’s activities. You can get details of the activities, plans, status, etc. from the wiki sections on the Labs, though some high level information is provided below.
[edit] How
The OSDV operates under an open source mandate. For us this means that all work is done through a collaborative, open, volunteer community to ensure complete transparency, public review, comment and participation, producing royalty-free publicly available results. That applies not only to availability of code, but to all work product (see below), and the transparency of how all work is conducted and all work product is created. We use the wiki to document everything that’s going on outside of the source code repository.
More: You can learn more about what we mean by “open source” and related matters in the FAQ.
[edit] Why
Why is it important to do work on computing technology used in voting? The main reason is the result of the current situation, where in most cases computing technology is poorly applied to voting. The result is an increase in voter confusion and concern, and a decrease in public confidence in elections and in the way elections are conducted using technology. We believe that most of this sad result is rooted in technical causes that can be corrected with primarily technical efforts, conducted in the open.
That answer leads to many more questions, but also a few key points. The questions include: What do we mean by poorly applied? Why do we believe in mainly technical corrections? Why are the current devices far from ideal? Answers are wide, varied, and to some extent available from OSDV blogs and wiki. But the key points are drivers for OSDV work. They are:
- Currently used voting devices were (for the most part) developed without any specifications, requirements, or guidelines on characteristics needed for public confidence.
- Currently used voting devices are proprietary closed systems that are difficult for anyone, even the technical experts, to understand or gain trust in.
- There is little practical guidance on how currently used voting devices should be assessed as fit for use, especially with respect to characteristics such as reliability, integrity, security, etc.
- There are no worked examples of voting devices developed to earn public trust.
The result is that it is very difficult in practice for anybody to build a trustworthy voting device, or to assess a device as potentially trustworthy.
That’s why we’ve chosen the approach and results (described below) to filling the resulting gaps: the lack of technical guidelines and specifications for trustworthy voting devices and services, and of worked examples of devices and of assessment of them.
[edit] Approach and Results
Our objective, simply stated, is to fill those gaps. Our approach combines elements of open source development and high-assurance systems. With respect to technical efforts, we use the term “T-Spec” to describe that approach. Filling the gaps includes these key activities in our approach:
- Assessing and resolving the root causes for the public confidence problems that exist in voting technology today,
- Developing guidelines for assessing, testing, and verifying whether an existing or new digital voting device, system, or service is truly high assurance and high veracity, and
- Defining, designing, and building a reference implementation of an open source high assurance digital voting system and service.
These activities create two results: three types of concrete work product as well as demonstrations of them.
- Specifications is term used to loosely cover a variety of types of documents including specifications, guidelines, architectures, designs, etc., all intended as guidance for the development of trustworthy, high-assurance voting devices.
- Software that is a reference implementation of specifications for a trustworthy, high-assurance voting device. Each reference implementation provides a worked example of how to build one kind of device, adhering to specifcations, and enabling independent assessment.
- Documented recommendations for conducting an independent review of a digital voting device or system, assessing whether it meets technical specifications and guidelinesfor trust and high assurance.
Demonstrating this work product consists of:
- Creating software to demonstrate the specifications;
- Creating demo systems using the software, and conducting public usage trials with them
- Performing assessment projects on demo systems, and documenting the feasibility of conducting such work.
In other words, results are both concrete work product that we write or code and activities to use the work product in a publicly visible or useful way, such as conducting technical demonstrations, assessment demonstrations, or building a demonstration digital voting service that meets or exceeds the guidelines, specifications, and potential standards for high assurance voting systems.
[edit] When
The first 12 months of operation of the developer community are intended to be focused on experimentation and parallel iterative development of both specifications (and other prescriptive documents) and reference implementations, mainly for the foundational technology for trustworthy voting technology, but also some relatively simple voting devices. Also included are tools for independent assessment. Either late in the first year or shortly thereafter, there should be demonstration assessment activities.
At the same time, non-technical efforts will be ongoing, to build awareness of the work, foster its adoption, get feedback, make interim results useful outside OSDV, etc.
Looking forward over a year is guesswork, based on expectations of results in the first year, but we expect that development work will continue towards more complete and robust reference implementations of a wider variety of types of devices, and more complete and useful versions of specs, assessment guidelines, etc.
